To see the other types of publications on this topic, follow the link: Model-free long-term prediction.
Journal articles on the topic 'Model-free long-term prediction'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 journal articles for your research on the topic 'Model-free long-term prediction.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
Szott, Wiesław, Piotr Łętkowski, Andrzej Gołąbek, and Krzysztof Miłek. "Modelling of the Long-Term Acid Gas Sequestration and Its Prediction: A Unique Case Study." Energies 13, no. 18 (September 9, 2020): 4701. http://dx.doi.org/10.3390/en13184701.
Full textAbstract:
A twenty-four-year on-going project of acid gas sequestration in a deep geological structure was subject to detailed modelling based upon a large set of geological, geophysical, and petrophysical data. The model was calibrated against available operational and monitoring data and used to determine basic characteristics of the sequestration process, such as fluid saturations and compositions, their variation in time due to fluid migrations, and the gas transition between free and aqueous phases. The simulation results were analysed with respect to various gas leakage risks. The contribution of various trapping mechanisms to the total sequestrated amount of injected gas was estimated. The observation evidence of no acid gas leakage from the structure was confirmed and explained by the simulation results of the sequestration process. The constructed and calibrated model of the structure was also used to predict the capacity of the analysed structure for increased sequestration by finding the optimum scenario of the risk-free sequestration performance.
2
Waldron, Levi, Paul A. Cooper, and Tony Y. Ung. "Prediction of long-term leaching potential of preservative-treated wood by diffusion modeling." Holzforschung 59, no. 5 (September 1, 2005): 581–88. http://dx.doi.org/10.1515/hf.2005.095.
Full textAbstract:
Abstract An approach to modeling leaching and leaching impacts of preservative components from treated wood is presented based on three simple laboratory determinations: the amount of preservative component available for leaching (Le), equilibrium dissociation of preservative into free water in wood (Di) and diffusion coefficients for component leaching in different wood directions (D). In this study, the following inorganic wood preservative systems were investigated: chromated copper arsenate (CCA), the copper component of copper azole (CA) and alkaline copper quaternary (ACQ), and boron in disodium octaborate tetrahydrate (DOT). Aggressive leaching of finely ground wood showed that amounts of preservative compounds available for leaching were highest for borates, followed by copper in copper amine systems and arsenic in CCA, copper in CCA and chromium in CCA. The equilibrium dissociation or solubility of components in free water in the wood was much higher for borates and copper amine, followed by copper and arsenic in CCA and chromium in CCA. Use of the applicable diffusion coefficient (D) and Di or Le values in a diffusion model allows the prediction of total amount leached and emission or flux rate at different times of exposure for products with different dimensions and geometries. The approach was tested and generally validated through application of the model to results of laboratory water spray leaching of full-size lumber samples. The approach explains the rapid leaching of boron compounds (large diffusion coefficient and high initial dissociated concentration) compared to other preservative components and predicts that ACQ will have higher initial leaching rates compared to CCA and CA, but the latter preservatives will continue to leach copper at a measurable rate for a much longer time. The practical implications and limitations of the approach are discussed.
3
Kim, Ji-Yeon, Kyunghee Park, Woong-Yang Park, Jeong Eon Lee, Seok Won Kim, Seok Jin Nam, Se-Kyung Lee, Zhengyan Kan, and Yeon Hee Park. "Genomic characteristics of breast cancer to predict response of neoadjuvant chemotherapy and long-term prognosis." Journal of Clinical Oncology 39, no. 15_suppl (May 20, 2021): 557. http://dx.doi.org/10.1200/jco.2021.39.15_suppl.557.
Full textAbstract:
557 Background: To precisely predict neoadjuvant chemotherapy (NAC) response and long-term prognosis, we developed prediction model with clinical and genomic characteristics of breast cancer (BC). Methods: We included early and locally advanced BC that would be scheduled to receive standard NAC (four cycles of anthracycline and cyclophosphamide and four cycles of docetaxel or docetaxel plus trastuzumab for HER2+ BC) followed by curative surgery. For each patient, tumor tissue and matched blood were prospectively collected three times: at diagnosis (T1), three weeks after the first cycle of chemotherapy (T2), and curative surgery (T3). Whole exome sequencing (WES) was performed to detect somatic mutation, mutational signature and tumor mutational burden (TMB) while RNASeq with PAM50 prediction was to classify intrinsic subtype. In terms of clinical variables, clinical stage and IHC subtype at diagnosis, residual cancer burden (RCB) class and distant recurrence free survival (DRFS) were used. Logistic regression was used for predicting RCB class with clinical and genomic variables at T1. Univariate and multivariate Cox regression were performed to identify prognostic factors for DRFS. Results: In total, 210 patients were enrolled and treated with NAC as scheduled. We successfully conducted WES in 231 BC tissues (T1:117, T2:101 and T3:13) from 117 patients. In NAC response, 13 patients were in RCB class 3, 39 in class 2, 14 in class 1 and 46 in class 0. Median follow up duration was 44months and distant recurrence was observed in 13 patients. TP53 mutation (68%) was the most commonly detected genetic alteration. ARID1A, CDH1, CSMD3, LRP1B, PIK3CA, RUNX1 and TP53 were significantly mutated genes in driver gene analysis. Median TMB was 87 (range, 14-570) and signature 3 was most frequently observed. Among genetic characteristics, high TMB was significantly associated with better NAC response compared with low TMB (hazard ratio[HR] for RCB class III: 0.11, 95% confident interval[CI]: 0.01, 0.74, p = 0.05). In prediction model, combination of seven variables: intrinsic subtype, TMB, LRRK1, OPLAH, and PIK3CA hotspot mutation, ERBB2 amplification, and clinical stage had 0.83 in area under curve (AUC) and 0.75 in accuracy. High clinical stage, PTEN and PIK3CA hotspot mutation negatively affected to DRFS while high TMB had protective effect (all ps < 0.05). Prediction model made with five variables: intrinsic subtype, TMB, PTEN mutation, PIK3CA hotspot mutation and clinical stage had 0.88 in c-index (95% CI: 0.81, 0.95). Conclusions: TMB, PIK3CA hotspot mutation and clinical stage showed predictive roles on NAC response and distant recurrence of BC in NAC setting. In prediction model, intrinsic subtype, TMB, LRRK1, OPLAH, and PIK3CA hotspot mutation, ERBB2 amplification, and clinical stage affected to RCB class while intrinsic subtype, TMB, PTEN, PIK3CA hotspot mutation and clinical stage did to DRFS. Clinical trial information: NCT02591966.
4
Huang, Yi-Cheng, and Yu-Hsien Chen. "Use of Long Short-Term Memory for Remaining Useful Life and Degradation Assessment Prediction of Dental Air Turbine Handpiece in Milling Process." Sensors 21, no. 15 (July 22, 2021): 4978. http://dx.doi.org/10.3390/s21154978.
Full textAbstract:
The complexity of the internal components of dental air turbine handpieces has been increasing over time. To make operations reliable and ensure patients’ safety, this study established long short-term memory (LSTM) prediction models with the functions of learning, storing, and transmitting memory for monitoring the health and degradation of dental air turbine handpieces. A handpiece was used to cut a glass porcelain block back and forth. An accelerometer was used to obtain vibration signals during the free running of the handpiece to identify the characteristic frequency of these vibrations in the frequency domain. This information was used to establish a health index (HI) for developing prediction models. The many-to-one and many-to-many LSTM frameworks were used for machine learning to establish prediction models for the HI and degradation trajectory. The results indicate that, in terms of HI predicted for the testing dataset, the mean square error of the many-to-one LSTM framework was lower than that that of a logistic regression model, which did not have a memory framework. Nevertheless, high accuracies were achieved with both of the two aforementioned approaches. In general, the degradation trajectory prediction model could accurately predict the degradation trend of the dental handpiece; thus, this model can be a useful tool for predicting the degradation trajectory of real dental handpieces in the future.
5
Duan, Jianchun, Hua Bai, Yiting Sun, Fei Gai, Shenya Tian, and Wenfang Xu. "Construction and validation of an eight-gene risk prediction model for stage II-IIIA lung adenocarcinoma." Journal of Clinical Oncology 39, no. 15_suppl (May 20, 2021): e20553-e20553. http://dx.doi.org/10.1200/jco.2021.39.15_suppl.e20553.
Full textAbstract:
e20553 Background: Clinical characters cannot precisely evaluate long-term survival of patients with resectable lung adenocarcinoma. Genomics studies of lung adenocarcinoma (LUAD) have advanced our understanding of LUAD's biology. Thus, genomics-based robust models predicting survival outcome for patients with operatable LUAD needs to be investigated. Here, we aimed to identify new gene signatures to construct a risk prediction model via integrating Omics data from The Cancer Genome Atlas (TCGA) to better evaluate the long-term clinical outcome of LUAD patients. Methods: A cohort of one hundred and eighty-nine stage II-IIIA lung adenocarcinoma cases receiving tumor resection were screened out and downloaded from TCGA database. Tumor samples without survival information and genes with low or no expression were removed. Genes associated with cancer and immune were further narrowed down using a Master Panel Gene Set (Amoydx). Lasso-Cox regression analysis was used to screen gene-survival outcome, and then a risk prediction model was established. LUAD cases were divided into high-risk or low-risk groups as per the scores, to assess differential expressed genes and pathways. Results: A total of 8 most survival outcome related genes (CLEC7A, PAX5, XCR1, KRT7, PLCG1, DKK1, CLEC10A, IKZF3) were identified after Lasso-Cox regression analysis and used for model construction. The overall survival (OS), progression-free survival (PFS) and disease-free survival (DFS) from the subgroups within the high- and low-risk groups were assessed and showed significant prolonged in low-risk group, the hazard ratio (HR) of OS was 2.72 (95%CI: 2.04-3.61, P = 5.91e-12) in high-risk group. Hierarchical clustering analysis, gene ontology (GO) analysis, gene set enrichment analysis (GSEA), and gene set variation analysis (GSVA) revealed that genes involved in immune responses were significantly suppressed in high-risk group, while as genes involved in antioxidative metabolism were activated, which gave us a hint that immune-metabolism interaction might play a vital role in determining the distal survival outcome of LUAD. Conclusions: Our risk prediction model enables precise evaluation of long-term survival for patients with LUAD. Further, it provides a novel and comprehensive understanding of biological impacts on LUAD prognosis, which offers new insights for future development of precise diagnostic and therapeutic approaches.[Table: see text]
6
Ali, Ghulam, Tariq Ali, Muhammad Irfan, Umar Draz, Muhammad Sohail, Adam Glowacz, Maciej Sulowicz, Ryszard Mielnik, Zaid Bin Faheem, and Claudia Martis. "IoT Based Smart Parking System Using Deep Long Short Memory Network." Electronics 9, no. 10 (October 15, 2020): 1696. http://dx.doi.org/10.3390/electronics9101696.
Full textAbstract:
Traffic congestion is one of the most notable urban transport problems, as it causes high energy consumption and air pollution. Unavailability of free parking spaces is one of the major reasons for traffic jams. Congestion and parking are interrelated because searching for a free parking spot creates additional delays and increase local circulation. In the center of large cities, 10% of the traffic circulation is due to cruising, as drivers nearly spend 20 min searching for free parking space. Therefore, it is necessary to develop a parking space availability prediction system that can inform the drivers in advance about the location-wise, day-wise, and hour-wise occupancy of parking lots. In this paper, we proposed a framework based on a deep long short term memory network to predict the availability of parking space with the integration of Internet of Things (IoT), cloud technology, and sensor networks. We use the Birmingham parking sensors dataset to evaluate the performance of deep long short term memory networks. Three types of experiments are performed to predict the availability of free parking space which is based on location, days of a week, and working hours of a day. The experimental results show that the proposed model outperforms the state-of-the-art prediction models.
7
D'Anna, Maurizio, Deborah Idier, Bruno Castelle, Goneri Le Cozannet, Jeremy Rohmer, and Arthur Robinet. "IMPACT OF UNCERTAINTIES IN MODEL FREE PARAMETERS AND SEA LEVEL RISE ON SHORELINE CHANGES: A 20-YEAR HINDCAST AT TRUC VERT BEACH, SW FRANCE." Coastal Engineering Proceedings, no. 36v (December 28, 2020): 10. http://dx.doi.org/10.9753/icce.v36v.management.10.
Full textAbstract:
Chronic erosion of sandy coasts is a continuous potential threat for the growing coastal communities worldwide. The prediction of shoreline evolution is therefore key issue for robust decision making worldwide, especially in the context of climate change. Shorelines respond to various complex processes interacting at several temporal and spatial scales, making shoreline reconstructions and predictions challenging and uncertain, especially on long time scales (e.g. decades or century). Despite the increasing progresses in addressing uncertainties related to the physics of Sea Level Rise, very little effort is made towards understanding and reducing the uncertainties related to wave driven coastal response. To fill this gap, we analyse the uncertainties associated with long-term (2 decades) modelling of the cross-shore transport dominated high-energy sandy coast around Truc Vert beach, SW France, which has been surveyed semi-monthly over the last 12 years.Recorded Presentation from the vICCE (YouTube Link): https://youtu.be/_NBJ2v-koMs
8
Lee, Junghoon, Min Soo Choo, Sangjun Yoo, Min Chul Cho, Hwancheol Son, and Hyeon Jeong. "Intravesical Prostatic Protrusion and Prognosis of Non-Muscle Invasive Bladder Cancer: Analysis of Long-Term Data over 5 Years with Machine-Learning Algorithms." Journal of Clinical Medicine 10, no. 18 (September 20, 2021): 4263. http://dx.doi.org/10.3390/jcm10184263.
Full textAbstract:
We aim to investigate the significance of intravesical prostate protrusion (IPP) on the prognosis of non-muscle invasive bladder cancer (NMIBC) after the transurethral resection of bladder tumors (TURBT). For newly diagnosed NMIBC, we retrospectively analyzed the association between prognosis and IPP for at least a 5-year follow-up. A degree of IPP over 5 mm in a preoperative CT scan was classified as severe. The primary endpoint was recurrence-free survival, and the secondary endpoint was progression-free survival. The machine learning (ML) algorithm of a support vector machine was used for predictive model development. Of a total of 122 patients, ultimately, severe IPP was observed in 33 patients (27.0%). IPP correlated positively with age, BPH, recurrence, and prognosis. Severe IPP was significantly higher in the recurrence group and reduced in the recurrence-free survival group (p = 0.038, p = 0.032). Severe IPP independently increased the risk of intravesical recurrence by 2.6 times. The addition of IPP to the known oncological risk factors in the prediction model using the ML algorithm improved the predictability of cancer recurrence by approximately 6%, to 0.803. IPP was analyzed as a potential independent risk factor for NMIBC recurrence and progression after TURBT. This anatomical feature of the prostate could affect the recurrence of bladder tumors.
9
Molot, Lewis A., and P. J. Dillon. "Nitrogen/Phosphorus Ratios and the Prediction of Chlorophyll in Phosphorus-Limited Lakes in Central Ontario." Canadian Journal of Fisheries and Aquatic Sciences 48, no. 1 (January 1, 1991): 140–45. http://dx.doi.org/10.1139/f91-019.
Full textAbstract:
The response of mean ice-free chlorophyll a in 15 stratified, P-limited oligotrophic and mesotrophic lakes in central Ontario to changes in mean epilimnetic total phosphorus (TPepi) within a lake was highly variable between years during the period 1976–87. The linear regression coefficient of determination, R2, using all annual means was only 0.36, and within-lake regressions revealed mostly random associations between chlorophyll a and TPepi. Neverthless, by using the long-term average of annual means for each lake, a bivariate linear regression model was developed relating the long-term, average response of chlorophyll a to the long-term, average TPepi concentration in these lakes (R2 = 0.78). Annual variation could not be explained by changes in epilimnetic total nitrogen to total phosphorus ratio (TN/TP). The R2 increased slightly from 0.78 to 0.82 with TN/TP as a second independent variable using long-term averages but remained at 0.78 with 1/TPepi as a second variable. Reanalysis of published data excluding lakes which were not P limited showed that TN/TP is of little or no benefit as an independent variable. A minimum of six consecutive years of sampling was required to avoid anomalously poor fits (defined as R2 < 0.6) for this set of lakes.
10
DeMaria, Mark. "A Simplified Dynamical System for Tropical Cyclone Intensity Prediction." Monthly Weather Review 137, no. 1 (January 1, 2009): 68–82. http://dx.doi.org/10.1175/2008mwr2513.1.
Full textAbstract:
Abstract A simplified dynamical system for tropical cyclone intensity prediction based on a logistic growth equation (LGE) is developed. The time tendency of the maximum sustained surface winds is proportional to the sum of two terms: a growth term and a term that limits the maximum wind to an upper bound. The maximum wind evolution over land is determined by an empirical inland wind decay formula. The LGE contains four free parameters, which are the time-dependent growth rate and maximum potential intensity (MPI), and two constants that determine how quickly the intensity relaxes toward the MPI. The MPI is estimated from an empirical formula as a function of sea surface temperature and storm translational speed. The adjoint of the LGE provides a method for finding the other three free parameters to make the predictions as close as possible to the National Hurricane Center best-track intensities. The growth rate is assumed to be a linear function of the vertical shear (S), a convective instability parameter (C) determined from an entraining plume, and their product, where both S and C use global model fields as input. This assumption reduces the parameter estimation problem to the selection of six constants. Results show that the LGE optimized for the full life cycle of individual storms can very accurately simulate the intensity variations out to as long as 15 days. For intensity prediction, single values of the six constants are found by fitting the model to more than 2400 Atlantic forecasts from 2001 to 2006. Results show that the observed intensity variations can be fit more accurately with the LGE than with the linear Statistical Hurricane Intensity Prediction Scheme (SHIPS) formulation, and with a much smaller number of constants. Results also show that LGE model solution (and some properties of real storms) can be explained by the evolution in the two-dimensional S–C phase space. Forecast and other applications of the LGE model are discussed.
11
Lei, Lei, Tzu-Ting Huang, Andre Ching-Hsuan Chen, Tzu-Pin Lu, and Skye Hung-Chun Cheng. "Prognostic value of a new clinical-genomic model to predict 10-year risk of recurrence in patients with operable breast cancer." Journal of Clinical Oncology 37, no. 15_suppl (May 20, 2019): 530. http://dx.doi.org/10.1200/jco.2019.37.15_suppl.530.
Full textAbstract:
530 Background: Searching for a specific biomarker to predict long-term risk of recurrence for all breast cancer subtypes is challenging. DGM-CM6 (Distant Genetic Model-Clinical variable Model 6) is a new clinical-genomic prognostic model developed from the 18-gene panel which was reported previously. This study aims to validate the long-term prognostic value of this new model in all subtypes of operable breast cancer patients. Methods: We included 752 operable breast cancer patients with stage I-III in all subtypes treated in a Cancer Center from 2005 to 2014 as the internal validation (IV) cohort. The median follow-up was 94.1 months. Meanwhile, Affymetrix U133P2 (n = 1139) data obtained from GEO (GSE9195/16391/17907/19615/20711/21653/42568, EMTAB365) were collected as the external validation (EV) dataset. The prognostic effect of DGM-CM6 was then evaluated by uni- and multivariate analyses. The low- and high-risk patients ( < 33 or ≥ 33 as cut-off value) classified by DGM-CM6 were evaluated by the 10-year distant relapse-free interval (DRFI), relapse-free interval (RFI), relapse-free survival (RFS) and distant relapse-free survival (DRFS), respectively. We further compared the predictive performance between DGM-CM6/DGM and PAM50-ROR score in our IV dataset. Results: In the IV dataset, DGM-CM6 was proved to be an independent prognostic factor by multivariate analysis with hazard ratios of 3.1 (1.6-6.0) for RFS (P = 0.0009) and 3.2 (1.6-6.3) for DRFS (P = 0.0009). Significant differences were observed between low- and high-risk groups with 10-year RFI (94.0% vs. 83.5%, P < 0.0001), RFS (90.0% vs. 80.5%, P = 0.0003), DRFI (94.1% vs. 85.0%, P < 0.0001), and DRFS (90.1% vs. 81.9%, P = 0.0004), respectively. The prognostic value of RFS was convinced in the EV dataset (HR = 1.34, P = 0.00052) by the DGM only. According to C-index estimate analysis, DGM appeared to have better performance comparing with PAM50 ROR score in prediction of long-term DR, DRFS, RFI, and RFS in N0 patients (C index for distant recurrence: 0.582 by DGM, 0.528 by ROR). Conclusions: DGM-CM6 could be a new long-term prognostic model to be applied in all subtypes of operable breast cancer patients. Further validation in a large scale of clinical trials is needed.
12
Ko, Bor-Sheng, Yu-Fen Wang, Chih-Chuan Lu, Jeng-Lin Li, Chi-Chun Lee, Jih-Luh Tang, and Hwei-Fang Tien. "Relapse and Mortality Prediction of Acute Myeloid Leukemia Patients Using Deep Bidirectional Long Short-Term Memory-Deep Neural Network Architecture." Blood 132, Supplement 1 (November 29, 2018): 2811. http://dx.doi.org/10.1182/blood-2018-99-115778.
Full textAbstract:
Abstract Introduction While continuous advancement has been made in the treatment of AML, the overall 5-year survival rate achieved in current days is still less than 40%. The ability to accurately assess AML patient's risk of mortality and relapse is crucial to act on the most appropriate comprehensive treatment strategy. However, the surge of clinical parameters and the scale of data collection are becoming too complex for us to stratify risk and deriving predictive model systematically with conventional statistical methods. In our study, we propose to utilize artificial intelligence techniques in creating risk stratification and outcome prediction model through retrospective data analysis. Method Retrospective clinical data of patients with AML, including basic demographics (age & gender), laboratory results of complete blood count (CBC), white blood composition (WBC) and cytogenetics, and treatment history such as Hematopoietic Stem Cell Transplant (HSCT) and other medication history, was collected at the National Taiwan University Hospital. A total of 479 patients with newly diagnosed de novo AML were enrolled in this study. The median age at diagnosis was 50 years, and the median follow-up duration was 26.6 months, 77.7% of them achieved CR and, 53.4% had relapsed, and 47.2% had received HSCT. In total, 479 cytogenetics tests at diagnosis, 43,518 CBC & WBC records, 28 types of anti-neoplastic medications in L01 & L03 ATC code category with a total of 69,546 medication records from these 479 AML patients were used in deriving the outcome prediction model. A vectorized representation that captures the static-dynamic clinical aspects, i.e., demographics, laboratory results, and treatments, of an AML patient can be learned directly from the collected data. The representation includes both static personal attributes (demographic, and cytogenetic) and time-varying progression of patient's clinical assessment across time (laboratory results, HSCT, and medication). The time-dependent representation was derived from training a deep network architecture of bi-directional long short-term memory network (BLSTM). By taking 10 days as a time step, the BLSTM took the input of Fisher-vector encoded time series of a patient's CBC and WBC, medication, and HSCT records separately. The last output layer, which summarized the relapse/mortality risk exhibited in the measurements of patient's clinical conditions over time, of each separately-learned BLSTM was used as the encoded dynamic clinical representation. The concatenation of the AML patient's static features and the time-dependent representations were fed into a deep neural network followed by a support vector machine to carry out the final prediction. The prediction models were conducted in 5-fold cross-validation experiments and further evaluated using metrics of accuracy (ACC). Results The median leukemia-free survival and median overall survival of these 479 patients are 7.2 months and 49.0 months respectively. By using the CBC & WBC data, the accuracy of next 3-months relapse and mortality prediction at any time point reached 77.9 and 77.0% respectively, while models built with CBC, WBC combined with medication records reached accuracies of 78.4% and 84.6% respectively. Furthermore, incorporating static features, i.e., demographics and cytogenetics, together with time series representation of CBC & WBC, and medication records in our prediction model, we achieved further improved performances of 82.2% and 85.0% accuracies in next 3-month relapse and mortality prediction respectively (Figure 2). Conclusions The BLSTM-DNN model is a novel approach that is capable of jointly taking into account multiple heterogeneous clinical measurements throughout the clinical courses of AML patients in order to derive relapse and mortality prediction. It could be applied at any given time point using past 3-month data to predict the next 3-month relapse or mortality. We also observed the model accuracy increased as we increased the number of clinical exams included in our model. Our study results demonstrated a potential model in facilitating precise and personalized risk assessment that could support physicians for better risk-averse intervention in the future. We plan to integrate other exams including genomic and pathology reports with a larger patient's cohort into our model in the next phase of model development. Disclosures Ko: GNT Biotech & Medicals Crop.: Research Funding; Roche: Research Funding; Abbevie: Research Funding; Mumdipharma Taiwan: Consultancy.
13
Zhou, Mengxi, Liyun Shen, Qin Jiao, Lei Ye, Yulin Zhou, Wei Zhu, Weiqing Wang, and Shu Wang. "ROLE OF MAGNETIC RESONANCE IMAGING IN THE ASSESSMENT OF ACTIVE THYROID-ASSOCIATED OPHTHALMOPATHY PATIENTS WITH LONG DISEASE DURATION." Endocrine Practice 25, no. 12 (December 2019): 1268–78. http://dx.doi.org/10.4158/ep-2019-0133.
Full textAbstract:
Objective: In thyroid-associated ophthalmopathy (TAO), long disease duration is negatively correlated with the response to immunosuppression treatment. The current treatment decision-making process does not involve magnetic resonance imaging (MRI); thus, we investigated the predictive value of MRI parameters for the immunosuppressive response in active moderate to severe TAO patients with different disease durations. Methods: We retrospectively analyzed the baseline MRI parameters of active TAO patients treated with guideline-recommended weekly glucocorticoid therapy in our center. Data were stratified by the quartile of disease duration. The signal intensity ratio (SIR) of T2-weighted images was used to describe the activity of extraocular muscles (EOMs). Results: Compared to the lowest quartile of disease duration, SIR values of EOMs were significantly lower in quartile 3 (Q3) and quartile 4 (Q4). Meanwhile, the clinical activity score (CAS) curve did not change in parallel and was not correlated with the SIR curve. In the highest quartile of disease duration, nonresponders had significantly lower SIR values of the most inflamed muscle ( P = .03) and the medial rectus ( P = .004) than did the responders, while no such significance was observed in patients within the lower 3 quartiles. A multivariable predictive model (including CAS, TAO duration, and SIR value) was established in each quartile. The fit of the model was better than CAS with regard to prognostic prediction and showed a high positive predictive value (Model 1: 86.67%; Model 2: 92.86%) and negative predictive value (Model 1: 88.89%; Model 2: 90%) in the top quartile. Conclusion: The anterior manifestation assessed by CAS is not always consistent with retro-orbital activity in long-term TAO patients. CAS is sufficient to reflect disease activity in short-term TAO patients. The supplementation of CAS with orbital MRI would be valuable in selecting appropriate active patients with a long disease duration. Abbreviations: AUC = area under the curve; CAS = clinical activity score; EOM = extraocular muscle; FT3 = free triiodothyronine; FT4 = free thyroxine; GC = glucocorticoid; ivGC = intravenous glucocorticoids; MRI = magnetic resonance imaging; NPV = negative predictive value; PPV = positive predictive value; SIR = signal intensity ratio; TAO = thyroid-associated ophthalmopathy; TRAb = thyroid-stimulating hormone receptor antibody; TSH = thyroid-stimulating hormone
14
Zhu, E., and D. Pi. "Photovoltaic Generation Prediction of CCIPCA Combined with LSTM." Complexity 2020 (September 15, 2020): 1–11. http://dx.doi.org/10.1155/2020/1929372.
Full textAbstract:
In order to remedy problems encompassing large-scale data being collected by photovoltaic (PV) stations, multiple dimensions of power prediction mode input, noise, slow model convergence speed, and poor precision, a power prediction model that combines the Candid Covariance-free Incremental Principal Component Analysis (CCIPCA) with Long Short-Term Memory (LSTM) network was proposed in this study. The corresponding model uses factor correlation coefficient to evaluate the factors that affect PV generation and obtains the most critical factor of PV generation. Then, it uses CCIPCA to reduce the dimension of PV super large-scale data to the factor dimension, avoiding the complex calculation of covariance matrix of algorithms such as Principal Component Analysis (PCA) and to some extent eliminating the influence of noise made by PV generation data acquisition equipment and transmission equipment such as sensors. The training speed and convergence speed of LSTM are improved by the dimension-reduced data. The PV generation data of a certain power station over a period is collected from SolarGIS as sample data. The model is compared with Markov chain power generation prediction model and GA-BP power generation prediction model. The experimental results indicate that the generation prediction error of the model is less than 3%.
15
Ding, Hangxing, Song Chen, Shuai Chang, Guanghui Li, and Lei Zhou. "Prediction of Surface Subsidence Extension due to Underground Caving: A Case Study of Hemushan Iron Mine in China." Mathematical Problems in Engineering 2020 (May 30, 2020): 1–10. http://dx.doi.org/10.1155/2020/5086049.
Full textAbstract:
Underground caving can potentially lead to large-scale surface destruction. To test the safety conditions of the surface construction projects near the circular surface subsidence zone in the Hemushan Iron Mine, this paper proposes an analytical model to analyze the stability of the cylindrical caved space by employing the long-term strength of the surrounding rock mass, the in situ stress, and the impact of caved materials as inputs. The proposed model is valid for predicting the orientation and depth where rock failure occurs and for calculating the maximum depth of the undercut, above which the surrounding rock mass of the caved space can remain stable for a long duration of time. The prediction for the Hemushan Iron Mine from the proposed model reveals that the construction projects can maintain safe working conditions, and such prediction is also demonstrated by the records from Google Earth satellite images. This means that the proposed model is valid for conducting such analysis. Additionally, to prevent rock failure above the free surface of caved materials, backfilling the subsidence zone with waste rocks is suggested, and such a measure is implemented in the Hemushan Iron Mine. The monitoring results show that this measure contributes to protecting the surrounding wall of the caved space from large-scale slip failure. The contribution of this work not only provides a robust analytical model for predicting the stability of rock around a cylindrical caved space but also introduces employable measures for mitigating the subsequent extension of surface subsidence after vertical caving.
16
Hattori, Masaya, Keitaro Matsuo, Mari Ichikawa, Takashi Fujita, Masataka Sawaki, Naoto Kondo, Akiyo Horio, Yasushi Yatabe, and Hiroji Iwata. "Distant disease-free survival (DDFS) discrimination capacity of various pathologic complete response (pCR) definitions according to breast cancer subtypes." Journal of Clinical Oncology 31, no. 26_suppl (September 10, 2013): 162. http://dx.doi.org/10.1200/jco.2013.31.26_suppl.162.
Full textAbstract:
162 Background: pCR has been postulated to be correlated with long-term clinical benefits in some subtypes of breast cancer. Here, we analyzed the discriminatory ability and the predictive power of various pCR definitions for distant disease-free survival (DDFS) according to breast cancer subtypes. Methods: We analyzed 326 (114 Luminal A: ER+/PR+/HER2-, 44 Luminal B/HER2-: ER+/PR-/HER2-, 51 Luminal B/HER2+: ER+/PR+ and/or -/HER2+, 51 HER2: ER-/PR-/HER2+, and 66 Triple negative: ER-/PR-/HER2-) non-metastatic breast cancer patients (pts) who had received neoadjuvant chemotherapy at our institution between January 2003 and June 2012. Four pCR definitions were used: ypT0ypN0, ypT0/isypN0, ypT0/isypN0/+, ypT<1micypN0/+. DDFS was estimated by Kaplan-Meier method, and analyzed by log-rank test and Cox proportional hazard model. The receiver operating characteristic (ROC) curves analysis was used for comparing DDFS prediction models with and without various pCR definitions in addition to other covariates (tumor stage, nodal status, BMI, tumor grade, use of trastuzumab) as variables. Results: The pCR rate was comparatively low in Luminal A and high in HER2. 94.1% of HER2 and 74.5% of Luminal B/HER2+ received total 1 year of trastuzumab therapy. In multivariate analysis, no pCR definitions were associated with improved DDFS significantly in Luminal A, Luminal B/HER2-, Luminal B/HER2+ and HER2, whereas each pCR definition was associated with improved DDFS in Triple negative (ypT0ypN0: HR0.12, p=0.043, ypT0/isypN0: HR0.06, p=0.007, ypT0/isypN0/+: HR0.107, p=0.004, ypT<1micypN0/+: HR0.104, p=0.003). In the ROC curves analysis of triple-negative, a DDFS prediction model including pCR defined as ypT0/isypN0 showed the highest accuracy, but low statistical significance (AUC: 0.834 vs.0.749 p=0.076). Conclusions: pCR could discriminate good and poor prognosis groups only in Triple negative and pCR defined as ypT0/isypN0 has the potential to provide better discrimination in this subtype. The predictive power of pCR for long term clinical benefit in other subtypes may not be obvious due to the influence of effective adjuvant therapies.
17
Elfarhani, Makram, Ali Mkaddem, Saeed Rubaiee, Abdessalem Jarraya, and Mohamed Haddar. "Prediction of foam impulse response through combination of hereditary and fractional derivative approaches." Multidiscipline Modeling in Materials and Structures 15, no. 4 (July 1, 2019): 800–817. http://dx.doi.org/10.1108/mmms-10-2018-0164.
Full textAbstract:
Purpose The purpose of this paper is to cover an experimental investigation of the impulse response of the foam-mass system (FMS) to unveil some of the foam dynamic behavior features needed to optimize the impact comfort of seat-occupant system. The equation of motion of the studied system is modeled as a sum of a linear elastic, pneumatic damping and viscoelastic residual forces. An identification methodology based on two separated calibration processes of the viscoelastic parameters was developed. Design/methodology/approach The viscoelastic damping force representing the foam short memory effects was modeled through the hereditary formulation. Its parameters were predicted from the free vibrational response of the FMS using iterative Prony method for autoregressive–moving–average model. However, the viscoelastic residual force resulting in the long memory effects of the material was modeled with fractional derivative term and its derivative order was predicted from previous cyclic compression standards. Findings The coefficients of the motion law were determined using closed form solution approach. The predictions obtained from the simulations of the impulse and cyclic tests are reasonably accurate. The physical interpretations as well as the mathematical correlations between the system parameters were discussed in details. Originality/value The prediction model combines hereditary and fractional derivative formulations resulting in short and long physical memory effects, respectively. Simulation of impulse and cyclic behavior yields good correlation with experimental findings.
18
Fan, Yanghua, Yichao Li, Xinjie Bao, Huijuan Zhu, Lin Lu, Yong Yao, Yansheng Li, et al. "Development of Machine Learning Models for Predicting Postoperative Delayed Remission in Patients With Cushing’s Disease." Journal of Clinical Endocrinology & Metabolism 106, no. 1 (October 1, 2020): e217-e231. http://dx.doi.org/10.1210/clinem/dgaa698.
Full textAbstract:
Abstract Context Postoperative hypercortisolemia mandates further therapy in patients with Cushing’s disease (CD). Delayed remission (DR) is defined as not achieving postoperative immediate remission (IR), but having spontaneous remission during long-term follow-up. Objective We aimed to develop and validate machine learning (ML) models for predicting DR in non-IR patients with CD. Methods We enrolled 201 CD patients, and randomly divided them into training and test datasets. We then used the recursive feature elimination (RFE) algorithm to select features and applied 5 ML algorithms to construct DR prediction models. We used permutation importance and local interpretable model–agnostic explanation (LIME) algorithms to determine the importance of the selected features and interpret the ML models. Results Eighty-eight (43.8%) of the 201 CD patients met the criteria for DR. Overall, patients who were younger, had a low body mass index, a Knosp grade of III–IV, and a tumor not found by pathological examination tended to achieve a lower rate of DR. After RFE feature selection, the Adaboost model, which comprised 18 features, had the greatest discriminatory ability, and its predictive ability was significantly better than using Knosp grading and postoperative immediate morning serum cortisol (PoC). The results obtained from permutation importance and LIME algorithms showed that preoperative 24-hour urine free cortisol, PoC, and age were the most important features, and showed the reliability and clinical practicability of the Adaboost model in DC prediction. Conclusions Machine learning–based models could serve as an effective noninvasive approach to predicting DR, and could aid in determining individual treatment and follow-up strategies for CD patients.
19
Patrick, Graeme, and Haithem Soliman. "Roughness prediction models using pavement surface distresses in different Canadian climatic regions." Canadian Journal of Civil Engineering 46, no. 10 (October 2019): 934–40. http://dx.doi.org/10.1139/cjce-2018-0697.
Full textAbstract:
The correlation between the international roughness index (IRI) and distress is inherent, as roughness is a function of both the changes in elevation of the distress-free pavement surface and the changes in elevation due to existing surface distress. In this way, a relationship between existing surface distress and IRI may be developed. However, the susceptibility of pavement to various types of surface distress is affected by many factors, including climatic conditions. A model that relates pavement surface distress to IRI for Canada needs to account for climatic conditions in different locations. This paper investigates the relationship between pavement surface distresses and IRI for different climatic conditions in Canada using historical data collected at numerous pavement test section locations sourced from the Long-Term Pavement Performance program database. Developed models were calibrated then validated and found to be statistically significant.
20
Arena, Felice, Giuseppe Barbaro, and Alessandra Romolo. "Return Period of a Sea Storm with at Least Two Waves Higher than a Fixed Threshold." Mathematical Problems in Engineering 2013 (2013): 1–6. http://dx.doi.org/10.1155/2013/416212.
Full textAbstract:
Practical applications in ocean engineering require the long-term analysis for prediction of extreme waves, that identify design conditions. If extreme individual waves are investigated, we need to combine long-term statistical analysis of ocean waves with short-term statistics. The former considers the distribution of standard deviation of free surface displacement in the considered location in a long-time span, of order of 10 years or more. The latter analyzes the distribution of individual wave heights in a sea state, which is a Gaussian process in time domain. Recent advanced approaches enable the combination of the two analyses. In the paper the analytical solution is obtained for the return period of a sea storm with at least two individual waves higher than a fixed level. This solution is based on the application of the Equivalent Triangular Storm model for the representation of actual storms. One of the corollaries of the solution gives the exact expression for the probability that at least two waves higher than fixed level are produced during the lifetime of a structure. The previous solution of return period and the relative probability of exceedance may be effectively applied for the risk analysis of ocean structures.
21
Ding, Ning, Chao Yang, Shoshana H. Ballew, Corey A. Kalbaugh, John W. McEvoy, Maya Salameh, David Aguilar, et al. "Fibrosis and Inflammatory Markers and Long-Term Risk of Peripheral Artery Disease." Arteriosclerosis, Thrombosis, and Vascular Biology 40, no. 9 (September 2020): 2322–31. http://dx.doi.org/10.1161/atvbaha.120.314824.
Full textAbstract:
Objective: Inflammatory markers, such as hs-CRP (high-sensitivity C-reactive protein), have been reported to be related to peripheral artery disease (PAD). Galectin-3, a biomarker of fibrosis, has been linked to vascular remodeling and atherogenesis. However, its prospective association with incident PAD is unknown; as is the influence of inflammation on the association between galectin-3 and PAD. Approach and Results: In 9851 Atherosclerosis Risk in Communities Study participants free of PAD at baseline (1996–1998), we quantified the association of galactin-3 and hs-CRP with incident PAD (hospitalizations with PAD diagnosis [ International Classification of Diseases - Ninth Revision : 440.2–440.4] or leg revascularization [eg, International Classification of Diseases - Ninth Revision : 38.18]) as well as its severe form, critical limb ischemia (PAD cases with resting pain, ulcer, gangrene, or leg amputation) using Cox models. Over a median follow-up of 17.4 years, there were 316 cases of PAD including 119 critical limb ischemia cases. Log-transformed galectin-3 was associated with incident PAD (adjusted hazard ratio, 1.17 [1.05–1.31] per 1 SD increment) and critical limb ischemia (1.25 [1.05–1.49] per 1 SD increment). The association was slightly attenuated after further adjusting for hs-CRP (1.14 [1.02–1.27] and 1.22 [1.02–1.45], respectively). Log-transformed hs-CRP demonstrated robust associations with PAD and critical limb ischemia even after adjusting for galectin-3 (adjusted hazard ratio per 1 SD increment 1.34 [1.18–1.52] and 1.34 [1.09–1.65], respectively). The addition of galectin-3 and hs-CRP to traditional atherosclerotic predictors (C statistic of the base model 0.843 [0.815–0.871]) improved the risk prediction of PAD (ΔC statistics, 0.011 [0.002–0.020]). Conclusions: Galectin-3 and hs-CRP were independently associated with incident PAD in the general population, supporting the involvement of fibrosis and inflammation in the pathophysiology of PAD.
22
Jaspers, Nicole E. M., Michael J. Blaha, Kunihiro Matsushita, Yvonne T. van der Schouw, Nicholas J. Wareham, Kay-Tee Khaw, Marie H. Geisel, et al. "Prediction of individualized lifetime benefit from cholesterol lowering, blood pressure lowering, antithrombotic therapy, and smoking cessation in apparently healthy people." European Heart Journal 41, no. 11 (May 18, 2019): 1190–99. http://dx.doi.org/10.1093/eurheartj/ehz239.
Full textAbstract:
Abstract Aims The benefit an individual can expect from preventive therapy varies based on risk-factor burden, competing risks, and treatment duration. We developed and validated the LIFEtime-perspective CardioVascular Disease (LIFE-CVD) model for the estimation of individual-level 10 years and lifetime treatment-effects of cholesterol lowering, blood pressure lowering, antithrombotic therapy, and smoking cessation in apparently healthy people. Methods and results Model development was conducted in the Multi-Ethnic Study of Atherosclerosis (n = 6715) using clinical predictors. The model consists of two complementary Fine and Gray competing-risk adjusted left-truncated subdistribution hazard functions: one for hard cardiovascular disease (CVD)-events, and one for non-CVD mortality. Therapy-effects were estimated by combining the functions with hazard ratios from preventive therapy trials. External validation was performed in the Atherosclerosis Risk in Communities (n = 9250), Heinz Nixdorf Recall (n = 4177), and the European Prospective Investigation into Cancer and Nutrition-Netherlands (n = 25 833), and Norfolk (n = 23 548) studies. Calibration of the LIFE-CVD model was good and c-statistics were 0.67–0.76. The output enables the comparison of short-term vs. long-term therapy-benefit. In two people aged 45 and 70 with otherwise identical risk-factors, the older patient has a greater 10-year absolute risk reduction (11.3% vs. 1.0%) but a smaller gain in life-years free of CVD (3.4 vs. 4.5 years) from the same therapy. The model was developed into an interactive online calculator available via www.U-Prevent.com. Conclusion The model can accurately estimate individual-level prognosis and treatment-effects in terms of improved 10-year risk, lifetime risk, and life-expectancy free of CVD. The model is easily accessible and can be used to facilitate personalized-medicine and doctor–patient communication.
23
Basu, Devraj, David Shimunov, Roger B. Cohen, Alexander Lin, Samuel Swisher-McClure, John Nicholas Lukens, Joshua Bauml, et al. "Outcomes and prediction of lethal recurrence after transoral robotic surgery for HPV+ head and neck cancer." Journal of Clinical Oncology 39, no. 15_suppl (May 20, 2021): 6047. http://dx.doi.org/10.1200/jco.2021.39.15_suppl.6047.
Full textAbstract:
6047 Background: Increasing use of transoral robotic surgery (TORS) for human papilloma virus-related (HPV+) head and neck squamous cell carcinomas (HNSCCs) is likely to impact recurrence patterns and outcomes. Profiling HPV+ HNSCC recurrences after TORS and identifying features predictive of lethal outcome would facilitate tailoring adjuvant therapy and guide surveillance post-therapy. This study uses long term follow-up of patients at the first institution to bring TORS into clinical use to describe the recurrence patterns, distinguish outcomes associated with distinct patterns, and create a risk model for lethal recurrence. Methods: This retrospective cohort study at a single academic tertiary center analyzed 634 consecutive, treatment-naïve HPV+ HNSCC patients receiving TORS and neck dissection for clinical features at presentation and pathologic traits identified by surgical resection. The main outcomes were distant metastatic recurrence (DMR) and locoregional recurrence (LRR). Multivariate logistic regression with backward stepwise elimination was used to identify features associated with recurrence. Results: 6.5% of patients developed DMR at a median of 12.4 months after surgery and had a 5-year overall survival (OS) of 52.5% (95% CI, 33.9%-68.2%), whereas the 6.2% patients developing LRR alone had 5-year OS of 83.3% (95% CI, 66.2%-92.2%; P =.01). After recurrence, 5-year progression-free survival was 24.7% (95% CI, 11.4%-40.7%) for DMR cases and 85.7% (95% CI, 65.1-94.6%) for cases with LRR alone (P <.001). Comparing recurrent cases to recurrence-free controls showed DMR to be independently associated with positive surgical margins (AOR 5.7; 95% CI, 2.1-15.7) and advanced clinical stage at presentation (AOR 6.5; 95% CI, 1.9-23.0). Positive margins increased DMR risk by 4.2-fold and reduced 5-year disease-free survival (P <.001) in early-stage cases (Table), which comprised 95% of the cohort. By contrast, isolated LRR was associated with failure to receive indicated adjuvant therapy and was usually controllable by salvage therapy. Conclusions: Based on the largest single institution cohort reported to date, long term oncologic outcomes for HPV+ HNSCCs after TORS are excellent overall. While DMR is often fatal, LRR is salvageable with durable disease control. In addition to standard staging criteria, positive margins indicate substantially higher risk of DMR but not LRR. A risk model for DMR that incorporates margin status after TORS is relevant for guiding clinical trial design and whole-body surveillance.[Table: see text]
24
Hughes, Maria F., Francisco Ojeda, Olli Saarela, Torben Jørgensen, Tanja Zeller, Tarja Palosaari, Mark G. O'Doherty, et al. "Association of Repeatedly Measured High-Sensitivity–Assayed Troponin I with Cardiovascular Disease Events in a General Population from the MORGAM/BiomarCaRE Study." Clinical Chemistry 63, no. 1 (January 1, 2017): 334–42. http://dx.doi.org/10.1373/clinchem.2016.261172.
Full textAbstract:
Abstract BACKGROUND High-sensitivity troponin I (hs-cTnI) concentrations reflect myocardial stress. The role of hs-cTnI in predicting long-term changes in the risk of cardiovascular disease (CVD) in general populations is not clearly defined. METHODS We investigated whether the change in 3 repeated measures of hs-cTnI collected 5 years apart in a prospective Danish study (3875 participants, initially aged 30–60 years, 51% female, disease free at baseline) improves 10-year prediction of incident CVD compared to using a single most recent hs-cTnI measurement. The change process was modelled using a joint (longitudinal and survival) model and compared to a Cox model using a single hs-cTnI measure adjusted for classic CVD risk factors, and evaluated using discrimination statistics. RESULTS Median hs-cTnI concentrations changed from 2.6 ng/L to 3.4 ng/L over 10 years. The change in hs-cTnI predicts 10-year risk of CVD (581 events); the joint model gave a hazard ratio of 1.31 per interquartile difference in hs-cTnI (95% CI 1.15–1.48) after adjustment for CVD risk factors. However, the joint model performed only marginally better (c-index improvement 0.0041, P = 0.03) than using a single hs-cTnI measure (c-index improvement 0.0052, P = 0.04) for prediction of CVD, compared to a model incorporating CVD risk factors without hs-cTnI (c-index 0.744). CONCLUSIONS The change in hs-cTnI in 5-year intervals better predicts risk of CVD in the general population, but the most recent measure of hs-cTnI, (at 10 years) is as effective in predicting CVD risk. This simplifies the use of hs-cTnI as a prognostic marker for primary prevention of CVD in the general population.
25
Morgan, Victoria L., Baxter P. Rogers, Adam W. Anderson, Bennett A. Landman, and Dario J. Englot. "Divergent network properties that predict early surgical failure versus late recurrence in temporal lobe epilepsy." Journal of Neurosurgery 132, no. 5 (May 2020): 1324–33. http://dx.doi.org/10.3171/2019.1.jns182875.
Full textAbstract:
OBJECTIVEThe objectives of this study were to identify functional and structural network properties that are associated with early versus long-term seizure outcomes after mesial temporal lobe epilepsy (mTLE) surgery and to determine how these compare to current clinically used methods for seizure outcome prediction.METHODSIn this case-control study, 26 presurgical mTLE patients and 44 healthy controls were enrolled to undergo 3-T MRI for functional and structural connectivity mapping across an 8-region network of mTLE seizure propagation, including the hippocampus (left and right), insula (left and right), thalamus (left and right), one midline precuneus, and one midline mid-cingulate. Seizure outcome was assessed annually for up to 3 years. Network properties and current outcome prediction methods related to early and long-term seizure outcome were investigated.RESULTSA network model was previously identified across 8 patients with seizure-free mTLE. Results confirmed that whole-network propagation connectivity patterns inconsistent with the mTLE model predict early surgical failure. In those patients with networks consistent with the mTLE network, specific bilateral within-network hippocampal to precuneus impairment (rather than unilateral impairment ipsilateral to the seizure focus) was associated with mild seizure recurrence. No currently used clinical variables offered the same ability to predict long-term outcome.CONCLUSIONSIt is known that there are important clinical differences between early surgical failure that lead to frequent disabling seizures and late recurrence of less frequent mild seizures. This study demonstrated that divergent network connectivity variability, whole-network versus within-network properties, were uniquely associated with these disparate outcomes.
26
Lamberink, Herm J., Willem M. Otte, Ada T. Geerts, Milen Pavlovic, Julio Ramos-Lizana, Anthony G. Marson, Jan Overweg, et al. "Individualised prediction model of seizure recurrence and long-term outcomes after withdrawal of antiepileptic drugs in seizure-free patients: a systematic review and individual participant data meta-analysis." Lancet Neurology 16, no. 7 (July 2017): 523–31. http://dx.doi.org/10.1016/s1474-4422(17)30114-x.
Full text
27
McCormack, J. P., S. D. Eckermann, D. E. Siskind, and T. J. McGee. "CHEM2D-OPP: A new linearized gas-phase ozone photochemistry parameterization for high-altitude NWP and climate models." Atmospheric Chemistry and Physics Discussions 6, no. 4 (July 17, 2006): 6627–94. http://dx.doi.org/10.5194/acpd-6-6627-2006.
Full textAbstract:
Abstract. The new CHEM2D-Ozone Photochemistry Parameterization (CHEM2D-OPP) for high-altitude numerical weather prediction (NWP) systems and climate models specifies the net ozone photochemical tendency and its sensitivity to changes in ozone mixing ratio, temperature and overhead ozone column based on calculations from the CHEM2D interactive middle atmospheric photochemical transport model. We evaluate CHEM2D-OPP performance using both short-term (6-day) and long-term (1-year) stratospheric ozone simulations with the prototype high-altitude NOGAPS-ALPHA forecast model. An inter-comparison of NOGAPS-ALPHA 6-day ozone hindcasts for 7 February 2005 with ozone photochemistry parameterizations currently used in operational NWP systems shows that CHEM2D-OPP yields the best overall agreement with Aura Microwave Limb Sounder ozone profile measurements. A 1-year free-running NOGAPS-ALPHA simulation using CHEM2D-OPP produces a realistic seasonal cycle in zonal mean ozone throughout the stratosphere. We find that the combination of a model cold temperature bias at high latitudes in winter and a warm bias in the CHEM2D-OPP temperature climatology can degrade the performance of the linearized ozone photochemistry parameterization over seasonal time scales despite the fact that the parameterized temperature dependence is weak in these regions.
28
Hu, Zhiming, Yingying Xu, Xiangui Liu, Xianggang Duan, and Jin Chang. "A Semianalytical Production Prediction Model and Dynamics Performance Analysis for Shale Gas Wells." Geofluids 2021 (September 23, 2021): 1–14. http://dx.doi.org/10.1155/2021/9920122.
Full textAbstract:
The shale gas productivity model based on shale gas nonlinear seepage mechanism is an effective way to reasonably predict productivity. The incomplete gas nonlinear effects considered in the current production prediction models can lead to inaccurate production prediction. Based on the conventional five-zone compound flow model, comprehensive gas nonlinearities were considered in the improved compound linear flow model proposed in the paper and a semianalytical solution for productivity was obtained. The reliability of the productivity model was verified by the field data, and then, the 20-year production performance analysis of the gas well was studied. Ultimately, the key influencing factors of the fracture control stage and matrix control stage have been analyzed. Research indicated the following: (1) the EUR predicted by the productivity model is higher than the EUR that the comprehensive nonlinear effects are not considered, which demonstrated that the various nonlinear effects cannot be neglected during the production prediction to ensure the greater calculation accuracy; (2) during the early production stage of shale reservoir, the adsorbed gas is basically not recovered, and the cumulative adsorption contribution rate does not exceed 10%. The final adsorption gas contribution rate is 23.28%, and the annual adsorption rate can exceed 50% in the 20th year, showing that free gas and adsorbed gas are, respectively, important sources of the early stage of production and long-term stable production; (3) the widely ranged three-dimensional fracturing reformation of shale reservoirs and reasonable bottom hole pressure in the later matrix development process should be implemented to increase the effective early production of the reservoir and ensure the earlier gas production process of the matrix development. The findings of this study can help for better ensuring the prediction accuracy of the estimated ultimate recovery and understanding the main influencing factors of the dynamic performance of gas wells so as to provide a theoretical reference for production optimization and development plan formulation of the shale gas reservoirs.
29
Ulmert, David, Angel M. Serio, Matthew F. O'Brien, Charlotte Becker, James A. Eastham, Peter T. Scardino, Thomas Björk, Göran Berglund, Andrew J. Vickers, and Hans Lilja. "Long-Term Prediction of Prostate Cancer: Prostate-Specific Antigen (PSA) Velocity Is Predictive but Does Not Improve the Predictive Accuracy of a Single PSA Measurement 15 Years or More Before Cancer Diagnosis in a Large, Representative, Unscreened Population." Journal of Clinical Oncology 26, no. 6 (February 20, 2008): 835–41. http://dx.doi.org/10.1200/jco.2007.13.1490.
Full textAbstract:
PurposeWe tested whether total prostate-specific antigen velocity (tPSAv) improves accuracy of a model using PSA level to predict long-term risk of prostate cancer diagnosis.MethodsDuring 1974 to 1986 in a preventive medicine study in Sweden, 5,722 men aged ≤ 50 gave two blood samples about 6 years apart. We measured free (fPSA) and total PSA (tPSA) in archived plasma samples from 4,907 participants. Prostate cancer was subsequently diagnosed in 443 (9%) men. Cox proportional hazards regression was used to evaluate tPSA and tPSAv as predictors of prostate cancer. Predictive accuracy was assessed by the concordance index.ResultsThe median time from second blood draw to cancer diagnosis was 16 years; median follow-up for men without prostate cancer was 21 years. In univariate models, tPSA level at second assessment and tPSAv between first and second assessments were associated with prostate cancer (both P < .001). tPSAv was highly correlated with tPSA level (r = 0.93). Twenty-year probabilities of cancer for men at 50th, 90th, and 95th percentile of tPSA and tPSAv were 10.6%, 17.1%, and 21.2% for tPSA, and 9.1%, 11.8%, and 14.1% for tPSAv, respectively. The concordance index for tPSA level was 0.771. Adding tPSAv, fPSA, %fPSA or velocities of fPSA and %fPSA did not importantly increase accuracy of tPSA to predict prostate cancer. Results were unchanged if the analysis was restricted to patients with advanced cancer at diagnosis.ConclusionAlthough PSA velocity is significantly increased in men with prostate cancer up to two decades before diagnosis, it does not aid long-term prediction of prostate cancer.
30
McCormack, J. P., S. D. Eckermann, D. E. Siskind, and T. J. McGee. "CHEM2D-OPP: A new linearized gas-phase ozone photochemistry parameterization for high-altitude NWP and climate models." Atmospheric Chemistry and Physics 6, no. 12 (October 30, 2006): 4943–72. http://dx.doi.org/10.5194/acp-6-4943-2006.
Full textAbstract:
Abstract. The new CHEM2D-Ozone Photochemistry Parameterization (CHEM2D-OPP) for high-altitude numerical weather prediction (NWP) systems and climate models specifies the net ozone photochemical tendency and its sensitivity to changes in ozone mixing ratio, temperature and overhead ozone column based on calculations from the CHEM2D interactive middle atmospheric photochemical transport model. We evaluate CHEM2D-OPP performance using both short-term (6-day) and long-term (1-year) stratospheric ozone simulations with the prototype high-altitude NOGAPS-ALPHA forecast model. An inter-comparison of NOGAPS-ALPHA 6-day ozone hindcasts for 7 February 2005 with ozone photochemistry parameterizations currently used in operational NWP systems shows that CHEM2D-OPP yields the best overall agreement with both individual Aura Microwave Limb Sounder ozone profile measurements and independent hemispheric (10°–90° N) ozone analysis fields. A 1-year free-running NOGAPS-ALPHA simulation using CHEM2D-OPP produces a realistic seasonal cycle in zonal mean ozone throughout the stratosphere. We find that the combination of a model cold temperature bias at high latitudes in winter and a warm bias in the CHEM2D-OPP temperature climatology can degrade the performance of the linearized ozone photochemistry parameterization over seasonal time scales despite the fact that the parameterized temperature dependence is weak in these regions.
31
Malumbres, Raquel, Nathalie A. Johnson, Laurie H. Sehn, Yaso Natkunam, Rob Tibshirani, Javier Briones, Joseph M. Connors, Ronald Levy, Randy D. Gascoyne, and Izidore S. Lossos. "Paraffin-Based 6-Gene Model Predicts Outcome of Diffuse Large B-Cell Lymphoma Patients Treated with R-CHOP." Blood 110, no. 11 (November 16, 2007): 49. http://dx.doi.org/10.1182/blood.v110.11.49.49.
Full textAbstract:
Abstract Background: Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous disease characterized by highly variable clinical outcomes. It is therefore of paramount importance to be able to predict the outcome of patients at the time of diagnosis. Previously, we constructed a 6 gene model for outcome prediction of DLBCL patients treated with anthracycline-based chemotherapies (Lossos et al NEJM 2004, 350:1829). However, the standard therapy has evolved to rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone (R-CHOP). Subset analyses of clinical trials have suggested that some of the prognostic factors lose their predictive power in R-CHOP treated patients. Consequently, the molecular gold standard for survival prediction in R-CHOP treated patients has not been established. Herein, we evaluated the predictive power of the 6-gene model in R-CHOP treated DLBCL patients. We have employed new methodology that allows quantitation of gene expression from paraffin embedded, fixed tissues. Methods: RNA was extracted from 100 paraffin-embedded specimens (Chen et al Diagn Mol Pathol 2007, 16:61), from patients with DLBCL treated with R-CHOP in British Columbia (73) and at the University of Miami (27). Expression of the 6 genes comprising the model was measured in these samples and the mortality-prediction score was calculated for each patient, as reported previously. Results: The study group consisted of 100 patients with a median age of 58 y (range, 16–92y) that were followed for a median of 2.1 years (range, 0.1–5.6). Distribution according to IPI: 0–1 factor, 42; 2 factors, 24; 3 factors, 19; and ≥4 factors, 14. RNA of sufficient quality and quantity was successfully extracted from all the 100 paraffin embedded specimens tested, some of which had been stored for up to 6 years. The mortality-prediction score derived from the model divided patients into low-risk (50%) and high-risk (50%) subgroups with significantly different overall survival (OS) (p=0.02) and progression free survival (PFS) (p=0.02). Notably the OS was similar between the groups during the 1st year, due to the inclusion of patients with advanced age and poor performance status, but it was markedly different beyond the first 2 years, with the 3 year OS of 80% and 50% in the low risk and high risk groups, respectively. The predictive power of the 6-gene model was independent of the IPI prognostic factors for prediction of OS (P=0.06) and PFS (p=0.01) in these patients. Conclusions: The prognostic value of the 6-gene model remains significant in the era of R-CHOP treatment. Further, using the new RNA extraction methodology, we demonstrate that the model can be applied to routinely available formalin-fixed paraffin blocks from initial diagnostic biopsies, even after long term storage. Following validation in an independent cohort of patients, the six gene model may be practically applied in routine clinical practice.
32
Weyhe, Dirk, Dennis Obonyo, Verena Nicole Uslar, Ingo Stricker, and Andrea Tannapfel. "Predictive factors for long-term survival after surgery for pancreatic ductal adenocarcinoma: Making a case for standardized reporting of the resection margin using certified cancer center data." PLOS ONE 16, no. 3 (March 18, 2021): e0248633. http://dx.doi.org/10.1371/journal.pone.0248633.
Full textAbstract:
Factors for overall survival after pancreatic ductal adenocarcinoma (PDAC) seem to be nodal status, chemotherapy administration, UICC staging, and resection margin. However, there is no consensus on the definition for tumor free resection margin. Therefore, univariate OS as well as multivariate long-term survival using cancer center data was analyzed with regards to two different resection margin definitions. Ninety-five patients met inclusion criteria (pancreatic head PDAC, R0/R1, no 30 days mortality). OS was analyzed in univariate analysis with respect to R-status, CRM (circumferential resection margin; positive: ≤1mm; negative: >1mm), nodal status, and chemotherapy administration. Long-term survival >36 months was modelled using multivariate logistic regression instead of Cox regression because the distribution function of the dependent data violated the requirements for the application of this test. Significant differences in OS were found regarding the R status (Median OS and 95%CI for R0: 29.8 months, 22.3–37.4; R1: 15.9 months, 9.2–22.7; p = 0.005), nodal status (pN0 = 34.7, 10.4–59.0; pN1 = 17.1, 11.5–22.8; p = 0.003), and chemotherapy (with CTx: 26.7, 20.4–33.0; without CTx: 9.7, 5.2–14.1; p < .001). OS according to CRM status differed on a clinically relevant level by about 12 months (CRM positive: 17.2 months, 11.5–23.0; CRM negative: 29.8 months, 18.6–41.1; p = 0.126). A multivariate model containing chemotherapy, nodal status, and CRM explained long-term survival (p = 0.008; correct prediction >70%). Chemotherapy, nodal status and resection margin according to UICC R status are univariate factors for OS after PDAC. In contrast, long-term survival seems to depend on wider resection margins than those used in UICC R classification. Therefore, standardized histopathological reporting (including resection margin size) should be agreed upon.
33
Maak, M., E. Zeestraten, M. Shibayama, T. Schuster, H. Friess, C. J. Van De Velde, H. Ishihara, R. Rosenberg, P. J. Kuppen, and K. Janssen. "Specific activity of cyclin dependent kinase 1 as a novel predictor of recurrence risk in stage II colon cancer." Journal of Clinical Oncology 29, no. 4_suppl (February 1, 2011): 402. http://dx.doi.org/10.1200/jco.2011.29.4_suppl.402.
Full textAbstract:
402 Background: Altered cell cycle dynamics and check points are typical features of solid tumors, and cyclin dependent kinases (CDKs) play pivotal roles in these processes. Previously we have demonstrated that CDK-based analysis, composed of CDK1 and CDK2, is useful in the prediction of outcomes in early breast cancer patients (Ann Oncol. 19(1):68-72, 2008, Br J Cancer. 100(3):494-500, 2009). Clinically, there is a need for risk stratification in patients with stage II colon cancer who have a recurrence risk of 20 to 30%. Therefore we investigated the use of CDK-based analysis for recurrence prediction of stage II colon cancer patients. Methods: Fresh frozen tissue samples of 254 patients with histologically confirmed adenocarcinoma of the colon, UICC stage II, who received primary tumor resection in Munich (217 cases), and Leiden (37 cases) were used. Protein expression and activity of CDK1 and CDK2 were determined by in vitro assays as previously described. Specific activity (SA) of CDKs was calculated as kinase activity in relation to its corresponding mass concentration. Results: Development of distant metastasis was observed in 27 patients (10.6%) after a median follow up of 86 months. We found that predictive performance of CDK1SA, but not CDK2SA, for the metastasis was substantial and almost constant for long-term event prediction (average area under the curve (AUC) = 0.69). Tumor recurrence risk analysis in association with CDK1SA identified a low- (41% of population) and high- risk group (59%). Cox proportional hazard model analysis retained the CDK-based patient classification as an independent prognostic factor for distant metastases-free survival (low vs. high-risk group: Hazard ratio = 6.2, 95% CI: 1.45 to 26.9, p=0.0049). Clinical parameters such as grading, T-categories, age, and sex were excluded as confounding factors for CDK1SA-risk. Conclusions: CDK1SA allows stratification of different risk subgroups of stage II colon cancer patients. CDK1SA-based analysis is useful for predicting patients with high risk of distant recurrence, who should be treated with chemotherapy. No significant financial relationships to disclose.
34
Lionis, Antonios, Konstantinos Peppas, Hector E. Nistazakis, Andreas Tsigopoulos, Keith Cohn, and Athanassios Zagouras. "Using Machine Learning Algorithms for Accurate Received Optical Power Prediction of an FSO Link over a Maritime Environment." Photonics 8, no. 6 (June 10, 2021): 212. http://dx.doi.org/10.3390/photonics8060212.
Full textAbstract:
The performance prediction of an optical communications link over maritime environments has been extensively researched over the last two decades. The various atmospheric phenomena and turbulence effects have been thoroughly explored, and long-term measurements have allowed for the construction of simple empirical models. The aim of this work is to demonstrate the prediction accuracy of various machine learning (ML) algorithms for a free-space optical communication (FSO) link performance, with respect to real time, non-linear atmospheric conditions. A large data set of received signal strength indicators (RSSI) for a laser communications link has been collected and analyzed against seven local atmospheric parameters (i.e., wind speed, pressure, temperature, humidity, dew point, solar flux and air-sea temperature difference). The k-nearest-neighbors (KNN), tree-based methods-decision trees, random forest and gradient boosting- and artificial neural networks (ANN) have been employed and compared among each other using the root mean square error (RMSE) and the coefficient of determination (R2) of each model as the primary performance indices. The regression analysis revealed an excellent fit for all ML models, indicative of their ability to offer a significant improvement in FSO performance modeling as compared to traditional regression models. The best-performing R2 model found to be the ANN approach (0.94867), while random forests achieved the most optimal RMSE result (7.37).
35
Shouval, Roni, Annalisa Ruggeri, Myriam Labopin, Mohamad Mohty, Guillermo Sanz, Gerard Michel, Eefke Petersen, et al. "A Machine Learning Based Model to Predict Two-Year Leukemia Free Survival in Cord Blood Transplantation for Acute Leukemia - a Data Mining Study, on Behalf of Eurocord, Cord Blood Committee and the Acute Leukemia Working Party of the EBMT." Blood 126, no. 23 (December 3, 2015): 3211. http://dx.doi.org/10.1182/blood.v126.23.3211.3211.
Full textAbstract:
Abstract Background: Umbilical cord blood transplantation (UCBT) is a potentially curative therapy acute leukemia (AL) patients. Transplantation benefit must be balanced against risks, such as transplant related mortality and relapse. The complex nature of hematopoietic stem cell transplantation data (HCT), rich in interactions and possibly nonlinear associations, has motivated us to apply machine learning (ML) for predictive modeling. ML is a field of artificial intelligence and is part of the data mining approach for data analysis. Our group has recently reported on a ML based prediction model for short term HCT outcomes (Shouval R et al; JCO 2015). Using a ML algorithm, the perspective of the current study was prediction of leukemia free survival (LFS) at 2 years after an UCBT, while exploring variables' importance and interactions. Patients & Methods: A cohort of 3,149 UCBT were analyzed. Inclusion criteria encompassed patients at all ages, undergoing an UCBT (single/double unit) in EBMT centers from the year 2004 to 2014, for AL, in all disease status. All conditioning and graft versus host disease prophylaxis regiments were included. A total of 24 variables were considered, including the number of total nucleated cell dose (TNC), donor and recipients HLA typing, as well as recipient, disease and transplant characteristics. The Random Survival Forest (RSF) ML algorithm was applied for model construction and data exploration. RSF is known to be adaptive to data, is able to automatically recover nonlinear effects and complex interactions among variables, and yields nonparametric prediction over test data. The analysis pipeline consisted of prediction model development, assessment of variable importance by their minimal depth from the tree trunk, and exploration of the top ranking variable with dependence plots. The latter promotes understanding of non-trivial associations between variables and outcomes. Results : The 2 years LFS was 49%, with a median follow up of 30 months. A RSF model of 1000 trees was developed, with each tree constructed on a bootstrap sample from the original cohort. A prediction error of 36.0% was calculated. The 10 most predictive variables (in ascending order) were disease status, age, TNC harvested and infused, recipient CMV serostatus, interval from diagnosis to UCBT, transplant year, previous autologous transplant, and use of anti-thymocyte globulin (ATG). Selected findings from exploration of variables-outcome relationship with dependence plots included a varying effect of TNCs in specific subpopulations. Increasing the number of infused TNCs had a positive effect on predicted LFS in patients receiving HLA mismatched (2 or more HLA mismatch) (figure) or single unit CB grafts, and patients in earlier disease status or older age. ATG administration was associated with worse LFS, whether unadjusted or adjusted to all other variables. However, there was an additional negative effect in advanced disease status patients, recipients of HLA mismatched or single CB units grafts, and older patients. Patients in 1st complete remission (CR) had higher predicted LFS as compared to those in 2nd CR. However, in patients receiving a HLA mismatched or a double CB graft, the difference in LFS between CR1 and CR2 was attenuated. Younger age had a favorable impact in early disease status, but lost its positive effect in advanced disease. Conclusions: A prediction model for LFS 2 years post UBCT was developed using the RSF ML algorithm. Variables were ranked according to their predictive contribution. Disease status, age, and TNC count were found to be the most important factors. Dependence plots revealed interactions and nonlinear associations between variables and the outcome, such as the effect of cell dose on HLA disparity. Apart from the study's clinical findings, it carries a methodological significance. A novel ML approach for prediction, variable selection and data exploration, accounting for long term time to event outcomes, has proved useful in the field of HCT. Figure 1. Variable marginal dependence coplot of predicted LFS at 2 years against TNC, conditional on HLA matching. Individual cases are marked with blue circles (alive or censored) and red `x's (event). Linear smooth (a linear extrapolation of the prediction function), with shaded 95% confidence band, indicates trends of variable dependence. Figure 1. Variable marginal dependence coplot of predicted LFS at 2 years against TNC, conditional on HLA matching. Individual cases are marked with blue circles (alive or censored) and red `x's (event). Linear smooth (a linear extrapolation of the prediction function), with shaded 95% confidence band, indicates trends of variable dependence. Disclosures Mohty: Janssen: Honoraria; Celgene: Honoraria. Sanz:JANSSEN CILAG: Honoraria, Research Funding, Speakers Bureau. Bader:Neovii: Other: Institutional grants; Medac: Other: Institutional grants; Riemser: Other: Institutional grants; Amgen: Consultancy; Novartis: Consultancy; Jazz Pharmaceuticals: Consultancy.
36
Kleiman, M., S. Shacham, M. Kushnir, E. Chapnik, and Z. Agur. "Use of the Optimata Virtual Patient (OVP) to predict effects of sunitinib malate in advanced pancreatic cancer." Journal of Clinical Oncology 29, no. 4_suppl (February 1, 2011): 341. http://dx.doi.org/10.1200/jco.2011.29.4_suppl.341.
Full textAbstract:
341 Background: Pancreatic cancer is considered to be incurable by available treatment modalities, with 5-year survival rate <4%. Sunitinib malate (Sutent) significantly increased progression-free survival in patients with advanced islet cell tumors (pancreatic); however, it demonstrated only modest single-agent effect in a phase II study in patients with metastatic pancreatic adenocarcinoma (PaC) that progressed after first-line therapy with gemcitabine. To improve the response of PaC patients to drugs, the interplay between biologic, pathologic, and pharmacologic processes underlying drug-patient interactions have been mathematically modeled, predicting efficacy responses, different toxicities, and long-term tolerability in clinical trials, allowing for improved dosing regimens and patient selection (OVP engine) (Agur Z., 2010). The OVP engine was used to predict Sunitinib malate single-agent activity in advanced PaC. Methods: OVP replicated the observed growth patterns of human PaC. Pharmacokinetics (PK) and pharmacodynamics (PD) of sunitinib malate were modeled based on literature (NDA 21-938). Effects of sunitinib malate on human PaC xenografts were scaled to model PK/PD in human. To evaluate drug efficacy in advanced PaC patient-population, a Virtual PaC Patient- Population was created by replacing the population averaged parameter values in the model by their distribution in the population. Using this procedure, a large set of virtual patients was generated. Model simulations with sunitinib malate therapy (50 mg QD/28 days; 14 days rest [1 cycle]) enabled the prediction and classification of patients' response according to RECIST criteria and compared with the actual clinical response (O'Reilly et al, 2008). Results: Simulations of the FDA-approved schedule, predicted a stable disease in 81% of the patients following one treatment cycle and 54% following two treatment cycles. Stable disease was predicted to be the best observed response, which was confirmed in the clinical study. Conclusions: Our model predictions are compatible with clinical results of a recent phase II trial with the same treatment regimen of sunitinib malate (O'Reilly et al, 2008) and further suggest the use of mathematical model during drug development. [Table: see text]
37
Vassallo, Diana, Robert N. Foley, and Philip A. Kalra. "Design of a clinical risk calculator for major clinical outcomes in patients with atherosclerotic renovascular disease." Nephrology Dialysis Transplantation 34, no. 8 (June 22, 2018): 1377–84. http://dx.doi.org/10.1093/ndt/gfy157.
Full textAbstract:
Abstract Background Risk stratification in atherosclerotic renovascular disease (ARVD) can influence treatment decisions and facilitate patient selection for revascularization. In this study, we aim to use variables with the best predictive value to design a risk calculator that can assist clinicians with risk stratification and outcome prediction. Methods Patients with a radiological diagnosis of ARVD referred to our tertiary renal centre were recruited into this prospective cohort study between 1986 and 2014. Primary clinical endpoints included: death, progression to end-stage kidney disease and cardiovascular events (CVE). A stepwise regression model was used to select variables with the most significant hazard ratio for each clinical endpoint. The risk calculator was designed using Hypertext Markup Language. Survival and CVE-free survival were estimated at 1, 5 and 10 years. Results In total, 872 patients were recruited into the Salford ARVD study with a median follow-up period of 54.9 months (interquartile range 20.2–96.0). Only models predicting death and CVE showed good performance (C-index >0.80). Survival probabilities obtained from the risk calculator show that most patients with ARVD have reduced long-term survival. Revascularization improved outcomes in patients with higher baseline estimated glomerular filtration rate and lower proteinuria but not in those with co-existing comorbidities and higher levels of baseline proteinuria. Conclusions Although this risk calculator requires further independent validation in other ARVD cohorts, this study shows that a small number of easily obtained variables can help predict clinical outcomes and encourage a patient-specific therapeutic approach.
38
Lin, Jingping, Jinsen Weng, Shaofeng Lin, Cuibo Lin, Jieping Huang, Chunxia Zhang, Shen Zhang, Chuanpeng Dong, Haizhou Ji, and Xi Ke. "Development and validation of a novel mRNA signature for predicting early relapse in non-small cell lung cancer." Japanese Journal of Clinical Oncology 51, no. 8 (May 25, 2021): 1277–86. http://dx.doi.org/10.1093/jjco/hyab075.
Full textAbstract:
Abstract Background Recurrence after initial primary resection is still a major and ultimate cause of death for non-small cell lung cancer patients. We attempted to build an early recurrence associated gene signature to improve prognostic prediction of non-small cell lung cancer. Methods Propensity score matching was conducted between patients in early relapse group and long-term survival group from The Cancer Genome Atlas training series (N = 579) and patients were matched 1:1. Global transcriptome analysis was then performed between the paired groups to identify tumour-specific mRNAs. Finally, using LASSO Cox regression model, we built a multi-gene early relapse classifier incorporating 40 mRNAs. The prognostic and predictive accuracy of the signature was internally validated in The Cancer Genome Atlas patients. Results A total of 40 mRNAs were finally identified to build an early relapse classifier. With specific risk score formula, patients were classified into a high-risk group and a low-risk group. Relapse-free survival was significantly different between the two groups in both discovery (HR: 3.244, 95% CI: 2.338-4.500, P < 0.001) and internal validation series (HR 1.970, 95% CI 1.181-3.289, P = 0.009). Further analysis revealed that the prognostic value of this signature was independent of tumour stage, histotype and epidermal growth factor receptor mutation (P < 0.05). Time-dependent receiver operating characteristic analysis showed that the area under receiver operating characteristic curve of this signature was higher than TNM stage alone (0.771 vs 0.686, P < 0.05). Further, decision curve analysis curves analysis at 1 year revealed the considerable clinical utility of this signature in predicting early relapse. Conclusions We successfully established a reliable signature for predicting early relapse in stage I–III non-small cell lung cancer.
39
Wu, Xiang-Fa, and Oksana Zholobko. "Experimental Study of the Probabilistic Fatigue Residual Strength of a Carbon Fiber-Reinforced Polymer Matrix Composite." Journal of Composites Science 4, no. 4 (November 21, 2020): 173. http://dx.doi.org/10.3390/jcs4040173.
Full textAbstract:
Degradation of the mechanical properties of fiber-reinforced polymer matrix composites (PMCs) subjected to cyclic loading is crucial to the long-term load-carrying capability of PMC structures in practice. This paper reports the experimental study of fatigue residual tensile strength and its probabilistic distribution in a carbon fiber-reinforced PMC laminate made of unidirectional (UD) carbon-fiber/epoxy prepregs (Hexcel T2G190/F263) with the ply layup [0/±45/90]S after certain cycles of cyclic loading. The residual tensile strengths of the PMC laminates after cyclic loading of 1 (quasistatic), 2000, and 10,000 cycles were determined. Statistical analysis of the experimental data shows that the fatigue residual tensile strength of the PMC laminate follows a two-parameter Weibull distribution model with the credibility ≥ 95%. With increasing fatigue cycles, the mean value of the fatigue residual strength of the PMC specimens decreased while its deviation increased. A free-edge stress model is further adopted to explain the fatigue failure initiation of the composite laminate. The present experimental study is valuable for understanding the fatigue durability of PMC laminates as well as reliable design and performance prediction of composite structures.
40
Heo, Tak Sung, Yu Seop Kim, Jeong Myeong Choi, Yeong Seok Jeong, Soo Young Seo, Jun Ho Lee, Jin Pyeong Jeon, and Chulho Kim. "Prediction of Stroke Outcome Using Natural Language Processing-Based Machine Learning of Radiology Report of Brain MRI." Journal of Personalized Medicine 10, no. 4 (December 16, 2020): 286. http://dx.doi.org/10.3390/jpm10040286.
Full textAbstract:
Brain magnetic resonance imaging (MRI) is useful for predicting the outcome of patients with acute ischemic stroke (AIS). Although deep learning (DL) using brain MRI with certain image biomarkers has shown satisfactory results in predicting poor outcomes, no study has assessed the usefulness of natural language processing (NLP)-based machine learning (ML) algorithms using brain MRI free-text reports of AIS patients. Therefore, we aimed to assess whether NLP-based ML algorithms using brain MRI text reports could predict poor outcomes in AIS patients. This study included only English text reports of brain MRIs examined during admission of AIS patients. Poor outcome was defined as a modified Rankin Scale score of 3–6, and the data were captured by trained nurses and physicians. We only included MRI text report of the first MRI scan during the admission. The text dataset was randomly divided into a training and test dataset with a 7:3 ratio. Text was vectorized to word, sentence, and document levels. In the word level approach, which did not consider the sequence of words, and the “bag-of-words” model was used to reflect the number of repetitions of text token. The “sent2vec” method was used in the sensation-level approach considering the sequence of words, and the word embedding was used in the document level approach. In addition to conventional ML algorithms, DL algorithms such as the convolutional neural network (CNN), long short-term memory, and multilayer perceptron were used to predict poor outcomes using 5-fold cross-validation and grid search techniques. The performance of each ML classifier was compared with the area under the receiver operating characteristic (AUROC) curve. Among 1840 subjects with AIS, 645 patients (35.1%) had a poor outcome 3 months after the stroke onset. Random forest was the best classifier (0.782 of AUROC) using a word-level approach. Overall, the document-level approach exhibited better performance than did the word- or sentence-level approaches. Among all the ML classifiers, the multi-CNN algorithm demonstrated the best classification performance (0.805), followed by the CNN (0.799) algorithm. When predicting future clinical outcomes using NLP-based ML of radiology free-text reports of brain MRI, DL algorithms showed superior performance over the other ML algorithms. In particular, the prediction of poor outcomes in document-level NLP DL was improved more by multi-CNN and CNN than by recurrent neural network-based algorithms. NLP-based DL algorithms can be used as an important digital marker for unstructured electronic health record data DL prediction.
41
Clancy, Kadie, Esmaeel Dadashzadeh, Christof Kaltenmeier, JB Moses, and Shandong Wu. "3132 Machine Learning for Prediction of Pathologic Pneumatosis Intestinalis Using CT Scans." Journal of Clinical and Translational Science 3, s1 (March 2019): 60–61. http://dx.doi.org/10.1017/cts.2019.142.
Full textAbstract:
OBJECTIVES/SPECIFIC AIMS: This retrospective study aims to create and train machine learning models using a radiomic-based feature extraction method for two classification tasks: benign vs. pathologic PI and operation of benefit vs. operation not needed. The long-term goal of our study is to build a computerized model that incorporates both radiomic features and critical non-imaging clinical factors to improve current surgical decision-making when managing PI patients. METHODS/STUDY POPULATION: Searched radiology reports from 2010-2012 via the UPMC MARS Database for reports containing the term “pneumatosis” (subsequently accounting for negations and age restrictions). Our inclusion criteria included: patient age 18 or older, clinical data available at time of CT diagnosis, and PI visualized on manual review of imaging. Cases with intra-abdominal free air were excluded. Collected CT imaging data and an additional 149 clinical data elements per patient for a total of 75 PI cases. Data collection of an additional 225 patients is ongoing. We trained models for two clinically-relevant prediction tasks. The first (referred to as prediction task 1) classifies between benign and pathologic PI. Benign PI is defined as either lack of intraoperative visualization of transmural intestinal necrosis or successful non-operative management until discharge. Pathologic PI is defined as either intraoperative visualization of transmural PI or withdrawal of care and subsequent death during hospitalization. The distribution of data samples for prediction task 1 is 47 benign cases and 38 pathologic cases. The second (referred to as prediction task 2) classifies between whether the patient benefitted from an operation or not. “Operation of benefit” is defined as patients with PI, be it transmural or simply mucosal, who benefited from an operation. “Operation not needed” is defined as patients who were safely discharged without an operation or patients who had an operation, but nothing was found. The distribution of data samples for prediction task 2 is 37 operation not needed cases and 38 operation of benefit cases. An experienced surgical resident from UPMC manually segmented 3D PI ROIs from the CT scans (5 mm Axial cut) for each case. The most concerning ~10-15 cm segment of bowel for necrosis with a 1 cm margin was selected. A total of 7 slices per patient were segmented for consistency. For both prediction task 1 and prediction task 2, we independently completed the following procedure for testing and training: 1.) Extracted radiomic features from the 3D PI ROIs that resulted in 99 total features. 2.) Used LASSO feature selection to determine the subset of the original 99 features that are most significant for performance of the prediction task. 3.) Used leave-one-out cross-validation for testing and training to account for the small dataset size in our preliminary analysis. Implemented and trained several machine learning models (AdaBoost, SVM, and Naive Bayes). 4.) Evaluated the trained models in terms of AUC and Accuracy and determined the ideal model structure based on these performance metrics. RESULTS/ANTICIPATED RESULTS: Prediction Task 1: The top-performing model for this task was an SVM model trained using 19 features. This model had an AUC of 0.79 and an accuracy of 75%. Prediction Task 2: The top-performing model for this task was an SVM model trained using 28 features. This model had an AUC of 0.74 and an accuracy of 64%. DISCUSSION/SIGNIFICANCE OF IMPACT: To the best of our knowledge, this is the first study to use radiomic-based machine learning models for the prediction of tissue ischemia, specifically intestinal ischemia in the setting of PI. In this preliminary study, which serves as a proof of concept, the performance of our models has demonstrated the potential of machine learning based only on radiomic imaging features to have discriminative power for surgical decision-making problems. While many non-imaging-related clinical factors play a role in the gestalt of clinical decision making when PI presents, we have presented radiomic-based models that may augment this decision-making process, especially for more difficult cases when clinical features indicating acute abdomen are absent. It should be noted that prediction task 2, whether or not a patient presenting with PI would benefit from an operation, has lower performance than prediction task 1 and is also a more challenging task for physicians in real clinical environments. While our results are promising and demonstrate potential, we are currently working to increase our dataset to 300 patients to further train and assess our models. References DuBose, Joseph J., et al. “Pneumatosis Intestinalis Predictive Evaluation Study (PIPES): a multicenter epidemiologic study of the Eastern Association for the Surgery of Trauma.” Journal of Trauma and Acute Care Surgery 75.1 (2013): 15-23. Knechtle, Stuart J., Andrew M. Davidoff, and Reed P. Rice. “Pneumatosis intestinalis. Surgical management and clinical outcome.” Annals of Surgery 212.2 (1990): 160.
42
Heimann, F. U. M., D. Rickenmann, J. M. Turowski, and J. W. Kirchner. "sedFlow – a tool for simulating fractional bedload transport and longitudinal profile evolution in mountain streams." Earth Surface Dynamics 3, no. 1 (January 12, 2015): 15–34. http://dx.doi.org/10.5194/esurf-3-15-2015.
Full textAbstract:
Abstract. Especially in mountainous environments, the prediction of sediment dynamics is important for managing natural hazards, assessing in-stream habitats and understanding geomorphic evolution. We present the new modelling tool {sedFlow} for simulating fractional bedload transport dynamics in mountain streams. sedFlow is a one-dimensional model that aims to realistically reproduce the total transport volumes and overall morphodynamic changes resulting from sediment transport events such as major floods. The model is intended for temporal scales from the individual event (several hours to few days) up to longer-term evolution of stream channels (several years). The envisaged spatial scale covers complete catchments at a spatial discretisation of several tens of metres to a few hundreds of metres. sedFlow can deal with the effects of streambeds that slope uphill in a downstream direction and uses recently proposed and tested approaches for quantifying macro-roughness effects in steep channels. sedFlow offers different options for bedload transport equations, flow-resistance relationships and other elements which can be selected to fit the current application in a particular catchment. Local grain-size distributions are dynamically adjusted according to the transport dynamics of each grain-size fraction. sedFlow features fast calculations and straightforward pre- and postprocessing of simulation data. The high simulation speed allows for simulations of several years, which can be used, e.g., to assess the long-term impact of river engineering works or climate change effects. In combination with the straightforward pre- and postprocessing, the fast calculations facilitate efficient workflows for the simulation of individual flood events, because the modeller gets the immediate results as direct feedback to the selected parameter inputs. The model is provided together with its complete source code free of charge under the terms of the GNU General Public License (GPL) (www.wsl.ch/sedFlow). Examples of the application of sedFlow are given in a companion article by Heimann et al. (2015).
43
Li, Xiyang, Bin Li, Tarlok Singh, and Kan Shi. "Predicting stock market returns in the US: evidence from an average correlation approach." Accounting Research Journal 33, no. 2 (February 10, 2020): 411–33. http://dx.doi.org/10.1108/arj-10-2018-0168.
Full textAbstract:
Purpose This study aims to draw on a less explored predictor – the average correlation of pairwise returns on industry portfolios – to predict stock market returns (SMRs) in the USA. Design/methodology/approach This study uses the average correlation approach of Pollet and Wilson (2010) and predicts the SMRs in the USA. The model is estimated using monthly data for a long time horizon, from July 1963 to December 2018, for the portfolios comprising 48 Fama-French industries. The model is extended to examine the effects of a longer lag structure of one-month to four-month lags and to control for the effects of a number of variables – average variance (AV), cyclically adjusted price-to-earnings ratio (CAPE), term spread (TS), default spread (DS), risk-free rate returns (R_f) and lagged excess market returns (R_s). Findings The study finds that the two-month lagged average correlation of returns on individual industry portfolios, used individually and collectively with financial predictors and economic factors, predicts excess returns on the stock market in an effective manner. Research limitations/implications The methodology and results are of interest to academics as they could further explore the use of average correlation to improve the predictive powers of their models. Practical implications Market practitioners could include the average correlation in their asset pricing models to improve the predictions for the future trend in stock market returns. Investors could consider including average correlation in their forecasting models, along with the traditional financial ratios and economic indicators. They could adjust their expected returns to a lower level when the average correlation increases during a recession. Social implications The finding that recession periods have effects on the SMRs would be useful for the policymakers. The understanding of the co-movement of returns on industry portfolios during a recession would be useful for the formulation of policies aimed at ensuring the stability of the financial markets. Originality/value The study contributes to the literature on three counts. First, the study uses industry portfolio returns – as compared to individual stock returns used in Pollet and Wilson (2010) – in constructing average correlation. When stock market becomes more volatile on returns, the individual stocks are more diverse on their performance; the comovement between individual stock returns might be dominated by the idiosyncratic component, which may not have any implications for future SMRs. Using the industry portfolio returns can potentially reduce such an effect by a large extent, and thus, can provide more reliable estimates. Second, the effects of business cycles could be better identified in a long sample period and through several sub-sample tests. This study uses a data set, which spans the period from July 1963 to December 2018. This long sample period covers multiple phases of business cycles. The daily data are used to compute the monthly and equally-weighted average correlation of returns on 48 Fama-French industry portfolios. Third, previous studies have often ignored the use of investors’ sentiments in their prediction models, while investors’ irrational decisions could have an important impact on expected returns (Huang et al., 2015). This study extends the analysis and incorporates investors’ sentiments in the model.
44
Muchtar, Eli, Terry Therneau, Dirk Larson, Morie A. Gertz, Martha Q. Lacy, Francis K. Buadi, David Dingli, et al. "Comparative Analysis of Staging Systems in AL Amyloidosis." Blood 132, Supplement 1 (November 29, 2018): 3228. http://dx.doi.org/10.1182/blood-2018-99-114262.
Full textAbstract:
Abstract Introduction: Immunoglobulin light chain (AL) amyloidosis is associated with significant morbidity and mortality. Several models were developed to allow survival prediction, but head-to-head model comparison is lacking. Patients and Methods: Five prognostic models were compared: Mayo 2004; Mayo 2012; 2013 European modification of Mayo 2004 (where Mayo 2004 stage 3 is subdivided into 3 subgroups based on NT-proBNP at 8500 ng/L and systolic blood pressure at 100 mmHg); 2015 European modification of Mayo 2004 (where Mayo 2004 stage 3 is subdivided into 2 subgroups based on NT-proBNP at 8500 ng/L); and organ model (a model based on number of involved organs). All patients with available baseline data for the 5 models were included in this study (n=1005). Improvements in predictive accuracy of the various models were determined by calculating a net improvement in survival prediction for each pair of prognostic models. Net improvement was considered as the proportion where model A outperforms model B minus the proportion where model B outperforms model A. Results: For all models, there was increasing risk for death with increasing stage. The organ model was the least powerful system with a hazard ratio of 2.8 for ≥4 organs compared to 1 involved organ. The organ model was consistently outperformed by the other 4 systems. In the cohort as a whole, the biomarker models (Mayo 2004; Mayo 2012; European 2013; European 2015) were comparable in ability to predict survival (Figure). However, a slightly superior performance of the European 2013 modification of the Mayo 2004 system was seen over the Mayo 2004 system in the group as a whole (net improvement 4.1%, P=0.023) and when restricted to non-ASCT population (net improvement 6.3% P=0.014). The European 2015 modification performed slightly better than the Mayo 2004 system in the ASCT population (net improvement 3.4%, P=0.025). 1-year mortality was better predicted by the European models (net improvement 11-14%, P<0.05), while the Mayo 2012 had better survival prediction to all other biomarker models in 3-year landmark comparison (net improvement 7-9%, P<0.05). Conclusion: Using biomarker based staging is a clear improvement over the organ staging system. The different biomarker staging systems' abilities to predict for survival were relatively comparable, solidifying the importance of soluble cardiac biomarkers and free light chain measurement as prognosticators in AL amyloidosis. Mayo 2004-based models better predict for early death while Mayo 2012 model has a better prediction for long-term survival. Figure. Figure. Disclosures Gertz: Ionis: Honoraria; celgene: Consultancy; spectrum: Consultancy, Honoraria; janssen: Consultancy; Amgen: Consultancy; Abbvie: Consultancy; Medscape: Consultancy; annexon: Consultancy; Teva: Consultancy; Research to Practice: Consultancy; Apellis: Consultancy; Alnylam: Honoraria; Prothena: Honoraria; Physicians Education Resource: Consultancy. Lacy:Celgene: Research Funding. Dingli:Millennium Takeda: Research Funding; Millennium Takeda: Research Funding; Alexion Pharmaceuticals, Inc.: Other: Participates in the International PNH Registry (for Mayo Clinic, Rochester) for Alexion Pharmaceuticals, Inc.; Alexion Pharmaceuticals, Inc.: Other: Participates in the International PNH Registry (for Mayo Clinic, Rochester) for Alexion Pharmaceuticals, Inc.. Kapoor:Celgene: Research Funding; Takeda: Research Funding. Russell:Vyriad: Equity Ownership. Kumar:Celgene: Membership on an entity's Board of Directors or advisory committees, Research Funding; KITE: Membership on an entity's Board of Directors or advisory committees, Research Funding; AbbVie: Membership on an entity's Board of Directors or advisory committees, Research Funding; Janssen: Membership on an entity's Board of Directors or advisory committees, Research Funding. Dispenzieri:Celgene, Takeda, Prothena, Jannsen, Pfizer, Alnylam, GSK: Research Funding.
45
Agrawal, Sushma, Prabhakar Mishra, Punita Lal, Gaurav Agarwal, Amit Agarwal, Anjali Mishra, and saroj K. Mishra. "Risk scoring model to predict recurrence in locally advanced breast cancer (LABC) treated with neoadjuvant chemotherapy (NACT)." Journal of Global Oncology 5, suppl (October 7, 2019): 98. http://dx.doi.org/10.1200/jgo.2019.5.suppl.98.
Full textAbstract:
98 Background: Complete response (CR) to NACT portends favorable long term outcomes in LABC. There is a need for a tool to risk categorise patients for recurrence risk (RR), so that intensification of treatment can be offered to women with high risk of recurrence. Methods: A prospectively maintained database of LABC (between January 2007 to December 2012), who received NACT followed by definitive surgery, radiotherapy and endocrine therapy in endocrine sensitive disease was retrospectively analyzed for clinico-pathological and treatment factors affecting disease free survival (DFS). A risk scoring model was developed on the basis of beta coefficients of identified independent risk factors for DFS. Results: The incidence of loco-regional relapse was 8% and that of distant metastases was 32% in a dataset of 206 patients at a median follow-up of 47 months (IQR 24-62 mo). The independent risk factors for recurrence were index T stage [HR 1.8 (0.9-3.6)], N stage [HR 1.7 (0.4 – 4.7)], grade [HR 1.8 (0.8-4.2)], age less than and more than 40 years [HR 1.6 (0.4-0.9)], pathologic CR [HR 4.3 (1.7- 10.7)], intrinsic subtype [HR 2.2 (1.3-3.7)], and type of surgery (BCS vs MRM) [HR 2.2 (1.3-3.6)]. The ROC of the model for the prediction of recurrence was 0.67 (95 % CI: 0.61-0.75). The results of this model were validated by dividing the population into 3 risk groups: low risk (score less than 12), intermediate risk group (score between 13-15), high risk group (score 16 or more). The chances of recurrence are 16% versus 34% versus 57% in low, intermediate and high risk group respectively. Presence of three risk factors implies low risk, five intermediate and more than five high risk. Conclusions: The risk scoring model developed by us predicts RR and can be used for selecting patients for treatment intensification in high risk category.
46
Turesson, Ingemar, Stephanie Kovalchik, Ruth M. Pfeiffer, Sigurdur Y. Kristinsson, Lynn Goldin, Mark Drayson, and Ola Landgren. "Monoclonal Gammopathy Of Undetermined Significance and Risk Of Lymphoid and Myeloid Malignancies: 743 Cases Followed For Up To 30 Years In Sweden." Blood 122, no. 21 (November 15, 2013): 3124. http://dx.doi.org/10.1182/blood.v122.21.3124.3124.
Full textAbstract:
Abstract Background Monoclonal gammopathy of undetermined significance (MGUS) is a non-malignant condition with a prevalence of about 3 % in individuals 50 years or older. It carries a risk of transformation to multiple myeloma (MM) and other lymphoproliferative disorders (LPD) that has been estimated to be 1% per year. However, there is considerable variation in the risk of progression, and differentiating low-risk patients, who may not need further follow-up, from high-risk patients, who may warrant close monitoring or enrolment in early intervention studies, is a challenge. Several risk models for progression have been published. Non-IgG isotype, M-protein concentration >1.5 g/dL, and an abnormal serum free light-chain (FLC) ratio (normal reference: 0.26-1.65) were included in a model developed by the Mayo Clinic, which has not been independently validated with a large number of MGUS cases with long-term follow-up. We therefore assessed established risk score and explored novel risk factors for progression using a large independent cohort of 743 MGUS patients followed for up to 30 years. Methods We identified 743 MGUS patients in Malmö, Sweden who had serum collected at the date of detection of the M-protein and stored at – 200Celcius. MGUS diagnosis was made according to IMWG criteria: M-protein concentration <3 g/dL, absence of CRAB-criteria or clinical signs of otherLPD, and <10% plasma cells if bone marrow examination was performed. M-protein isotype, M-protein concentration and serum levels of FLC, albumin, creatinine, ß2-microglobulin, C-reactive protein and non-involved immunoglobulins were analysed in stored serum. We obtained the date of diagnosis of incident cancers from three sources: the nationwide Swedish Cancer Registry, the nationwide Patient Registry, and the Patient Registry of Malmö University Hospital. Using Cox regression models, we examined associations of demographic and laboratory factors with progression and determined the discriminatory power of three prediction models for progression. Results During 8,240 person-years of follow-up (median 11 years per subject), we observed 92 lymphoid and 10 myeloid malignancies in the study cohort, representing a cumulative risk of 16.0%. MM occurred in 53 patients and the 30-year cumulative risk was 10.8%; an approximate 0.5% annual risk. Three factors were significantly associated with progression: abnormal FLC-ratio (< 0.26 or >1.65), M-protein concentration >1.5g/dL, and depression of non-involved immunoglobulin isotype levels (immunoparesis). A prediction model with separate effects for these three factors and the M-protein isotype had higher discriminatory power than other models, though the differences were not statistically significant. The 30-year cumulative risk for myeloid malignancies was <2%. Conclusion In a large cohort of MGUS patients with long-term follow-up, we confirmed that an abnormal FLC-ratio predicts risk for progression of MGUS. This is an important observation as serum FLC assays are commonly used in the clinical setting. Furthermore, our results confirm the predictive value of a high M-protein concentration. A novel observation in our study is that the addition of immunoparesis to a multivariable model that includes independent effects for the factors of the Mayo Clinic risk model increases the discriminatory power to identify high-risk (versus low-risk) MGUS patients. Disclosures: Turesson: Celgene Corp: Honoraria.
47
Caron, Pablo A., Marcela A. Cruchaga, and Axel E. Larreteguy. "Sensitivity analysis of finite volume simulations of a breaking dam problem." International Journal of Numerical Methods for Heat & Fluid Flow 25, no. 7 (September 7, 2015): 1718–45. http://dx.doi.org/10.1108/hff-10-2014-0308.
Full textAbstract:
Purpose – The present work is a numerical study of a breaking dam problem. The purpose of this paper is to assess the effect of turbulence and surface tension models in the prediction of the interface position in a long-term analysis. Additionally, dimensional effects are analyzed by carrying out both 2D and 3D simulations. Design/methodology/approach – Finite volume simulations performed with the different models are compared between them and contrasted with numerical results computed using other numerical techniques and experimental data. Findings – The reported numerical results are in general in good agreement with experimental results available in the literature. They are also consistent with numerical solutions of other authors obtained using different numerical techniques. The results show that the laminar simulations exhibit strong mesh size dependency, while the turbulence models seem to help in producing mesh-independent solutions. Surface tension modeling does not seem to play a relevant role in the interface evolution. Practical implications – Model validation. Originality/value – The value of the present work encompass the comparison of different flow conditions used to simulate a free surface problem and their validation by contrasting numerical results with experiments. Also, the results shown in the present work are a contribution to the understanding of the role of some specific aspects of the models in the simulation of the proposed problem.
48
Bagheri, Ayoub, T. Katrien J. Groenhof, Folkert W. Asselbergs, Saskia Haitjema, Michiel L. Bots, Wouter B. Veldhuis, Pim A. de Jong, and Daniel L. Oberski. "Automatic Prediction of Recurrence of Major Cardiovascular Events: A Text Mining Study Using Chest X-Ray Reports." Journal of Healthcare Engineering 2021 (July 9, 2021): 1–11. http://dx.doi.org/10.1155/2021/6663884.
Full textAbstract:
Background and Objective. Electronic health records (EHRs) contain free-text information on symptoms, diagnosis, treatment, and prognosis of diseases. However, this potential goldmine of health information cannot be easily accessed and used unless proper text mining techniques are applied. The aim of this project was to develop and evaluate a text mining pipeline in a multimodal learning architecture to demonstrate the value of medical text classification in chest radiograph reports for cardiovascular risk prediction. We sought to assess the integration of various text representation approaches and clinical structured data with state-of-the-art deep learning methods in the process of medical text mining. Methods. We used EHR data of patients included in the Second Manifestations of ARTerial disease (SMART) study. We propose a deep learning-based multimodal architecture for our text mining pipeline that integrates neural text representation with preprocessed clinical predictors for the prediction of recurrence of major cardiovascular events in cardiovascular patients. Text preprocessing, including cleaning and stemming, was first applied to filter out the unwanted texts from X-ray radiology reports. Thereafter, text representation methods were used to numerically represent unstructured radiology reports with vectors. Subsequently, these text representation methods were added to prediction models to assess their clinical relevance. In this step, we applied logistic regression, support vector machine (SVM), multilayer perceptron neural network, convolutional neural network, long short-term memory (LSTM), and bidirectional LSTM deep neural network (BiLSTM). Results. We performed various experiments to evaluate the added value of the text in the prediction of major cardiovascular events. The two main scenarios were the integration of radiology reports (1) with classical clinical predictors and (2) with only age and sex in the case of unavailable clinical predictors. In total, data of 5603 patients were used with 5-fold cross-validation to train the models. In the first scenario, the multimodal BiLSTM (MI-BiLSTM) model achieved an area under the curve (AUC) of 84.7%, misclassification rate of 14.3%, and F1 score of 83.8%. In this scenario, the SVM model, trained on clinical variables and bag-of-words representation, achieved the lowest misclassification rate of 12.2%. In the case of unavailable clinical predictors, the MI-BiLSTM model trained on radiology reports and demographic (age and sex) variables reached an AUC, F1 score, and misclassification rate of 74.5%, 70.8%, and 20.4%, respectively. Conclusions. Using the case study of routine care chest X-ray radiology reports, we demonstrated the clinical relevance of integrating text features and classical predictors in our text mining pipeline for cardiovascular risk prediction. The MI-BiLSTM model with word embedding representation appeared to have a desirable performance when trained on text data integrated with the clinical variables from the SMART study. Our results mined from chest X-ray reports showed that models using text data in addition to laboratory values outperform those using only known clinical predictors.
49
Schinke, Maximilian, Inga Promny, Stefanie Hieke, Johannes M. Waldschmidt, Gabriele Ihorst, Milena Pantic, Justus Duyster, Ralph Wäsch, Martin Schuhmacher, and Monika Engelhardt. "Conditional Survival Analysis of 816 Multiple Myeloma (MM) Patients Constitutes a Different Way to Provide More Specifically Determined Prognosis." Blood 126, no. 23 (December 3, 2015): 2091. http://dx.doi.org/10.1182/blood.v126.23.2091.2091.
Full textAbstract:
Abstract Introduction: Disease monitoring based on genetics or other molecular markers obtained by noninvasive or minimally invasive methods will potentially allow the early detection of treatment response or disease progression in cancer patients. Investigations in order to identify prognostic factors, e.g. patient's baseline characteristics or molecular markers, contributing to long-term survival potentially provide important information for patients with multiple myeloma. Overall survival (OS) is not very informative for patients who already survived one or more years. To better characterize long-term survival respectively long-term survivors, conditional survival (CS) analyses are useful. Conditional survival (CS) describes probabilities of surviving t additional years given they survived s years and provides information, how prognosis evolves over time. We have demonstrated the use of CS in a large data set of multiple myeloma patients with long-term survival which is mandatory for the calculation of CS (Hieke,... Engelhardt, Schumacher. CCR 2015). Methods: We evaluated 816 consecutive multiple myeloma patients treated at our department from 1997 to 2011 with follow-up until the end of 2011. Patients' data were assessed via electronic medical record (EMR) retrieval within an innovative research data warehouse. Our platform, the University of Freiburg Translational Research Integrated Database Environment (U-RIDE), acquires and stores all patient data contained in the EMR at our hospital and provides immediate advanced text searching capacity. We assessed 21 variables including gender, age, stage and admission period. We calculated 5-years CS and stratified 5-years CS according to disease- and host-related risks. Component-wise likelihood-based boosting and variables selected by boosting were investigated in a multivariable Cox model. Results: The OS probabilities at 5- and 10- years were 50% and 25%, respectively. The 5-year CS probabilities remained almost constant over the years a patient had already survived after initial diagnosis (~50%). According to baseline variables, conditional survival estimates showed no gender differences. The estimated 5-year survival probabilities varied substantially, from 25% for patients ages 70 or older to 65% for patients younger than 60 years. Similarly, patients with D&S stage I have an estimated 5-year survival probability of about 75% compared with 40% for patients with D&S stages II and III. Significant risk factors via Cox proportional hazard model were D&S stage II+III, age >70 years, hemoglobin <10g/dl, ß2-MG ≥5.5mg/dl, LDH ≥200U/l. Renal impairment, low albumin and unfavorable cytogenetics increased the risk, but failed to reach significance. Cytogenetics, response, response duration and other risk parameters post treatment are currently included in our assessment. Of note, over the study period, admission of patients <60 years decreased from 60% to 34%, but increased for those ≥70 years from 10% to 35%, respectively, illustrating that not only young and fit, but also elderly patients are increasingly treated within large referral and university centers and that patient cohorts and risks do not remain constant over time. Conclusions: Conditional survival has attracted attention in recent years either in an absolute or relative form where the latter is based on a comparison with an age-adjusted normal population being highly relevant from a public health perspective. In its absolute form, conditional survival constitutes the quantity of major interest in a clinical context. We defined conditional survival by using the fact that the patient is alive at the prediction time s as the conditioning event. Alternatively, one could determine conditional survival, given that the patient is alive and progression-free or alive, but has progression at time s (Zamboni et al. JCO 2010). Analysis of the above and additional variables from diagnosis to prediction time s may refine conditional survival towards an even more specifically determined prognosis; follow-up response and risk parameters most likely further refining these CS analyses. Figure 1. Figure 1. Disclosures Wäsch: MSD: Research Funding; Janssen-Cilag: Research Funding; Comprehensiv Cancer Center Freiburg: Research Funding; German Cancer Aid: Research Funding.
50
Zhang, Mingzheng, Dehai Zhu, Wei Su, Jianxi Huang, Xiaodong Zhang, and Zhe Liu. "Harmonizing Multi-Source Remote Sensing Images for Summer Corn Growth Monitoring." Remote Sensing 11, no. 11 (May 28, 2019): 1266. http://dx.doi.org/10.3390/rs11111266.
Full textAbstract:
Continuous monitoring of crop growth status using time-series remote sensing image is essential for crop management and yield prediction. The growing season of summer corn in the North China Plain with the period of rain and hot, which makes the acquisition of cloud-free satellite imagery very difficult. Therefore, we focused on developing image datasets with both a high temporal resolution and medium spatial resolution by harmonizing the time-series of MOD09GA Normalized Difference Vegetation Index (NDVI) images and 30-m-resolution GF-1 WFV images using the improved Kalman filter model. The harmonized images, GF-1 images, and Landsat 8 images were then combined and used to monitor the summer corn growth from 5th June to 6th October, 2014, in three counties of Hebei Province, China, in conjunction with meteorological data and MODIS Evapotranspiration Data Set. The prediction residuals ( Δ P R K ) in NDVI between the GF-1 observations and the harmonized images was in the range of −0.2 to 0.2 with Gauss distribution. Moreover, the obtained phenological curves manifested distinctive growth features for summer corn at field scales. Changes in NDVI over time were more effectively evaluated and represented corn growth trends, when considered in conjunction with meteorological data and MODIS Evapotranspiration Data Set. We observed that the NDVI of summer corn showed a process of first decreasing and then rising in the early growing stage and discuss how the temperature and moisture of the environment changed with the growth stage. The study demonstrated that the synthesized dataset constructed using this methodology was highly accurate, with high temporal resolution and medium spatial resolution and it was possible to harmonize multi-source remote sensing imagery by the improved Kalman filter for long-term field monitoring.