Дисертації з теми "SaaS tool"

Le numérique est devenu une partie intégrante de notre vie quotidienne, mais il a un impact environnemental important, notamment en termes de consommation d’énergie, de production de gaz à effet de serre, de déchets électroniques, de pollution de l’eau et de l’air, de déforestation, et de diminution de la biodiversité. Les organisations prennent conscience de leur responsabilité environnementale et cherchent à réduire leur empreinte carbone liée au SI (système d’information). Pour aider les organisations à piloter leur empreinte carbone SI, cette thèse propose un cadre d’aide à la décision basé sur une modélisation fine de l’impact environnemental du système d’information. Ce cadre permet d’identifier les actions les plus efficaces pour réduire l’impact environnemental du SI en prenant en compte les contraintes opérationnelles et financières des organisations. Les outils WeNR sont présentés comme des applications concrètes de ce cadre d’aide à la décision, qui permettent de mesurer l’impact environnemental du SI au sein des organisations et de proposer des mesures concrètes pour le réduire. Enfin, l’auteur présente les perspectives d’un outil SaaS d’aide à la décision dans le numérique responsable qui permettrait de collecter automatiquement les données relatives à l’empreinte carbone numérique de l’organisation, de les analyser et de proposer des actions spécifiques pour la réduire, ainsi que de fournir des indicateurs clés de performance environnementale<br>Digital technology has become an integral part of our daily lives, but it has a significant environmental impact, particularly in terms of energy consumption, greenhouse gas production, electronic waste, water and air pollution, deforestation, and biodiversity loss. Organizations are becoming aware of their environmental responsibility and are seeking to reduce their carbon footprint related to information systems. To help organizations manage their IS carbon footprint, this thesis proposes a decision support framework based on a detailed modeling of the environmental impact of the information system. This framework identifies the most effective actions to reduce the environmental impact of the IS, taking into account the operational and financial constraints of the organizations. The WeNR tools are presented as concrete applications of this decision support framework, which allow measuring the environmental impact of the IS within organizations and proposing concrete measures to reduce it. Finally, the author presents the prospects of a responsible digital SaaS decision support tool that would automatically collect data on the organization’s digital carbon footprint, analyze it, propose specific actions to reduce it, and provide key environmental performance indicators

Breast cancer is the most common cause of cancer death in Australian women. Current pathological analysis examines a small section of tissue for cellular and plasma abnormalities using a light microscope. However, this method of diagnosis, despite being the current gold standard, has its limitations, where human error and professional experience can influence a patient’s diagnosis. A potential alternative or adjunct to conventional histopathology for classifying tissue disease status is offered by Small Angle X-ray Scattering (SAXS). At the time of commencement of this work, there had been several small scale studies which examined the potential of SAXS to classify the disease status of breast tissue. These tended to focus on the supramolecular structure of collagen fibrils found in the breast, where it is known that the degradation of these fibres is related to the spread of disease. Most previous studies also used a synchrotron as an X-ray source, due to the intense and highly collimated flux available. This study used a synchrotron source, but also evaluated the use of a laboratory X-ray source, as a more convenient and relatively inexpensive alternative that could one day find application in the clinic. The work presented in this thesis analyses the largest cohort of patients and breast tissue samples studied to date using SAXS: 130 patients with 543 tissue samples. Tissues were sourced from surgical waste and classified into four groups: invasive carcinoma, benign, normal, and mammoplasty. Mammoplasty tissue samples were harvested from patients undergoing breast reduction and/or reconstruction, where no history or presence of disease was indicated. Normal tissue was sampled from patients with known disease, but pathological analysis of the tissue core diagnosed it as normal. A comprehensive analysis of the scattering patterns was carried out, analysing features arising from the collagen structure and orientation, the total scattered intensity, and adipose tissue in the breast. Features related to the axial D-spacing of the collagen fibrils within the breast tissue as well as the integrated scattering intensity (called amorphous scatter) demonstrated the highest ability to discriminate tissue types, in SAXS images acquired from both the synchrotron source and the laboratory X-ray source. The amorphous scatter intensities obtained using a synchrotron source showed highly significant differences (p < 0.01) for almost all of the tissue pair comparisons: invasive carcinoma vs. benign, invasive carcinoma vs. normal, invasive carcinoma vs. mammoplasty, benign vs. mammoplasty, and normal vs. mammoplasty. However, no significant difference was seen in the amorphous scatter between benign versus normal tissues (p = 0.30). The amorphous scatter values increased with severity of disease, i.e. it was the highest for invaded tissues and decreased progressively from benign to normal to mammoplasty. There was a significant difference between normal and mammoplasty tissue types using the amorphous scatter as a discriminator (p = 0.0025). Pathological assessment cannot differentiate between these two tissue types, which suggests that there may be changes occurring in these tissue structures at the supramolecular level that can be characterised using SAXS. The ability of SAXS to reveal structural differences between normal and mammoplasty tissue types is highly significant, for both disease diagnosis and treatment, as well as for understanding disease progression. For example, these differences might aid in determining surgical margin clearance of excised breast lesions as well as potentially provide a means of pre-screening or perhaps improve false-negative rates of diagnosis. The potential of SAXS to reveal macroscopic extent and directional spread of disease was explored using two-dimensional mapping of the amorphous scatter. These maps showed broad agreement with histopathological diagnosis, but further investigation regarding their reliability and interpretation for clinical utility is still needed. Changes in both the amorphous scatter and the axial D- spacing were seen in tissue samples up to 6 cm away from the primary site of disease. In particular, a significant decrease in both parameters was seen between the centre of the tumour (at 0 cm) and 2 cm away, suggesting that closer examination of the tissue structures over the disease/healthy tissue border may provide information regarding the mechanisms of metastasis and growth of cancerous tumours. The combination of the amorphous scattering results from the two X-ray sources indicates that the size of the scatterers may be the key in classifying tissue types. The synchrotron source was able to access a lower q-range (q = 0.1-0.6 nm-1) and the laboratory source covered a larger q-range (q = 0.25-2.3 nm-1). Mammoplasty tissues appear to be characterised by large scattering components (d > 25.13 nm), whereas normal tissues are characterised by slightly smaller scattering components (10.47 nm < d < 25.13 nm) and benign tissues by even smaller scattering components (4.83 < d < 10.47 nm). It appears that the size of the scatterers contributing to the total scattering intensity decreases with severity of disease, which was seen independently with both X-ray sources. Further investigation is warranted to determine the biological origin of these differences. These results also suggest that the optimum SAXS instrument may need to cover a scattering vector range of q < 0.25 nm-1 to identify differences in healthy tissue types, and q > 2.3 nm-1 to possibly investigate invasive carcinoma tissue types. A SAXS apparatus that can examine a large q-range may provide all of the necessary information from the amorphous scatter to differentiate between tissue groups. The periodic structure of collagen fibrils along their longitudinal axis can be characterised by the axial D-spacing, where this spacing was found to change with the presence of disease. The axial D-spacing for healthy breast tissues was found to be significantly lower in normal and mammoplasty tissues compared to invaded tissues (p = 0.0050 and p = 0.0093, respectively). However, no significant differences between the other tissue group pairs were seen (p > 0.05). These differences were evident in classification modelling of the four tissue groups, where the amorphous scatter and the amplitude of a collagen axial peak were used to build a probability model for disease status. The model showed high sensitivities (> 70%) and widely variable specificities (ranged from 18-97%) for the data examined with the synchrotron source. This means that the model was a good indicator of disease, but poor at indentifying healthy tissue types. The work presented in this thesis shows that SAXS is capable of distinguishing breast tissue types with high sensitivity and has the potential to become a significant tool for the investigation of cancer progression or even diagnosis. Further investigation into the amorphous scatter and axial D-spacing in particular may provide insight into the biological mechanisms related to tissue degradation associated with invasive disease.