En médecine, le diagnostic est la démarche par laquelle le vétérinaire, médecin, généraliste ou spécialiste, le kinésithérapeute, la sage-femme ou le chirurgien-dentiste, ou encore le psychologue au Canada, détermine l'affection dont souffre le patient, et qui permet de proposer un traitement. Il repose sur la recherche des causes (étiologie) et des effets (symptômes) de l'affection ; on parle aussi de « tableau clinique ». Le diagnostic du médecin Typiquement, le diagnostic se déroule en deux parties :

Anamnèse (ou histoire de la maladie) : c'est l'étape la plus cruciale de la consultation. Le médecin y écoute son patient, d'abord librement, puis en l'orientant par des questions. L'anamnèse permet de connaître les antécédents du patient, les symptômes ressentis, l'ancienneté de la maladie et son évolution, les traitements déjà suivis, les attentes du patient…

Examen physique (à l'issue de l'anamnèse, le médecin a souvent une idée assez précise de la maladie ; l'exa

BPMN process generation and optimization for readability in the context of telemedecine

Charline Montial

In business, automated processes can be specified using the BPMN standard, which stands for Business Process Model and Notation. It is presented as flowcharts with standardized graphical elements to show the possible flows and activities of a process. These diagrams can be seen as a graph with logical expressions conditioning their edges. The telemedecine start-up Soignez-moi.ch uses BPMN to describe complex medical forms in a readable way. Soignez-moi.ch offers medical consultations via an online platform. The first step for the patient is to fill a questionnaire about his/her symptoms. In the second step, a doctor accesses to the answers and proceeds to call the patient for the teleconsultation. The complexity of these medical forms comes from the fact that the questions are adapted to the patient answers, which leads to different possible flows. Complex forms from another telemedecine website, Diaana, had to be translated to BPMN to be compatible and used at Soignez-moi.ch. Diaana is a contraction of the words diagnosis and anamnesis. The way these forms are specified is very different from what has been modeled in Soignez-moi.ch: instead of having separated branches, there is a sequential list of questions. In this list, every question is evaluated on its own to see if it is asked or not. The question conditions are Boolean expressions. If a direct translation to BPMN is performed, there is no branching to show the relationships between the questions. The aim of my work is to highlight similarities in the branch conditions to make the diagrams more readable and editable by humans. Doctors and collaborators read the medical BPMN processes and may modify them. It should be a pleasant experience for them to go through the diagrams. They should quickly understand in a visual way if questions are related, and how they are related. The BPMN process readability has been achieved by extracting patterns, showing the implications between the questions to have a readable diagram with meaningful branches. This translation has been applied to 18 questionnaires. Two questionnaires were not improved because they were only composed of one question. One questionnaire was also not improved because all its questions were unconditional, meaning that no relationship between the question conditions could be highlighted. Another questionnaire with conditional questions was not improved because its conditions were unrelated. 14/18 questionnaires have had their readability improved.

2022

Graph Representation Learning in Computational Pathology

Guillaume Jaume

Advances in scanning systems have enabled the digitization of pathology slides into Whole-Slide Images (WSIs), opening up opportunities to develop Computational Pathology (CompPath) methods for computer-aided cancer diagnosis and prognosis. CompPath has been primarily developed using models based on Convolutional Neural Networks (CNNs), building on the recent successes in Computer Vision. A series of promising approaches have been proposed for nuclei segmentation and classification, tumor detection, tumor grading, among others. However, CNN-based methods suffer from several limitations. First, it is challenging to model both fine-grained nuclei-level information and long-range inter-glandular dependencies. Second, there is a discrepancy between the pixel-based analysis of CNNs and the histological entity-centered analysis employed for pathological diagnosis, which in turn can hinder model transparency. Third, the inherent complexity of training networks on large histology images with limited annotations constrains its learning capabilities.Instead, we propose an analytical paradigm shift, where we view and analyze histology images as a set of biological entities interacting with each other. Specifically, an image is represented as an entity-graph where nodes depict biological entities and edges encode interactions between these entities. Entity-graphs are further processed by a Graph Neural Network (GNN) model to jointly encode the entity morphological attributes and topological distribution, towards tissue phenotyping. In this thesis, we study three research directions in CompPath, namely, scalability, interpretability and explainability, and weakly-supervised learning.First, histology images are orders of magnitude larger than natural images, where diagnostically relevant regions can represent only a fraction of the image. We propose a scalable hierarchical cell-to-tissue representation (HACT) and GNN model, HACT-Net, for learning on arbitrary large inputs. We show the capabilities of HACT-Net on our proposed BR ACS dataset, the largest cohort to date for breast tumor Regions-of-Interest subtyping.Second, computer-aided diagnostic tools must be transparent and their decision-making process justified. By shifting the analysis from pixel- to entity-based, we make the input space interpretable for pathologists that can better relate to the model input. We further propose entity-centric graph explainers, exemplified with cell-graph model explainability, along with novel metrics to evaluate explanations based on entity-level pathological concepts.Third, acquiring ground-truth data to train deep learning systems requires pathologists to provide specific annotations, which is time-consuming, expensive, and subject to inter- and intra-observer variability. We therefore propose WholeSIGHT, a method that reduces annotation requirements to WSI-level labels only, for joint classification and segmentation of WSIs. We show the capabilities of WholeSIGHT for Gleason pattern segmentation and grading on multi-source WSI prostate datasets. The generalization properties of WholeSIGHT are further evaluated on unseen cohorts, and compared to Bayesian variants to strengthen the estimation of the model uncertainty. Finally, we introduce HistoCartography, a novel python library designed to accelerate the development of graph analytics in CompPath.

EPFL2022

Automation of the diagnosis process in the railway system : Detection of defects in concrete sleepers using vision-based machine learning models

Linah Charif

inspectors that walk over the track and check the defects on the rail surface, fasteners and sleepers. In the case of concrete sleepers, rail inspectors classify defects according to their size and occurrence over 20 sleepers. The manual inspection is error prone, it is time consuming and it can be sometimes risky. In an eort to overcome these issues, SBB has developed a project called AISI (Articial Intelligence Streckeninspection) in order to automatize the rail inspection by collecting and analysing images with diagnosis trains equipped with cameras. The objective of this thesis is integrated within the AISI Project, since its goal is the conception of a deep learning model to detect and classify concrete sleepers defects using computer vision and more specically, object detection. The defects studied in this thesis are cracks and akings (or spallings). In the literature, many researchers have tackled the subject of defect detection in concrete, especially in civil infrastructures. The most common methods a combination of image-processing and Machine Learning (image pre-processing and a classier) whereas the latest methods use deep learning (CNN, FCN and RPN, YOLO). The proposed methodology uses a Fast-RCNN model developed with Tensor ow Object Detection API. The evaluation of the model is given on a total of 996 defects and its precision and recall are given. The model shows average results as the number of false positives is high. It has been noted that the labels provided needed to be improved. The proposed approach has a great potential of development since a large amount of training data can be added thanks to the produced relabelling process. Also, the designed pipeline establishes a base that can be adjustable for various algorithms and congurations.

2021

BPMN process generation and optimization for readability in the context of telemedecine

Charline Montial

2022

Graph Representation Learning in Computational Pathology

Guillaume Jaume

EPFL2022

Diagnostic (médecine)

Examen physique (à l'issue de l'anamnèse, le médecin a souvent une idée assez précise de la maladie ; l'exa

BPMN process generation and optimization for readability in the context of telemedecine

Graph Representation Learning in Computational Pathology

Automation of the diagnosis process in the railway system : Detection of defects in concrete sleepers using vision-based machine learning models

BPMN process generation and optimization for readability in the context of telemedecine

Automation of the diagnosis process in the railway system : Detection of defects in concrete sleepers using vision-based machine learning models

Graph Representation Learning in Computational Pathology