Vous êtes ici : Réunions » Réunion

Identification

Identifiant: 
Mot de passe : 

Mot de passe oublié ?
Détails d'identification oubliés ?

Recent advances in machine learning for computer aided diagnosis and prognosis based on medical imaging

Nous vous rappelons que, afin de garantir l'accès de tous les inscrits aux salles de réunion, l'inscription aux réunions est gratuite mais obligatoire.

Inscriptions closes à cette réunion.

Inscriptions

58 personnes membres du GdR ISIS, et 50 personnes non membres du GdR, sont inscrits à cette réunion.
Capacité de la salle : 200 personnes.

Annonce

La réunion aura lieu le 01 Décembre 2020 à 9h15 sur Zoom


In recent years, artificial intelligence, especially machine learning has received a lot of attention to explore and structure multidimensional and multimodality medical imaging data, especially for the design of diagnosis models, aiming at detecting, localizing and characterizing pathological patterns in the data. Some academic works also recently explored the potential of artificial intelligence for predicting the course and outcome of diseases. This one-day workshop intends to gather researchers in deep machine learning, computer vision and/or medical image analysis as well companies and AI-based startups in the medical image field. We will start by reviewing state-of-the art achievements in the domain of computer aided diagnosis (including patient screening, detection, segmentation..) and prognosis models based on medical imaging for different clinical applications. Then, we will cover some challenges that need to be addressed to foster the development of these diagnosis and prognosis models. This includes methodological questions such as uncertainty and interpretability of the deep learning based models, as well as strategies regarding the evaluation framework of models performance (challenges, standardisation of the performance metrics..) and the access to structured medical image database.

This one-day workshop is organized jointly by the action "Analysis, processing and decision for massive and multimodal data in life sciences" of theme B Image and Vision, and the transversal theme T Machine learning for signal and image analysis.

We will have three keynote presentations by:

Mathieu De Craene, Research Scientist at Philips, Paris

Pierrick Coupé, DR CRNS au LabRI, Bordeaux

Ivana Isgum, University Professor AI and Medical Imaging at Amsterdam University Medical Center

Organisateurs

Programme

9h15-9h30 : Introduction

9h30-10h30 : AI for analysis of cardiovascular disease

Ivana Isgum, University Professor AI and Medical Imaging at Amsterdam University Medical Center

10h30-11h30 : Use of AI in Cardiology: standardizing image acquisition and workflow in echocardiography to open perspectives in computer-aided diagnosis

Mathieu de Craene, Philips, Paris

11h30-12h30 : Quantitative MRI Analysis:From voxel to knowledge

Pierrick Coupé, LabRI, Bordeaux

12h30-12h45 : Conclusion

Résumés des contributions

Quantitative MRI Analysis:From voxel to knowledge

Pierrick Coupé, LabRI, Bordeaux

- This talk will present our work dedicated to quantitative MR analysis. First, I will introduce our brain segmentation methods based on patch-based framework and large ensemble of deep networks. Afterwards, I will describe our open access platform integrating the presented brain segmentation tools. The volBrain platform is now used by more than 4500 users worldwide. Finally, I will present our recent BigData studies using the volBrain platform dedicated to normal ageing and Alzheimer?s diseases. Here, I will present the new knowledge that we produced thanks to such massive processing strategy.

AI for analysis of cardiovascular disease

Ivana Isgum, University Professor AI and Medical Imaging at Amsterdam University Medical Center

-

Use of AI in Cardiology : standardizing image acquisition and workflow in echocardiography to open perspectives in computer-aided diagnosis

Mathieu de Craene, Philips, Paris

-Dans cette présentation, nous présenterons quelques exemples d'application d'algorithmes de Machine Learning à différentes phases de l'examen échocardiographique. Tout d'abord, lors de la phase d'acquisition, des classifieurs ou régresseurs peuvent estimer en temps réel la qualité de la vue, ou de l'image. Ensuite lors de la phase de quantification, nous montrerons des applications de l'apprentissage profond en vue de quantifier et automatiser certaines tâches, comme la quantification du mouvement cardiaque par speckle tracking et le contourage de traces Doppler. Enfin, dans la phase de stratification et comparaison de patients, nous donnerons l'exemple d'une approche de réduction de dimensionalité afin d'analyser des échos de stress.

Date : 2020-12-01

Lieu : en visio-conférence


Thèmes scientifiques :
B - Image et Vision
T - Apprentissage pour l'analyse du signal et des images

Inscriptions closes à cette réunion.

Accéder au compte-rendu de cette réunion.

(c) GdR IASIS - CNRS - 2024.