« avril 2024 »
L M M J V S D
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30 1 2 3 4 5
 
Tous les évènements de Physique à venir

Tous les évènements de Biologie / Chimie à venir

Tous les évènements à venir

Les évènements relevant de la Physique et de la Biologie / Chimie sont représentés en turquoise

Deep learning approaches to improve the connectivity and robustness of vascular segmentation

Lundi 29 janvier 14:00 - Duree : 1 heure
Lieu : Conference room 2nd floor - LIPhy - 140 Avenue de la Physique - St Martin d’Hères + VISIO-CONFERENCE

Orateur : Odyssée MERVEILLE (Creatis, INSA Lyon)

You can attend the seminar in the LIPhy conference room or you can follow the live stream :

https://meet.univ-grenoble-alpes.fr/b/dor-7r9-f2b-dqa. The link will be activated shortly before the seminar.

Résumé :

The segmentation of blood vessels in medical images is challenging as they are thin, tortuous and organized in networks.

Despite more than two decades of research, achieving a complete and connected vascular network segmentation remains a challenge, in particular when dealing with 3D images. Yet, a geometrically accurate segmentation is critical for clinical applications such as cardiovascular disease computer-aided diagnosis and surgical planning.

End-to-end deep learning approaches have been extensively studied for medical image segmentation. However, state-of-the-art approaches yield sub-optimal results when applied to vascular segmentation. The segmentations tend to be disconnected, leading to poor results in downstream tasks such as blood flow simulations. Moreover, these learned models are not robust to domain shift, which drastically reduce their performances when applied to datasets without annotations.

In this talk, I will present several strategies that we proposed to improve the segmentation of vascular network with deep learning approach. First, I will discuss two approaches designed to enforce the connectivity of vascular networks. Then, I will present domain adaptation strategies that can be used to segment vascular image datasets from which no annotation is available.

Contact : Aurelien.Gourrier@univ-grenoble-alpes.fr



Prévenir un ami par email

Télécharger dans mon agenda

Cafés sciences de Grenoble | UdPPC de Grenoble | Sauvons Le Climat | Cafe des sciences de Vizille
Accueil du site | Secretariat | Espace privé | Suivre la vie du site RSS 2.0 : Tous les evenements Suivre la vie du site RSS 2.0 : Evenements de Physique Suivre la vie du site RSS 2.0 : Evenements de Biologie & Chimie