In the same section
- Extraction of robust morphological feature from medical images (CT-scan) of the aorta in the context of the Marfan syndrome.
Heritable Thoracic Aortic Diseases (HTAD) are genetic vascular diseases presenting thoracic aortic dilation, aneurysms or dissections on one or more aortic segments. Mortality of people with HTAD is mostly determined by aortic root aneurysm dissection and rupture. Aortic dilation is generally not detected at an early stage.
Existing aorta measurement techniques are limited, and can deviate up to 10% depending on the examiner. This has implications for early diagnosis and for an optimal timing of preventive surgery.
Measurement inconsistency hampers the re-use of data and building large consistent datasets.
The project aims to help in extracting robust morphological feature from medical images (CT-scan) of the aorta.
The project will be done in collaboration with Fondation 101 Génomes.
Collaborations:
Assistance Publique des Hôpitaux de Paris (APHP), ETRO-VUB Department of Electronics and Informatics and FARI – AI for the Common Good Institute
Contact:
Christine Decaestecker cdecaes@ulb.ac.be, odebeir@ulb.ac.be
- MRI/CT-Style transfer
- Fiabilité des mesures d’un nouveau logiciel de détection et suivi semi- automatique des paramètres architecturaux musculaires.
- Implémenter une détection automatique des éléments structurelles du muscle (Par le biais d’un réseau de neurone). Actuellement l’utilisateur du logiciel doit détecter manuellement, sur la première image de la vidéo, les fascicules et les aponévroses avant de lancer le tracking.
- Optimiser l’algorithme de tracking et les mask utilisés pour améliorer la reproductibilité du tracking d’image en image.
Contacts:
olivier.debeir@ulb.be, Hans.Bourgeois@ulb.be
- Évaluation volumétrique méniscale par reconstruction RMN-3D
- Deep neural network image processing
Please register to UV for PROJH419 for more details
contact:
Olivier.debeir@ulb.be
- Mise en oeuvre de segmentation 3D organites cellulaires (imagerie electronique) FLIPSEM
https://openorganelle.janelia.org/
https://www.nature.com/articles/s41586-021-03992-4
Useful Dragonfly links:
https://pages.zeiss.com/Support.html
en collaboration avec CMMI (David Perez)
- tbd [ASSIGNED]
tbd
- tbd [ASSIGNED]
tbd
- tbd [ASSIGNED]
tbd
- RSNA-MICCAI brain tumor radiogenomic classification (MRI) [ASSIGNED]
Challenge : RSNA-MICCAI brain tumor radiogenomic classification (MRI)
Link : https://www.kaggle.com/c/rsna-miccai-brain-tumor-radiogenomic-classification
- tbd [ASSIGNED]
tbd
- Chest radiography [ASSIGNED]
- a short title of the selected challenge for your project :
1) Lung pattern classification for different types of pneumonia (Virus or Bacteria) using Convolutional Neural Network (CNN)
2) Lung pattern classification for covid-19 detection using Convolutional Neural Network (CNN)
link to the challenge description :
1) https://www.kaggle.com/c/pneumoniabacteriavirus/overview
Dataset : https://www.kaggle.com/paultimothymooney/chest-xray-pneumonia
2) https://www.kaggle.com/c/1056lab-covid19-chest-xray-recognit
Dataset : https://www.kaggle.com/tawsifurrahman/covid19-radiography-database
I am equally interested in the 2 subjects. Therefore, I let you choose the one that suits you best.
BIOMED/INFO : BIOMED
- Analyse statistique des objets présents sur une lame (wholeslide)
Le projet consiste à faire l'inventaire sur base d'une recherche non supervisée d'objets (clustering). Un pré-processing consistant à un encodage par DNN sera envisagé. Les images sont acquise par un scanner de lame à champ clair.
Contact:
olivier.debeir@ulb.be
- Tracking of endoscopic position using morphologic landmark recognition