Skip to Main content Skip to Navigation
Conference papers

Organ Segmentation in CT Images With Weak Annotations: A Preliminary Study

Rosana El Jurdi 1 Caroline Petitjean 2, 1 Paul Honeine 1 Fahed Abdallah 3, 4
1 DocApp - LITIS - Equipe Apprentissage
LITIS - Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes
2 QuantIF-LITIS - Equipe Quantification en Imagerie Fonctionnelle
LITIS - Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes
Abstract : Medical image segmentation has unprecedented challenges, compared to natural image segmentation, in particular because of the scarcity of annotated datasets. Of particular interest is the ongoing 2019 SegTHOR competition, which consists in Segmenting THoracic Organs at Risk in CT images. While the fully supervised framework (i.e., pixel-level annotation) is considered in this competition, this paper seeks to push forward the competition to a new paradigm: weakly supervised segmentation, namely training with only bounding boxes that enclose the organs. After a pre-processing step, the proposed method applies the GrabCut algorithm in order to transforms the images into pixel-level annotated ones. And then a deep neural network is trained on the medical images, where several segmentation loss functions are examined. Experiments show the relevance of the proposed method, providing comparable results to the ongoing fully supervised segmentation competition.
Complete list of metadatas

Cited literature [10 references]  Display  Hide  Download

https://hal-normandie-univ.archives-ouvertes.fr/hal-02183031
Contributor : Paul Honeine <>
Submitted on : Sunday, July 14, 2019 - 6:18:51 PM
Last modification on : Friday, May 15, 2020 - 12:00:03 PM

File

19.gretsi.weak.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02183031, version 1

Citation

Rosana El Jurdi, Caroline Petitjean, Paul Honeine, Fahed Abdallah. Organ Segmentation in CT Images With Weak Annotations: A Preliminary Study. 27-ème Colloque GRETSI sur le Traitement du Signal et des Images, Aug 2019, Lille, France. ⟨hal-02183031⟩

Share

Metrics

Record views

194

Files downloads

108