Annonce

Les commentaires sont clos.

Few shot medical image classification

20 Novembre 2023

Catégorie : Stagiaire

Nowadays, deep learning has achieved a breakthrough in Artificial Intelligence (AI). However, huge, labeled datasets are needed to train on [1]. Collecting such extensive annotated data is time and resources consuming and it is not feasible in the medical domain. Indeed, unlike the case of natural images, where annotations can be easily performed by non-experts, medical images require careful and time-consuming analysis from experts such as radiologists. The limited availability of annotated medical imaging data remains the biggest challenge to the success of deep in the medical domain [2].

This internship aims to develop few shot medical image classification methods for computer aided diagnosis. This internship focuses on creating a representation learning approach that extracts clinically relevant information from medical images. Inspired by recent advancements in AI, we will explore the self-supervised learning technique [3], leveraging domain-specific knowledge and the 3D context of medical images [4][5]. The proposed approach will be evaluated on some medical diagnosis tasks using medical images. we will focus on two trendy and widely used medical images for diagnosis and prognosis: Computed Tomography (CT) [7] and Magnetic Resonance Imaging (MRI) [8].

This internship provides an exciting opportunity to tackle a critical challenge in medical image analysis while gaining valuable experience in deep learning and healthcare. We look forward to collaborating with passionate individuals dedicated to innovation in patient care.

References :

[1]Ouahab, Achraf, Olfa Ben-Ahmed, and Christine Fernandez-Maloigne. "A Self-attentive Meta-learning Approach for Image-Based Few-Shot Disease Detection." MICCAI Workshop on Resource-Efficient Medical Image Analysis. 2022..

[2]Razzak, M. I., Naz, S., & Zaib, A. (2018). Deep learning for medical image processing: Overview, challenges and the future. Classification in BioApps, 323-350.

[3]Azizi, Shekoofeh, et al. "Big self-supervised models advance medical image classification." Proceedings of the IEEE/CVF International Conference on Computer Vision. 2021..

[4]Xie, Xiaozheng, et al. "A survey on incorporating domain knowledge into deep learning for medical image analysis." Medical Image Analysis 69 (2021): 101985

[5]Williams, Lauren H, and Trafton Drew. “What do we know about volumetric medical image interpretation? a review of the basic science and medical image perception literatures.” Cognitive research: principles and implications vol. 4,1 21. 8 Jul. 2019, doi:10.1186/s41235-019-0171

[6]Jiang, Yifan, et al. "Few-shot learning for CT scan based covid-19 diagnosis." ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2021.

[7]Chen, D., Zhang, L., & Ma, C. (2020, November). A multimodal diagnosis predictive model of Alzheimer’s disease with few-shot learning. In 2020 International Conference on Public Health and Data Science (ICPHDS) (pp. 273-277). IEEE.

Location: XLIM, University of Poitiers, France
Duration: 6 months (starting from March 2023), with the potential for transition into a PhD program based on performance and mutual interest.
Salary: 600 € monthly + housing expenses covered for the 3 first months
Keywords: Deep Learning, medical images, diagnosis

Qualifications:

Master/Engineering degree in the following fields: computer science, data science, or applied mathematics
Experience with deep learning; Skills and proven experience in programming (Python and frameworks such as Pytorch/Tensorflow);
Interest in DL mathematical theory would be a plus
Experience in medical image processing will be appreciated
Good communication skills (spoken/written English) are required

Application:

Send CV, motivation letter, Last University transcripts, and two reference letters, as attachments of an email with "Application__ANR_MIMIC_Inter position " as subject to olfa.ben.ahmed@univ-poitiers.fr