Ocular diseases leading progressively to the loss of vision such as the diabetic retinopathy destroy lives and have a huge economic impact on society by stopping the patient’s social activity. To preserve the sight usually implies to detect the disease before it became irremediable. However the current means available to detect the point where the disease turns into a maculopathy, and threatens the sight have too many limitations: 1) detection is too late; 2) fundus diagnosis can underestimate or miss the extent of the damages as the disease spreads below the tissues surface; 3) there is a growing quantity of images to analyze, which challenges the capacities of the ophthalmologists; 4) each disease is treated individually by Machine Learning (ML) laboratories, while they have symptoms in common. In this project, we aim at determining Artificial Intelligence methods to allow using ML to address such diagnosis problem on population of small sizes using OCT. Elaborating smart symptoms-specific (SYS) features extractors (FE) in place or generic ones would compact the problem and allow solving new Deep Learning (DL) diagnosis problems through a quick recomposition of the set of SYS-FE to use by matching them to the disease-associated symptoms list. The project will be collaborating with the OCTAGON project of the OEIl laboratory in the National University of Singapour, which is developing AI solutions for ophthalmology, and with Dr Jean Louis Uzel, of the Cardella Clinic in Tahiti.
Keywords: Image processing, artificial intelligence, machine learning, Median Learning, Diabetic Maculopathy.
Funding: This project will be funded by a three years doctoral grant from the Fondation pour la Recherche Médicale (Foundation for Medical Research). The Project + Candidate must be selected by the FRM for final acceptation.
Supervision: The project will be co-supervised by Dr. Jean Martial Mari
(http://www.upf.pf/fr/content/jean-martial-mari), Associate Professor at the University of French Polynesia, and by Dr. Michael Girard (http://www.bioeng.nus.edu.sg/OEIL/), Assistant Professor at the National University of Singapore. Researches will be performed in collaboration with Dr Alexandre Thiery, from the National University of Singapore, and Dr Jean Louis Uzel, from the Cardella Clinic, Papeete, Tahiti. The research will take place in the
GePaSud laboratory, Tahiti, French Polynesia, and may lead to scientific visits to Singapore (funds already obtained).
Applicant: The applicant should have a master degree in Computer Science; experience in Machine Learning and/or related fields, Medical Imaging and Mathematics would be desirable, but all profiles will be considered for a first round of interviews. Requests from people requiring specific arrangements will be answered within our capacities, but overseas interviews will be on Skype and the applicant must be able to travel to Tahiti.
Application: Applicants should email Dr. Jean Martial Mari (email@example.com), along with a CV and cover letter.
Application deadline: April 06, 2018 (06/04/2018).