Title of research project

Virtual assistant for shoulder arthroscopy

Research engineer in AI for health

Contract type Contrat (CDD) Category A Workload Fulltime

Scientific field Informatique

WorklJob Descriptionoad

Description of the offer The shoulder is a complex joint stabilized by multiple bones, tendons and ligaments. An injury to any of these structures can have a disabling effect and produce significant pain in the patient. When pain medications and physical therapy no longer relieve pain or improve function, surgical treatment with arthroscopic surgery is required. Arthroscopy [DeMarinis2024, Levin2024] is a surgical procedure used to diagnose and treat, among other things, problems of the shoulder joint. It allows the surgeon to examine the inside of it using a small camera (arthroscope) inserted through a small incision. Recent advances in Deep Learning [LeCun2005] make it possible to study approaches based on neural networks to solve complex problems. These networks are resource intensive, often making them difficult to deploy for real-time analytics on interactive devices. This thesis focuses on five scientific challenges: - Analyze the quality of the video and images captured during the examination; - Identify, using a supervised learning method, the different tissues that make up the shoulder and which appear on the arthroscope image; - Learn to differentiate healthy tissues from diseased tissues; - Take part in the production deployment of the solution on an arthroscopy column; - Evaluate and validate the model during the different phases of the process. Advances in Artificial Intelligence now make it possible to consider a virtual assistant to help young surgeons learn, understand and perform shoulder injury surgeries.

Field of research IA, Health, Deep Learning, Vision

Skills Programming, Research

Knowledge IA, Deep Learning, Python

Work environment

Location Bâtiment Georges Sand

Laboratory La Maison des Sciences Numériques

Degree level

Degree required Be a graduate of an Engineering course or a Master 2 in computer science with an Artificial Intelligence focus and demonstrate an excellent academic profile.

Experience required Have initial experience (which may have been acquired during training) in video processing using AI and deep learning.

WorkloadInformations complémentaires

Start date 02 01 26

Job Type Contrat (CDD)

Composition of the selection jury

(Président(e)) Melanie Courtine

Younès BENNANI

Application deadline 02 01 26

Apply

Candidater à cette offre

You have a USPN account:

USPN connection

You already have an account:

You don't have an account, create new

New account

Back to list Fermer