how artificial intelligence helps fight knee osteoarthritis

Last May, the project Development of a functional score to guide the diagnosis and help the therapeutic recommendation of knee pathologies “ he was among the winners of the “For innovations for the future” call by Axelys, a non-profit organization from Quebec. The scientific team of the project included professors from the Department of Science and Technology of the TÉLUQ University, Khadidja Henni and Neila Mezghani, also a researcher at the research center in Chum, as well as Nicola Hagemeister, from the École de technologie supérieure (ÉTS). researcher at CHUM, with which the Quebec company EMOVI was associated.

Osteoarthritis of the knee, or gonarthrosis, is the wear and therefore the destruction of the articular cartilage in various areas of the knee joint. This chronic disease often linked to aging affects women more often, it can also be due to overweight mini-traumas, repeated during professional or sports activities or of genetic origin.

Neila Mezghani and Nicola Hagemeister jointly conducted various studies, they are currently working as part of the project team ” Automatic classification of the kinematic data of the knee and its application to the diagnosis of pathologies “.

The goal is to develop an automatic kinematic data classification system, apply it for the implementation of a new knee pathology diagnostic technology and integrate it with the KneeKG knee motion analysis system.


Kneegraphy, an evaluation of the movement of the knee, allows to understand the origin of the pain felt by the patient and to create a personalized exercise program based on the observed deficits.

It is performed using the KneeKG System, an FDA approved, Health Canada approved, CE marked wearable medical device that quickly provides accurate, reliable, real-time data that quantifies knee joint motion. This tool, developed at the Imaging and Orthopedics Research Laboratory (LIO) of the École de technologie supérieure (ÉTS) in particular by Nicola Hagemeister, is the result of twenty years of collaborative work involving ÉTS, the Center of the Hospital Center of the University of Montreal (CHUM) and TÉLUQ University.

The KneeKG system consists of three elements: an exoskeleton on which motion sensors are attached, a surgical-grade NDI camera and a computer.

A sling is attached to the patient’s leg, the reflections returned by the motion sensors are then recorded by the infrared camera which transmits the information to the computer. The algorithms of the KneeKG system then analyze the data automatically and extract mechanical biomarkers known to be linked to the progression of specific diseases and symptoms related to the knee injury.

Marketed by Emovi, it is distributed in France, Germany, Italy, Poland, Spain and the United Kingdom by Macopharma.

The project “Development of a functional score to guide the diagnosis and help in the therapeutic recommendation of knee diseases”

This innovative project, which brings together Neila Mezghani, Khadidja Henni, Nicola Hagemeister and Emovi, draws on the experience of Neila Mezghani, who holds the Canadian chair of biomedical data analysis. The professor is interested in the analysis and classification of data in biomedical engineering and the development of tools based on artificial intelligence methods for the development of decision support systems.

The AI-assisted knee assessment tools, which the project team is working on, will provide faster and more accurate diagnoses for people with knee osteoarthritis. The team wants to develop quantified measurements (functional score) associated with various knee pathologies that can serve as a benchmark for the clinician. This score would also make it possible to classify the risk of disease progression more precisely and therefore to prioritize the follow-up to be carried out.

A grant of $ 471,901 was awarded to the science team. Most of this funding comes from the Quebec Ministry of Economy and Innovation (80%) and from private partners (20%).

Leave a Reply

Your email address will not be published. Required fields are marked *