Grant: Discovery Grants Program - Individual
École Polytechnique de Montréal
$53,000.00
Graph-based representations learning for an interpretable diagnosis and prognosis of ocular diseases
EF 2024-25
Résumé
The overall objective of this research program is to develop novel computational tools for interpretable diagnosis and prognosis of sight-threatening diseases from multi-view and multimodal images. To achieve this, we will leverage recent theoretical advancements in graphs and their classification to build a multimodal representation of the retina. The goal is to gain a better understanding of the mechanisms of classification and the evolution of retinal diseases, especially diabetic retinopathy (DR), age-related macular degeneration (AMD) stage conversion, and glaucoma pathogenesis.
The short-term objectives are: OBJ 1) Perform an automated diagnosis of ocular diseases using a graph-based representation of the retina from automatically segmented structures in the image. This involves studying both the graph's structure and the classification architecture that maximizes diagnostic performance; OBJ 2) Build and validate multi-view and multimodal graph learning architectures for an automatic diagnosis of multiple retinal diseases to leverage cross-modal dependencies; OBJ 3) Extend the model's applications to provide disease prognosis in cases where longitudinal monitoring is crucial: AMD, glaucoma, and DR.
The long-term objective is to promote and contribute to better incorporation of medical expertise in the construction of medical classification models, ultimately facilitating their deployment and adoption in clinical practice and leading to timelier patient care. Our goal is to obtain a graph-based representation that is sufficiently robust and flexible to monitor the patient's retinal health and to integrate new information as the patient undergoes consultations, including medical records.
The long-term vision is to contribute to a paradigm shift in healthcare by promoting preventive medicine instead of palliative care.
We believe that our research program will lead to major knowledge advancements in modeling physiological structures of the retina and to novel technology that addresses a significant healthcare need. Semantic graph-based representations will model relationships between multimodal biomarkers and enhance interpretability, fostering trust in AI-assisted medical decision-making. This will lead to powerful tools for early diagnosis of ocular diseases that are common causes of blindness in the working-age population. Indeed, DR, AMD and glaucoma affect millions of people worldwide and their incidence will increase in an ageing population. Thus, the proposed research will lead to significant cost savings to the Canadian healthcare system by preventing serious sight-threatening diseases. HQP involved in this program will work in an interdisciplinary and inclusive environment combining technology development at Polytechnique and state-of-the art imaging performed at Maisonneuve Rosemont Hospital. Developed algorithms will be disseminated publicly through open-source software projects.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Quebec, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2024-06559
Name nom
Cheriet, Farida
Department département
Génie informatique et génie logiciel
Organization id
64
Institution établissement
École Polytechnique de Montréal
Province en
Québec
Province fr
Québec
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
53000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1510
Committee name en
Electrical and Computer Engineering
Committee name fr
Génie électrique et informatique
Area of application code
1102
Area of application group en
Health, education and social services
Area of application group fr
Santé publique, éducation et services sociaux
Area of application en
Biomedical engineering
Area of application fr
Génie biomédical
Research subject code
2800
Research subject group en
Artificial intelligence
Research subject group fr
Intelligence artificielle
Research subject en
Artificial intelligence (computer vision, use2603)
Research subject fr
Intelligence artificielle (vision artificielle, utiliser 2603)
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Graph-based representations learning for an interpretable diagnosis and prognosis of ocular diseases
Keywords
medical imaging; artificial intelligence; deep learning; graph representations; multimodal imaging; spatiotemporal analysis; statistical modeling; semantic segmentation; graph neural networks; ocular diseases
Application summary
The overall objective of this research program is to develop novel computational tools for interpretable diagnosis and prognosis of sight-threatening diseases from multi-view and multimodal images. To achieve this, we will leverage recent theoretical advancements in graphs and their classification to build a multimodal representation of the retina. The goal is to gain a better understanding of the mechanisms of classification and the evolution of retinal diseases, especially diabetic retinopathy (DR), age-related macular degeneration (AMD) stage conversion, and glaucoma pathogenesis.
The short-term objectives are: OBJ 1) Perform an automated diagnosis of ocular diseases using a graph-based representation of the retina from automatically segmented structures in the image. This involves studying both the graph's structure and the classification architecture that maximizes diagnostic performance; OBJ 2) Build and validate multi-view and multimodal graph learning architectures for an automatic diagnosis of multiple retinal diseases to leverage cross-modal dependencies; OBJ 3) Extend the model's applications to provide disease prognosis in cases where longitudinal monitoring is crucial: AMD, glaucoma, and DR.
The long-term objective is to promote and contribute to better incorporation of medical expertise in the construction of medical classification models, ultimately facilitating their deployment and adoption in clinical practice and leading to timelier patient care. Our goal is to obtain a graph-based representation that is sufficiently robust and flexible to monitor the patient's retinal health and to integrate new information as the patient undergoes consultations, including medical records.
The long-term vision is to contribute to a paradigm shift in healthcare by promoting preventive medicine instead of palliative care.
We believe that our research program will lead to major knowledge advancements in modeling physiological structures of the retina and to novel technology that addresses a significant healthcare need. Semantic graph-based representations will model relationships between multimodal biomarkers and enhance interpretability, fostering trust in AI-assisted medical decision-making. This will lead to powerful tools for early diagnosis of ocular diseases that are common causes of blindness in the working-age population. Indeed, DR, AMD and glaucoma affect millions of people worldwide and their incidence will increase in an ageing population. Thus, the proposed research will lead to significant cost savings to the Canadian healthcare system by preventing serious sight-threatening diseases. HQP involved in this program will work in an interdisciplinary and inclusive environment combining technology development at Polytechnique and state-of-the art imaging performed at Maisonneuve Rosemont Hospital. Developed algorithms will be disseminated publicly through open-source software projects.