businesses
Project of $395,255 over 2 years
- Supported by CQDM through:
Ministère de l’Économie, de l’Innovation et de l’Énergie du Québec (MEIE) - And by co-funding partners:
Vega BioImagerie, Aurora mScope - And by project partner:
IVADO
The project in detail
The diagnosis of pancreatic neuroendocrine tumours involves classifying them by stage—G1 to G3—with G3 representing the most active and dangerous tumours. To determine the stage, pathologists assess the tumour’s proliferation rate using two criteria: the number of mitoses and the expression level of the Ki-67 protein. Mitosis counting is currently done manually by the pathologist, which is time-consuming and leads to significant variability in diagnosis. Although artificial intelligence has been studied for mitosis counting in breast cancer biopsies, there are very few data banks for pancreatic neuroendocrine tumours, which limits the implementation of such algorithms for this type of cancer.
Montreal-based companies Vega BioImagerie and Aurora mScope, in collaboration with the Centre Hospitalier de l’Université de Montréal, are developing a database of digitized pathology samples to create an automated algorithm for tumour grading. This project could significantly improve pathologists’ work by speeding up diagnosis and reducing errors. It also has strong commercial potential, with economic benefits for the Montreal-based companies involved.
