INESC TEC participates in the development of First Artificial Intelligence prototype in the field of Digital Pathology with application in colorectal cancer developed in Portugal

The Institute of Systems and Computer Engineering, Technology and Science (INESC TEC) and the pathological anatomy laboratory IMP Diagnostics developed a solution that allows optimising the diagnosis of colorectal pathologies, using Artificial Intelligence (AI). This solution was developed within the scope of the CADPath.AI project, according to the best and latest technology in terms of the anatomopathological diagnosis of histological samples.

AI shows significant potential to improve healthcare. The digitalisation of medical records allows doctors and patients to make faster and better decisions, and to seek immediate help from experts in diagnosing digital images around the world. Aresearch by Max Tegmark, a professor at the Massachusetts Institute of Technology (MIT), predicts that AI will play an increasingly important role in the diagnosis of clinical pathologies, thanks to the progress in terms of computer vision.

The model included in the prototype is based on an AI algorithm developed from the analysis of approximately 4500 colorectal biopsies and polypectomies, with a sensitivity of 98.9% and an accuracy rate of 90.2%. “With this tool, the pathologist will be able to analyse the digitalised image of a colorectal histological sample and obtain the classification of the sample in real time, either without neoplasia, low-grade or high-grade lesions”, explained Isabel Macedo Pinto, clinical director of IMP Diagnostics. “The system is also capable of generating a map representing the specific classification of the different parts of the tissue, guiding the pathologist to areas that are important to the respective diagnosis”, she stated.

Jaime Cardoso, researcher at INESC TEC, explained that “the prototype will also allow the pathologists to report if the automatic diagnosis is correct, enabling the integration of new data in the development of the algorithm, thus improving its performance over time”.

The testing phase of the prototype at IMP Diagnostics will begin this month to assess its day-to-day operation in the laboratory.

CADPath.AI is a Research and Technological Development project, funded by the Portugal 2020 programme.

The INESC TEC researcher mentioned in this news piece is associated with UP-FEUP.