Accelerating implementation of AI-based methods in clinical practice
The project “Accelerating precision treatment for colorectal cancer with deep learning and digital pathology” aims to facilitate faster and safer implementation of a validated AI-based method into clinical practice for the treatment of colorectal cancer.
Photo: Erling Sæthre-Hansen
The project has been awarded funding from the Norwegian Cancer Society through the call “Accelerating Methods into Practice”, which targets methods that are ready for implementation within healthcare services.
Among other activities, the project will integrate the CAPAI (Combined Analysis of Pathologists and Artificial Intelligence) system into Sectra’s pathology module. The CAPAI system includes the digital biomarker DoMore v1/Histotype Px® Colorectal, which analyses digital histological images using deep learning and provides prognostic information to support the decision of adjuvant chemotherapy after resection for early-stage colorectal cancer.
Among other activities, the project will integrate the CAPAI (Combined Analysis of Pathologists and Artificial Intelligence) system into Sectra’s pathology module. The CAPAI system includes the digital biomarker DoMore v1/Histotype Px® Colorectal, which analyses digital histological images using deep learning and provides prognostic information to support the decision of adjuvant chemotherapy after resection for early-stage colorectal cancer.
In addition, safety mechanisms will be established to alert users in cases of inadequate image quality, and health economic analyses will be conducted to assess the benefits and implications of clinical use.
The project builds on more than 20 years of research in digital pathology and artificial intelligence at the institute. From 2016 to 2021, the institute led the “Lighthouse” project DoMore!, funded by the Research Council of Norway, in which methods for automated analysis of cancer tissue using deep learning were developed and validated. As a result of the DoMore project, DoMore Diagnostics AS was established to commercialise the technology.
The project team consists of researchers and clinicians at Oslo University Hospital in collaboration with DoMore Diagnostics AS. User representative Solveig Rostøl Bakken and the patient organisation NORILCO contribute to ensuring the patient perspective in the implementation process.