Owkin

OWKIN MSIntuit (Colorectal Cancer)

Colorectal cancer (CRC) is the second most common cause of cancer-related deaths worldwide. A key factor in determining the best treatment for CRC is a genomic biomarker called microsatellite instability (MSI). Currently, identifying MSI and other important genetic markers requires specialised laboratory testing. Owkin, a company specialising in artificial intelligence (AI) for healthcare, is developing machine learning (ML) tools to help analyse routine tissue samples. They aim to create a tool that can predict genetic and clinical data from standard pathology slides, reducing the need for extensive genetic testing.

To achieve this, Owkin is working with University Hospitals Birmingham (UHB), which holds valuable clinical data and pathology slides from past CRC patients. This project will digitise these slides and pair them with existing genetic data, including mismatch repair (MMR) protein expression, BRAF gene mutations, and MLH1 gene methylation status.

The goal is to train and test AI models that can detect key cancer characteristics from a simple tissue slide, potentially speeding up diagnosis and improving treatment decisions. Additionally, this project will help establish a framework for future AI research in digital pathology, benefiting cancer patients and advancing medical science.

Owkin has developed the MSIntuit v2 pre-screening solution to assist pathologists in identifying microsatellite instability (MSI) tumours in patients with primary colorectal cancer. This innovative tool, powered by a deep learning algorithm, classifies patients based on routine haematoxylin and eosin (HE)-stained histology slides, representing a new category of AI-driven diagnostic technology within the emerging digital pathology ecosystem.

As part of its development, Owkin seeks to evaluate the performance of MSIntuit v2 on an external cohort of patients managed by the pathology laboratory at University Hospitals Birmingham (UHB). By leveraging real-world clinical data, this study aims to validate and enhance the tool’s accuracy, supporting the advancement of digital pathology and AI in cancer diagnostics. Ultimately, this project has the potential to improve early detection and treatment decision-making for colorectal cancer, benefiting both patients and healthcare providers.

04/06/2023

SDE022a

Data Access Provided

Owkin