Open Sharing Icons

Multi-omics data to AI disease signature discovery

DHIF 1630

Multi-omics data to AI disease signature discovery

Using AI to predict which cancer treatments will work best for patients with lung, breast, or colorectal cancer by analyzing their genes, proteins, and metabolism together.

This project uses artificial intelligence (AI) to predict which cancer treatments are most likely to work for individual patients with lung, breast, or colorectal cancer. We analyze blood samples to identify biological patterns that help guide more personalized treatment decisions.

Cancer treatments do not work the same way for everyone. By combining multiple types of biological data and AI, this project aims to reduce trial and error in cancer care, improve treatment success, and accelerate precision medicine research in Canada.

By the end of the project, we expect to identify blood-based biomarkers and validate AI models that can help predict treatment response, while establishing a secure and standardized data-sharing framework across hospitals.

Key details

● Disease area: Oncology (lung, breast, colorectal cancer)

● Data used: Blood-based proteomics and metabolomics, linked to clinical outcomes

● Data security: Fully anonymized data; AI models trained on anonymized data

● Method: AI and machine learning are used to detect patterns across large datasets; data are standardized so results are comparable across institutions

● Impact: More precise treatment selection, fewer ineffective therapies, and better outcomes for cancer patients

"We see that many patients receive treatments that don’t work well for them. Our team wants to change that by using better data and technology to support more personalized and effective care for patients."

 

Project Images

↑