Case Study: American Cancer Society discovers breast cancer patterns with Slalom

Finding breast cancer patterns with machine learning

Slalom partnered with the American Cancer Society to tackle the challenge of analyzing more than 1,700 breast cancer tissue slides, a process that would have taken a pathologist years to complete manually. Using Slalom and Google Cloud Machine Learning, ACS sought a scalable way to uncover patterns in tumor images that could support future breast cancer research and treatment.

Slalom converted the slide images into a usable format on Google Cloud, cleaned inconsistencies, and built an unsupervised machine learning model to cluster tissue patterns. The project delivered insights in about three months instead of an estimated three years, and left the American Cancer Society with reusable machine learning capabilities for future cancer research.

American Cancer Society

Mia Gaudet

Scientific Director of Epidemiology Research