The Northwestern Medicine study introduces a novel digital biomarker, the Histomic Prognostic Signature (HiPS), using a new deep learning-based AI tool designed to improve breast cancer prognosis. Traditional methods, like the Nottingham criteria used by pathologists, focus on grading breast tissue based on its microscopic appearance but don’t consider noncancerous elements in the tumor microenvironment.
Histomic Prognostic Signature (HiPS) addresses this by providing a detailed scoring of survival risk based on the morphology of the breast tumor microenvironment. Using deep learning to map cellular and tissue structures accurately, it analyzes various features such as epithelial, stromal, immune, and spatial interactions within the tissue. The tool’s effectiveness is primarily attributed to its focus on stromal and immune features in the tumor microenvironment.
This comprehensive validation process showed that Histomic Prognostic Signature (HiPS) is more effective than traditional methods in predicting survival outcomes for breast cancer patients, regardless of the tumor–node–metastasis stage and other significant variables. Researchers’ plans include prospective evaluation of this model for clinical use and developing models for specific breast cancer types.
Ref: Amgad M, Hodge JM, Elsebaie MAT, et al. A population-level digital histologic biomarker for enhanced prognosis of invasive breast cancer. Nat Med. Published online November 27, 2023:1-13. doi:10.1038/s41591-023-02643-7
You must be logged in to post a comment.