Breast Imaging and Omics for Non-Invasive Integrated Classification (BIONIC)

Breast cancer is often diagnosed by invasive tissue biopsies, and treatment decisions are subsequently guided by molecular subtyping. Delays, procedural risks, and limited accessibility in resource-constrained contexts limit this method, notwithstanding its accuracy. In this paper we propose a novel non-invasive diagnostic approach for the molecular subtyping of breast tumors that integrates AI-enhanced Doppler ultrasound imaging with multi-omics data, including transcriptomics, proteomics, and genomes. By facilitating real-time, systemic evaluation of tumor biology, this technique offers a scalable and convenient alternative to tissue biopsy. Our results show how AI-integrated imaging and omics can be used to accurately stratify breast cancer subtypes, speed up diagnostic times, and enhance global equity in cancer care.