Do deep embedding-derived imaging phenotypes reveal breast cancer risk heterogeneity beyond scalar risk stratification?
Published in Proceedings of the 18th International Workshop on Breast Imaging (IWBI 2026), Thessaloniki, Greece, 2026
First author. Peer-reviewed proceedings paper, pp. 423-432. Presented as a poster at IWBI 2026, Thessaloniki, 28 June - 1 July 2026.
My contribution: Conceived and led the study, designed and ran the clustering and stability analyses, constructed the HIPIMR analytic cohort, and wrote the manuscript.
A negative result that directly motivates my dissertation work: clustering was highly stable (ARI > 0.98), but the recovered groups tracked predicted risk rather than distinct clinical phenotypes, which constrains what Mirai’s learned representation can be said to encode.
Recommended citation: Sollis LJ, Shvetsov Y, Quon B, Wolfgruber TK, Bunnell A, Hernandez BY, Shepherd JA. (2026). "Do deep embedding-derived imaging phenotypes reveal breast cancer risk heterogeneity beyond scalar risk stratification?" In: Sechopoulos I, ed. Proceedings of the 18th International Workshop on Breast Imaging (IWBI 2026), Thessaloniki, Greece. Zenodo. pp. 423-432.
View Paper