0130 Utilizing self-supervised machine learning to improve risk stratification in cutaneous squamous cell carcinoma
Staging systems for cutaneous squamous cell carcinoma (cSCC) often fail to reliably predict poor outcomes. Up to 40% of poor-outcome cases are classified within low-risk groups (BWH), while high-risk cSCC cases show a 46% upstaging rate1,2. We developed the rank-aware contextual reasoning (RACR) multiple instance learning (MIL) approach to predict metastasis risk and identify novel histological phenotypes3. Our approach combines self-supervised histological features with hierarchical graph networks to capture the tumor microenvironment and structural context.
Authors
- Nneka I. Comfere (ORCID: https://orcid.org/0000-0002-8001-2639)
- A.R. Mangold
- David J. DiCaudo (ORCID: https://orcid.org/0000-0001-8088-9353)
- Zachary Leibovit‐Reiben (ORCID: https://orcid.org/0000-0002-9229-1145)
- Olayemi Sokumbi (ORCID: https://orcid.org/0000-0001-8060-4311)
- A. Choudhary
- S. Nelson
- P. Arnold
- Lucas Swanson (ORCID: https://orcid.org/0000-0002-7074-4991)
- Andrew Hwang (ORCID: https://orcid.org/0000-0002-8437-6800)
- K. Saboo
- Javier Cañueto (ORCID: https://orcid.org/0000-0003-0037-7353)
- Jacob Kechter (ORCID: https://orcid.org/0000-0002-2646-0166)
- R. Iyer
- Eric Johnson
- Alyssa Stockard
Institutions
- Mayo Clinic (US)
- University of Illinois Urbana-Champaign (US)
- Illinois College (US)
- Jacksonville College (US)
- Complejo Hospitalario de Salamanca (ES)
- Mayo Clinic in Arizona (US)
- Mayo Clinic in Florida (US)
Publication Details
- Journal
- Journal of Investigative Dermatology
- Published
- 2025-07-21
- DOI
- https://doi.org/10.1016/j.jid.2025.06.132
- Primary Topic
- Cutaneous Melanoma Detection and Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00