Diego Mendez

M.D.

Hospital El Pino

Dermatology Resident

PLCO (Learn more about this study)

PLCO-1855

Mar 18, 2025

Development and Validation of a Machine Learning Model for Predicting Sentinel Lymph Node Metastasis in Melanoma Patients

Melanoma is a highly aggressive skin cancer, and sentinel lymph node biopsy (SLNB) remains a crucial step in staging and treatment decisions. However, unnecessary SLNBs may expose patients to surgical risks without clear benefits. This project aims to develop and validate a machine learning-based predictive model that estimates the risk of sentinel lymph node metastasis using clinical and histopathological variables such as Breslow thickness, ulceration, mitotic index, age, sex, regression, and lymphovascular invasion.

By leveraging data from the CDAS repository, we will train and validate our model on real-world patient data to improve its predictive performance. The ultimate goal is to provide clinicians with a decision-support tool that can optimize SLNB recommendations, reducing unnecessary procedures while ensuring accurate detection of metastatic disease. This model will be compared against existing clinical criteria to assess its performance and reliability.

To develop a machine learning model that predicts sentinel lymph node metastasis in melanoma patients based on key clinicopathological factors.

To validate the model using real-world patient data from the CDAS repository.

To compare the predictive performance of the machine learning model against standard clinical decision criteria.

To assess the potential impact of the model on reducing unnecessary SLNBs while maintaining high sensitivity in detecting metastatic cases.

To provide an interpretable and clinically relevant risk stratification tool for melanoma management.