Study on Precision Medicine and Survival Analysis for Melanoma in All Ages Using AI Techniques

What this Trial Is About

Researchers are creating a comprehensive database for melanoma that includes children, adolescents, and young adults along with adult patients. This database aims to analyze survival rates and evaluate sentinel lymph node positivity in younger patients. The study also focuses on adapting existing precision medicine algorithms for younger populations, using transfer learning to improve models with limited pediatric data, and integrating expert medical opinions into the decision support system. Artificial intelligence will be used to develop models that predict melanoma outcomes across all age groups. The study uses retrospectively collected data including demographic, genetic, medical procedures, cancer treatments, and biopsy information to build a detailed dataset. Advanced data engineering techniques will organize the data into a model-ready format that preserves patient history over time. The team will apply Gradient Boosting Survival Analysis, a specialized machine learning method for survival data, to handle censored information and estimate survival probabilities. Model performance will be evaluated using the concordance index, which measures how well predicted outcomes match actual events. Participants medical records from adults and younger patients will be included, with 6,000 adult melanoma cases and 120 cases from children, adolescents, and young adults. Researchers will analyze survival curves and sentinel lymph node positivity over a 24-month period. Data integrity is maintained through identifiers connecting patient events, and different transformations enrich the dataset. This observational study does not involve any direct treatment but focuses on data analysis to improve melanoma prognosis models.

Precision Medicine for L/GCMN and Melanoma 1