TY  - JOUR
T1  - Explainable Machine Learning for Predicting Skin Cancer Invasiveness: Insights into Regional Heterogeneity in Brazil
A1  - Maria Gonzales
A1  - Jose Santos
A1  - Anna Cruz
A1  - Carlos Reyes
JF  - Journal of Medical Sciences and Interdisciplinary Research
JO  - J Med Sci Interdiscip Res
SN  - 3108-4826
Y1  - 2025
VL  - 5
IS  - 2
DO  - 10.51847/5v9ZOjqCoE
SP  - 163
EP  - 176
N2  - Skin cancer constitutes the most frequent neoplasm in Brazil, and its potential for invasion is shaped by several elements, among which geographical factors play a notable role. Recognizing these elements is essential for advancing diagnostic accuracy and therapeutic effectiveness. This investigation examined how geographical region influences skin cancer invasiveness in Brazil by identifying distinct predictive patterns associated with different regions. Data obtained from the Hospital Cancer Registries (RHC) maintained by Brazil’s National Cancer Institute (INCA) underwent thorough analysis and preprocessing. Machine learning algorithms were then applied, and model interpretations were generated using the SHapley Additive exPlanations (SHAP) technique. The analysis indicated that geographical location is a key determinant in forecasting skin cancer invasiveness, underscoring the need to incorporate regional factors in future research. This study established that regional characteristics within Brazil significantly affect predictions of skin cancer invasiveness. While challenges such as data imbalance were encountered, the outcomes provide valuable insights for crafting more targeted policies to combat skin cancer across the country.
UR  - https://smerpub.com/article/explainable-machine-learning-for-predicting-skin-cancer-invasiveness-insights-into-regional-heterog-hwonptun7eyy80d
ER  -