SKCMcrpred web-server can be used to distinguish high risk skin cutaneous melanoma (SKCM) patients from low risk SKCM patients based on the clinically measured (AJCC approved) features for a given patient viz. N stage, M stage, Breslow Thickness and Ulceration status. The risk estimation is based on an ensemble model that integrates all these features by employing statistical and survival analysis techniques, on a recent SKCM dataset. See supplement section for more information.
SKCM or Melanoma is one of the fatal forms of skin-cancers and requires immediate attention. New prognostic markers and strategies are needed for more accurate risk prediction. The model implemented here achieves this risk stratification with a minimal number of features and a greater accuracy than before. Following this risk estimation, better therapy choices can be implemented and survival chances of melanoma patients can be improved.
Users are provided with four input fields 1) N stage 2) M stage 3) Ulceration Status and 4) Breslow Thickness. The results displayed show the risk category, risk grade and life expectancy of the patient. This prediction is based on survival analysis using the prognostic marker function RG. See algorithm section for more details.