Application of machine learning algorithms for predicting the outcome of therapeutic intervention of colorectal cancer patients in Palestine with a global comparison

This is the authors' abstract. We don't add key points for this paper.

Colorectal cancer (CRC) represents a significant global health challenge, being one of the leading causes of cancer-related deaths worldwide. Advances in predictive modeling are crucial for improving treatment outcomes. The objective of this study is to assess the efficacy of machine learning algorithms in predicting the outcomes of therapeutic interventions for colorectal cancer patients, comparing results from Palestinian patient data with global benchmarks. This study analyzes data sourced from the Ministry of Health in Palestine and the Cancer Data Access System (CDAS), providing a diverse clinical setting. We applied KNeighbors Classifier, Random Forest Classifier and XGBoost models to predict treatment success, recurrence, and survival rates. These models were evaluated using accuracy metrics and area under the receiver operating characteristic (ROC) curve. The results demonstrate that machine learning models can significantly enhance prediction accuracy for treatment outcomes in CRC. Factors such as age and stage of cancer were identified as critical predictors of patient outcomes. Machine learning offers substantial potential to improve the precision of therapeutic interventions in colorectal cancer by enabling personalized treatment plans. However, integration of real-time data and further validation of models are necessary to optimize clinical applications.