Application of Artificial Intelligence in Predicting Survival Outcomes in Multiple Myeloma Cancer Patients: A Systematic Review

Multiple Myeloma (MM) is a blood disorderliness or malignancy with significant clinical and heterogeneity making it difficult for making strong, reliable, and accurate prognosis. While the usual or conventional staging systems provide significant risk stratification, they are inadequate in the capturing of the complexities of Multiple Myeloma. Artificial Intelligence and Machine Learning tools and models have emerged as capable technologies for the analysis of high dimensional data to improve prognostic accuracy. The study utilized a systematic review approach to investigate the application of Artificial Intelligence and Machine Learning based technologies in the prediction of survival outcomes for Multiple Myeloma patients. PRISMA 2020 framework was adopted for the study. A systematic search of Google scholar for relevant studies was conducted between 2015 to July, 2025. Using relevant inclusion and exclusion criteria, and screening process, a total number of 10 research articles were considered to be very relevant to the study. The study result shows that artificial and machine based models such as Random Forest and Gradient Boosting algorithms demonstrate strong predictive power, with reported Area Under the Curve (AUC) and Concordance Index (C-index) values often ranging from 0.72 to 0.85. The models is capable of integrating different data types, including clinical parameters, laboratory results, high- dimensional genomics, and advanced imaging. The evidence suggests that AI/ML models can significantly enhance risk stratification, identify novel prognostic biomarkers, and offer more personalized survival predictions compared to conventional methods. Integrating AI into clinical practice holds the potential to optimize treatment strategies and improve outcomes for MM patients.