Enhanced Prognostication of Liver Metastases Survival: A Fusion of Machine Learning Techniques

Secondary liver cancer survival prediction is an essential research area in oncology. However, insufficient research has been conducted using machine learning to support these predictions. In the field of healthcare, this remains an underexplored area. The ability to predict survival outcomes holds promise to benefit the healthcare industry and help tailor treatment decisions. Exploration into the role machine learning plays in predicting cancer survival is limited, with the extension of hybrid modelling even less so. The potential it has to offer through clinical decision-making and treatment strategies is critical to the future. This paper investigates the accuracy of hybrid machine learning classifiers in predicting the survival of secondary liver cancer, with research leveraging existing models’ strengths and weaknesses. The findings indicate that a weighted average hybrid model, combining Gradient Boosting Machines, Support Vector Machines and KNN, achieved an accuracy of 62.54% and an F1 score of 72.19%, the strongest patient survival predictor. The results demonstrate the importance of using machine learning algorithms and highlight the value of hybrid models for greater robustness and predictive accuracy. The implications of this research extend beyond predicting secondary liver cancer survival, advancing into other types of cancer survival prediction and a more comprehensive range of diseases.