Predictive Models Using Interpretable Machine Learning for Cardiovascular Survival in Breast Cancer Patients with Secondary Malignancies: Insights from a SEER Analysis - Report - MDSpire
Advertisement
Predictive Models Using Interpretable Machine Learning for Cardiovascular Survival in Breast Cancer Patients with Secondary Malignancies: Insights from a SEER Analysis
Predictive Models Using Interpretable Machine Learning for Cardiovascular Survival
Overview
This study developed and compared predictive models for cardiovascular survival in breast cancer patients with secondary malignancies using data from the SEER database. Machine learning algorithms demonstrated superior performance over traditional Cox regression models in assessing cardiovascular-specific survival.
Background
Breast cancer survivors face increased risks of secondary primary cancers and cardiovascular disease, which significantly impact long-term health outcomes. Traditional statistical methods struggle to accurately predict these risks due to the complexity of interactions between clinical characteristics and treatment-related toxicities. The integration of machine learning offers a promising approach to enhance risk stratification and improve clinical decision-making.
Data Highlights
No numerical data available.
Key Findings
Machine learning models outperformed Cox regression in predicting cardiovascular-specific survival in breast cancer patients with secondary malignancies.
Random survival forest (RSF), support vector machine (SVM), XGBoost, and DeepSurv were evaluated for their predictive capabilities.
SHAP analysis was utilized to interpret the contribution of various risk factors in the optimal model.
An interactive web-based prediction tool was developed to facilitate personalized risk assessment.
Cardiovascular disease mortality is a leading cause of non-cancer mortality among breast cancer survivors.
Clinical Implications
Healthcare providers should consider implementing machine learning-based predictive models to enhance risk assessment for cardiovascular complications in breast cancer survivors. The development of interactive tools can aid in personalized patient management and improve clinical outcomes.
Conclusion
The study highlights the potential of machine learning in improving cardiovascular survival predictions for breast cancer patients with secondary malignancies, emphasizing the need for integrated risk assessment strategies.
