Predictive Models Using Interpretable Machine Learning for Cardiovascular Survival in Breast Cancer Patients with Secondary Malignancies: Insights from a SEER Analysis - Summary - MDSpire
Advertisement
Predictive Models Using Interpretable Machine Learning for Cardiovascular Survival in Breast Cancer Patients with Secondary Malignancies: Insights from a SEER Analysis
To develop and compare predictive models for cardiovascular-specific survival (CSS) in breast cancer patients with secondary malignancies using machine learning techniques.
Key Findings:
Machine learning models (RSF, SVM, XGBoost, DeepSurv) outperformed traditional Cox regression in predicting cardiovascular-specific survival.
SHAP analysis provided insights into the contribution of various risk factors in the optimal model.
An interactive web-based prediction tool was developed for personalized risk assessment.
Interpretation:
The study demonstrates the potential of machine learning in enhancing predictive accuracy for cardiovascular outcomes in breast cancer patients with secondary malignancies, addressing a critical gap in existing research.
Limitations:
The study is based on retrospective data from a single database, which may limit generalizability.
Potential biases in data collection and patient selection may affect the results.
Conclusion:
The integration of machine learning techniques offers a promising approach to improve risk stratification and management of cardiovascular complications in breast cancer survivors with secondary malignancies.
This twice-monthly newsletter highlights recently published research where Dana-Farber faculty are listed as first or senior authors. The information is pulled from PubMed and this issue notes papers published from April 16 - 30.
