Development and internal validation of machine learning models for personalized survival predictions in spinal cord glioma patients
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.Machine learning models accurately predict survival for spinal cord gliomas (SCG) patients. An online calculator is available to help physicians assess individual patient prognosis.
Area Of Science
- Machine learning applications in oncology
- Predictive modeling for neuro-oncology
- Survival analysis in rare tumors
Background
- Spinal cord gliomas (SCG) survival is influenced by numerous factors.
- Accurate prognostic models are essential for SCG patient management.
- Identifying key determinants of SCG survival is crucial for personalized medicine.
Purpose Of The Study
- Develop interpretable machine learning (ML) models to predict survival outcomes in SCG patients.
- Create an accessible online calculator integrating these ML models for clinical use.
- Enhance clinical decision-making through precise individual survival predictions.
Main Methods
- Retrospective, population-based cohort study using the National Cancer Database (NCDB) (2010-2019).
- Utilized five ML algorithms (TabPFN, CatBoost, XGBoost, LightGBM, Random Forest) with hyperparameter optimization.
- Evaluated model performance using ROC curves, PRCs, AUROC, AUPRC, and Brier Score; employed SHAP and PDP for interpretability.
Main Results
- Top ML models achieved high AUROCs: 0.938 (1-year), 0.907 (3-year), and 0.902 (5-year).
- Key predictors identified include histology, tumor grade, age, surgery, radiotherapy, and tumor size.
- The study included thousands of SCG patients, with varying mortality rates at 1, 3, and 5 years postdiagnosis.
Conclusions
- ML models demonstrate high accuracy and discriminatory ability for SCG patient prognostication.
- An interactive online calculator is available for physician use, enhancing clinical applicability.
- This interpretable ML approach supports precision medicine and future research in SCG patient care.
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.