Neural network and spline-based regression for the prediction of salivary hypofunction in patients undergoing radiation therapy

Abstract

Background This study leverages a large retrospective cohort of head and neck cancer patients in order to develop machine learning models to predict radiation induced hyposalivation from dose-volume histograms of the parotid glands. Methods The pre and post-radiotherapy salivary flow rates of 510 head and neck cancer patients were used to fit three predictive models of salivary hypofunction, (1) the Lyman-Kutcher-Burman (LKB) model, (2) a spline-based model, (3) a neural network. A fourth LKB-type model using literature reported parameter values was included for reference. Predictive performance was evaluated using a cut-off dependent AUC analysis. Results The neural network model dominated the LKB models demonstrating better predictive performance at every cutoff with AUCs ranging from 0.75 to 0.83 depending on the cutoff selected. The spline-based model nearly dominated the LKB models with the fitted LKB model only performing better at the 0.55 cutoff. The AUCs for the spline model ranged from 0.75 to 0.84 depending on the cutoff chosen. The LKB models had the lowest predictive ability with AUCs ranging from 0.70 to 0.80 (fitted) and 0.67 to 0.77 (literature reported). Conclusion Our neural network model showed improved performance over the LKB and alternative machine learning approaches and provided clinically useful predictions of salivary hypofunction without relying on summary measures.

Development of a risk prediction model for radiation dermatitis following proton radiotherapy in head and neck cancer using ensemble machine learning

“…Although Logistic Regression offers greater interpretability, machine learning models, particularly XGBoost, excel in predictive accuracy, aligning with the growing emphasis on machine learning for complication prediction in medical research. Studies like Poolakkad et al [ 15 ], who achieved a higher AUC using gradient boosting for predicting mucositis post-chemotherapy, and Smith et al [ 16 ], who demonstrated superior performance using neural networks for post-radiation xerostomia, support this trend. Dean et al’s use of penalized logistic regression, SVM, and random forest to predict dysphagia with strong external validation further validates the efficacy of advanced models [16].…”