Machine Learning Methods Uncover Radiomorphologic Dose Patterns in Salivary Glands that Predict Xerostomia in Patients with Head and Neck Cancer

Abstract

Purpose Patients with head-and-neck cancer (HNC) may experience xerostomia after radiation therapy (RT), which leads to compromised quality of life. The purpose of this study is to explore how the spatial pattern of radiation dose (radiomorphology) in the major salivary glands influences xerostomia in patients with HNC. Methods and materials A data-driven approach using spatially explicit dosimetric predictors, voxel dose (ie, actual radiation dose in voxels in parotid glands [PG] and submandibular glands [SMG]) was used to predict whether patients would develop xerostomia 3 months after RT. Using planned radiation dose data and other nondose covariates including baseline xerostomia grade of 427 patients with HNC in our database, the machine learning methods were used to investigate the influence of dose patterns across subvolumes in PG and SMG on xerostomia. Results Of the 3 supervised learning methods studied, ridge logistic regression yielded the best predictive performance. Ridge logistic regression was also preferred to evaluate the influence pattern of highly correlated dose on xerostomia, which showed a discriminative pattern of influence of doses in the PG and SMG on xerostomia. Moreover, the superior–anterior portion of the contralateral PG and medial portion of the ipsilateral PG were determined to be the most influential regions regarding dose effect on xerostomia. The area under the receiver operating characteristic curve from a 10-fold cross-validation was 0.70 ± 0.04. Conclusions Radiomorphology, combined with machine learning methods, is able to suggest patterns of dose in PG and SMG that are the most influential on xerostomia. The influence pattern identified by this data-driven approach and machine learning methods may help improve RT treatment planning and reduce xerostomia after treatment.

Application of Radiomics for the Prediction of Radiation-Induced Toxicity in the IMRT Era: Current State-of-the-Art

“…In particular, parotid glands with low metabolic activity and a low fat-to-functional parenchymal ratio were matched by more heterogeneous intensity and texture imaging features: overall, these hypothesis-generating studies showed that pre-treatment radiomics-based prediction outperformed conventional NTCP models. Finally, a machine-learning approach integrating dosiomics, radiomics, and morphological data in predicting both acute and late injury to salivary glands has recently shown promising results ( 44 , 45 ). Interestingly, by applying a novel artificial intelligence methodology (“likelihood-fuzzy analysis”), Pota et al ( 46 ) identified quantitative predictors of 12-month toxicity through a longitudinal assessment of parotid glands in a dual institution experience.…”

Machine Learning-Based Models for Prediction of Toxicity Outcomes in Radiotherapy

“…Jiang et al (21) utilized a data set of 427 H&N cancer patients treated with RT to predict xerostomia. Ridge LR, LASSO LR, and RF classifiers were trained with planned radiation dose data and non-dosimetric features to investigate the influence of dose patterns on xerostomia.…”

Emerging Concepts and Novel Strategies in Radiation Therapy for Laryngeal Cancer Management

“…The ridge logistic regression method was used to evaluate the influence patterns of doses in the salivary glands on xerostomia. It was found that the superior–anterior portion of the contralateral parotid gland and the medial portion of the ipsilateral parotid gland were the most influential areas regarding dose effect on xerostomia [ 76 ]. Furthermore, the spatial radiation dose influence on the recovery of xerostomia eighteen months after treatment was also analyzed.…”