A new machine learning approach can predict the negative side effects of radiation treatment in patients with head and neck cancers. The findings, presented at the American Society for Radiation Oncology (ASTRO) annual meeting, can help select patients who might need a more tailored care approach.