Machine Learning Methods for Detecting Radiation-Induced Tissue Inflammation in Patients with Lung Cancer

Abstract

Patients undergoes radiation therapy as measure of their lung cancer are at hazard of developing radiation pneumonitis, a condition caused by lung radiation damage (RP). RP is a theoretically terminal adverse influence of medicine. As a result, new strategies for guiding clinicians in administering personalized treatment doses to individuals at high risk of RP are necessary. Several prediction models have been constructed utilizing machine learning and traditional statistical processes, but no explanation for performance variances has been provided. In this study, we analyze a variety of well-known organization algorithms in the field of deep learning in order to identify several RP risk categories. The usefulness of these cataloguing algorithms is evaluated in combination with different segment assortment approaches, and the influence of technique collection on routine is then estimated further.