Prognostic prediction for lung stereotactic body radiotherapy by using CT-based radiomic features

Abstract

Personalized medicine is increasingly becoming a subject of intense interest, and studies on the utilization of data mining for implementing personalized medicine are on the rise. The increasing view that medical images are minable data on patients' personal traits have led to increased focus on quantitative analysis of these data, a field known as radiomics. Radiomics is a hybrid of the term "radiology" and the suffix "-omics", and aims to integrate and comprehensively analyze biological data. Researchers extract high-dimensional quantitative image information, known as the radiomic features, which are not detectable upon visual examination of regions of interest containing tumors. This information is subsequently linked to clinical, histopathological and molecular biological data and investigated regarding tumor phenotype, prognosis and treatment efficacy. The aim of this study is to predict prognosis for early-stage lung cancer patients after stereotactic body radiotherapy in multi-institution by CT-based radiomic features.