Construction and Multi-Center Validation of the Radiomics Model for Non-Invasive Identification of Active Surveillance Candidates
Liang Wang, MD, PhD, presents current data on the use of noninvasive approaches with radiomics models to identify prostate cancer in active surveillance (AS) patients. Dr. Wang begins by sharing data on the risk reduction that early detection provides, but notes risks of overdiagnosis and overtreatment. He then addresses the role of magnetic resonance imaging (MRI) in prostate cancer management, noting improved techniques and better image interpretation by the Prostate Imaging Reporting & Data System (PI-RADS). However, Dr. Wang highlights that other biomarkers along with MRI must guide further diagnosis and treatment.
Dr. Wang discusses the rapidly evolving field of radiomics, explaining it enables the digital decoding of images into quantitative features that may uncover disease characteristics unseen by the naked eye. Further, it assesses a broad set of predefined features to define patterns relevant to pathology using statistical methods.
Dr. Wang concludes by cautioning that current data on the use of radiomics were from single-institution retrospectives with small cohort sizes and an absence of independent, external validation. Dr. Wang mentions broader, ongoing research which may lead to a non-invasive, radiomics-based tool that may be used to identify AS candidates with prostate cancer in the future.
Read More