Historical and Clinical Contributions of the JSRT Chest Radiograph Database to Medical Imaging Research

The Japanese Society of Radiological Technology (JSRT) chest radiograph database (JSRT-DB), developed in 1998, has been widely used in medical imaging research for nearly three decades and has been cited more than 800 times. Although recent advances in artificial intelligence have shifted the focus toward large-scale datasets, the JSRT-DB has played an important role in establishing methodological standards for diagnostic performance evaluation, particularly in lung nodule detection. This review summarizes the historical background, design philosophy, and significant contributions of the JSRT-DB to observer performance studies, computer-aided diagnosis, and radiological education. It also discusses the lessons learned from the JSRT-DB that remain relevant to the development and utilization of large-scale medical imaging databases.