Re-use of Clinical Imaging Information for Research: Structured Reporting, Machine Learning, and Natural Language Processing

April 27, 2016 11:00 a.m. - 12:00 p.m. ET

The imaging report is an essential source of clinical imaging information. It documents critical information about the patient's health and provides a professional interpretation of the images. However, the vast majority of report information remains narrative, a major obstacle to the rapid extraction and re-use of discrete imaging data. Structured reporting facilitates linking of imaging observations to clinical and genomic data, and is increasingly being adopted by clinical imaging practices. However, most imaging reports are used only once by the clinician who ordered the imaging study and are rarely used again for research, clinical care, or analytics. This presentation will describe the likely future of the imaging report, including efforts underway to standardize radiology report information, and the use of machine learning and natural language processing techniques to extract the semantic elements of the radiology report. These novel technologies enable connections between images and the electronic health record, and represent a vital part of the future of medical research.

Curtis Langlotz, M.D., Ph.D.

r. Langlotz has published over 100 scholarly articles, and is author of the recently published book "The Radiology Report: A Guide to Thoughtful Communication for Radiologists and Other Medical Professionals." Over the past decade, Dr. Langlotz has led many national and international efforts to improve the quality of radiology reporting, including the RadLex terminology standard, the RadLex Playbook of radiology exam codes, and the report template library of the Radiological Society of North America. His research is focused on improving the accuracy and consistency of radiology communication through real-time decision support systems and other information technologies. His biomedical informatics laboratory develops novel deep learning and natural language processing algorithms that provide intelligent assistance to radiologists, clinicians, patients, and other consumers of images and reports.

Was this page useful?