Principal Investigator Videos: Andreas Rauschecker on AI tools to improve MRI

At UCSF, Andreas Rauschecker, MD, PhD, is a leader in developing AI tools that help specialists diagnose and monitor neurological diseases. From brain tumors to multiple sclerosis to neurodegenerative diseases and even rare inherited childhood diseases, Rauschecker's work has created ways to better extract information from MRI scans. These AI tools can collaborate with the radiologist and provide them with additional insights.

One such project is an AI support program, the Adaptive Radiology Interpretation and Education System (ARIES), which helps trainees significantly improve the accuracy of their brain MRI diagnosis, particularly for rare diseases. This AI tool – which in some ways simulates years of diagnostic experience – showed promise in helping laypeople approach the accuracy of subspecialists in many areas.

In another project, Rauschecker developed new AI imaging methods to predict genetic biomarkers of glioblastoma from preoperative brain MRI, a potentially useful advance for providing individualized precision therapies for brain tumors.

Rauschecker quickly realizes that UCSF and especially the Center for Intelligent Imaging (ci2) were an optimal setting for his work. Leveraging UCSF's extensive database of medical records to create well-trained AI tools and image analysis programs enables sophisticated and confident predictions, all while adhering to HIPAA privacy standards.

This is the second in a series of videos highlighting our principal investigators and their groundbreaking work in the UCSF Department of Radiology and Biomedical Imaging. These researchers bring imaging insights to the most challenging clinical questions in many different disciplines and collaborate with experts from nearly every department at UCSF.