Gagandeep Singh Selected for the 2025 CERRAF Award
Gagandeep Singh, MD, MBA, has been selected by the Association of Academic Radiology (AAR) for the 2025 Clinical Effectiveness in Radiology Research Academic Fellowship (CERRAF) Award. The annual award supports junior radiology faculty who are committed to developing their research careers, with a focus on projects that encompass health services research, care delivery, and technology assessment.
Singh, an assistant professor of radiology at Columbia’s Vagelos College of Physicians and Surgeons, will receive $140,000 over two years for his project, “Adaptable Tumor Segmentation and Radiogenomic Classification Model to Handle Real-World Challenges.” CERRAF fellows also receive structured mentoring and dedicated academic time to complete training and research.
Singh’s project aims to validate and implement previously developed artificial intelligence (AI) models for comprehensive brain tumor characterization. The performance of current state-of-the-art segmentation models is often contingent upon the availability of all MRI sequences that are used during training. This poses challenges in clinical settings, where datasets frequently lack some sequences due to artifacts, acquisition variability, or institutional differences.
“With the rapid advancement and integration of AI tools into clinical practice, a key challenge remains creating models that are adaptable, generalizable, and capable of functioning in diverse clinical settings,” says Singh. “By improving generalizability of diagnostic AI models, the proposed scalable framework will enhance AI integration into diverse clinical environments, ultimately improving patient care and clinical outcomes.”
Accurately segmenting brain tumors is a vital aspect of both diagnosing and treating brain tumor patients and AI-algorithms areincreasingly being used to automate the process of segmenting tumors from medical images. This adaptable AI framework is poised to pave the way for broader clinical adoption of diagnostic AI models that rely on accurate segmentation and subtype classification by addressing data heterogeneity across institutions.
Singh will be mentored by Ruth Carlos, MD, MS, and Dorothy Sippo, MD, MPH, from the Department of Radiology, and Jeff Goldsmith, PhD, from Columbia University’s Mailman School of Public Health.
Singh will be recognized as a CERRAF Fellow during the awards presentation at the American Association of Radiology Annual Meeting in Los Angeles on March 13, 2025. He is the first member of Columbia’s faculty to receive the award.