Evaluation of AI-based Direct Contrast Synthesis in the Pediatric Brain

Motivation: Direct contrast synthesis (AI-DCS) enables the derivation of synthetic images from a trained neural network. Its utility in clinical practice has yet to be adequately explored. Goal(s): To evaluate the image quality and diagnostic confidence of these synthetic images in the brain in the pediatric population. Approach: The study radiologist compares the image quality and diagnostic confidence of AI-DCS images to standard of care (SOC) and SyMRI. Results: AI-DCS demonstrates better overall image quality and diagnostic confidence than SyMRI on T1, T2 and FLAIR in majority of patients. AI-DCS demonstrates better overall image quality on FLAIR when compared to SOC. Impact: AI-DCS demonstrates improved image quality and diagnostic confidence compared to SyMRI and comparable image quality to the current standard of care.