HERON: High-Efficiency Real-Time mOtion quantification and re-acquisitioN for Fetal dMRI

Motivation: Fetal diffusion MRI (dMRI) is essential for studying brain development but is highly sensitive to motion artifacts from fetal movement and maternal breathing, impacting data quality and limiting diagnostic accuracy. Goal(s): To create a robust, real-time automatic pipeline (HERON) that enhances fetal brain dMRI by performing motion assessment and re-acquisition. Approach: A pipeline using AI-driven brain localization, segmentation, and motion assessment on a clinical 0.55T scanner to autoplan, quality-check, and reacquire motion-affected dMRI volumes. Results: Tested in 6 cases, the pipeline effectively improved image quality, reducing inter-volume motion and enabling more reliable ADC analysis even in challenging cases. Impact: Enhancing fetal diffusion MRI through an automatic AI-driven pipeline enabling automatic planning, real-time quality checks and prospective re-acquisition of corrupted volumes enables higher diagnostic quality and wider use.