Motivation: Fetal fMRI offers critical insights into early brain functional development, but unpredictable fetal motion often distorts images, reducing data reliability and limiting research potential. Goal(s): This study aims to develop a prospective motion-correction (PMC) system for fetal fMRI to mitigate motion artifacts, enhancing data quality and enabling accurate study of fetal brain development. Approach: The PMC system integrates U-Net-based segmentation and rigid registration to track fetal head motion and adjust slice positioning in real-time. Results: PMC improved imaging quality with a 23% increase in temporal SNR and a 22% increase in Dice similarity index in an fMRI time series compared to uncorrected data. Impact: Real-time fetal brain segmentation and registration with one-TR latency enable effective motion-correction in fetal fMRI, allowing motion data from one repetition to guide adjustments in subsequent frames. This significantly enhances data reliability and usability in fetal fMRI studies.
Fan et al. (Tue,) studied this question.