AI Insight
This study evaluated SynthSeg, a deep learning brain segmentation tool, on a large dataset of 26,000 MRI scans spanning infancy through adulthood and found significant performance degradation during early development, with only 36% of infant scans passing automated quality control. Researchers discovered that rescaling infant brain scans to adult brain sizes and cropping them to match adult fields of view dramatically improved segmentation accuracy, increasing the success rate for infant scans from 36% to 91%. These preprocessing steps are necessary because current AI segmentation tools were not optimized for the rapid changes in brain size, morphology, and imaging contrast that occur during early human development.
Why it matters
This work provides a practical solution for researchers studying brain development and neurodevelopmental disorders, enabling more accurate automated analysis of pediatric brain MRI scans. The findings also highlight an important consideration for training future AI medical imaging tools: algorithms must account for developmental variations in organ size and structure to maintain performance across age ranges.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
The human brain undergoes rapid developmental changes through early life, underpinning the emergence of function but also marking a period of vulnerability to a range of neurodevelopmental disorders. With dynamic changes to brain size, morphology, and imaging contrast, consistent and accurate computational neuroanatomy remains a challenge. Deep learning tools for segmentation, like SynthSeg, offer robustness to heterogeneously acquired MRI contrast but remain unproven in early development. Here, we aggregated a large cohort (26k) of MRI scans spanning infant to adult development, and evaluated SynthSeg performance. Automated quality control scores, visual inspection, and spatial overlap with expert-segmented MRI scans revealed poor quality output segmentations during development. In the infant period only 36% of scans (1094/3069) passed automated QC. Rescaling infant scans to adult brain sizes significantly improved spatial overlap, and cropping scans to match adult fields of view retrieved automated quality control. Evaluation of the SynthSeg rescale + crop pipeline demonstrated visible and quantitative improvements in segmentation throughout infancy and childhood. There were marked increases in successful segmentations in infant scans, with 91% of scans now passing QC (2803/3069). These findings facilitate computational analysis of typical and disrupted neurodevelopment and should be considered when training the next generation of computational tools.