Rice University’s MetaSeg matches U-Net accuracy using 90% fewer parameters, making AI medical image segmentation faster, cheaper, and more accessible for hospitals and researchers.