Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
A GAN-based image translation technique was used to synthesize MRI-like images from low-dose CT scans, which were then combined with real MRI data to train a U-Net for automated segmentation of the thoracolumbar spine (T1-L5). This approach enables accurate, continuous 3D reconstructions from MRI alone, overcoming the limitations of manual workflows and data scarcity. The resulting model achieves an 88% Dice score and reduces processing time to under one minute, facilitating radiation-free deformity assessment in pediatric scoliosis.
Ditch the CT scans: AI now enables high-resolution, 3D spinal reconstructions from MRI alone, opening up radiation-free assessments for pediatric scoliosis.
MRI is preferred over CT in paediatric imaging because it avoids ionising radiation, but its use in spine deformity assessment is largely limited by the lack of automated, high-resolution 3D bony reconstruction, which continues to rely on CT. MRI-based 3D reconstruction remains impractical due to manual workflows and the scarcity of labelled full-spine datasets. This study introduces an AI framework that enables fully automated thoracolumbar spine (T1-L5) segmentation and 3D reconstruction from MRI alone. Historical low-dose CT scans from adolescent idiopathic scoliosis (AIS) patients were converted into MRI-like images using a GAN and combined with existing labelled thoracic MRI data to train a U-Net-based model. The resulting algorithm accurately generated continuous thoracolumbar 3D reconstructions, improved segmentation accuracy (88% Dice score), and reduced processing time from approximately 1 hour to under one minute, while preserving AIS-specific deformity features. This approach enables radiation-free 3D deformity assessment from MRI, supporting clinical evaluation, surgical planning, and navigation in paediatric spine care.