Gated fusion improves lumbar MRI report generation accuracy
GateSPINE: Gated Cross-View Fusion for Lumbar Spine MRI Report Generation
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Radiologists examine many types of MRI images to find problems in the spine, but it can be hard and time-consuming to write reports from all this data. The authors created GateSPINE, a computer program that looks at different MRI views together and decides which parts are most important when creating a report. This helps the program catch more abnormalities, making the generated reports more accurate and useful to doctors. They tested GateSPINE on multiple MRI datasets and found it improved detection without losing language quality in reports.
What this means in practice
- •For radiology software developers: Enhance lumbar spine MRI reporting tools by integrating gated cross-view fusion to improve detection of abnormalities from multiple imaging planes.
- •For medical imaging platform providers: Incorporate multi-sequence MRI fusion techniques to increase clinical accuracy of automated spine MRI report generation services.$Commercial implications: Enables more reliable automated spine MRI reports that can be sold to hospitals seeking AI-assisted diagnostic tools.
Authors
Hoang Nguyen Van, Cuong Vuong Tuan, Trang Mai Xuan, Bien Tran Van, Nam Tran Van, Thien Van Luong
Abstract
Automated report generation can ease the burden radiolo gists face when interpreting multi-sequence MRI studies. Unlike CT, MRI examinations comprise multiple sequences and imaging planes, each con tributing complementary diagnostic information. Existing methods en code a study as a single volume and combine multiple acquisitions by fixed rules. Findings visible in only one plane are thus diluted and of ten missed, lowering recall on clinical efficacy metrics, where a missed abnormality is most costly. We propose GateSPINE, a vision-language framework that fuses sagittal T1 and T2 volumes with a training-free operator, encodes the fused sagittal and axial volumes with two parallel 3D encoders, and decodes their combined representation into a report. Its core mechanism is a gated cross view fusion module that predicts, per feature channel and token, how much of each view to admit, so the more informative view dominates at each spatial location. We evaluate GateSPINE on three lumbar MRI datasets, comprising two public bench marks and a private cohort collected from Phenikaa University Hospital, using both natural language generation (NLG) and clinical efficacy (CE) metrics. GateSPINE achieves the highest CE F1 through improved re call on all three datasets; on SPIDER, which lacks an axial sequence, this reflects the sagittal fusion component rather than the gated cross-view mechanism, which is validated on the two cohorts with both imaging planes. GateSPINE also remains competitive on standard NLG metrics.