GateSPINE: Gated Cross-View Fusion for Lumbar Spine MRI Report Generation

[NEW PUBLICATION]

Congratulations to the group of authors from Applied AI Lab - Phenikaa University, Center for Diagnostic Imaging & Business Information - Phenikaa University Hospital and Business AI Lab - National Economics University for having a research article officially accepted for publication at the CSoNet 2026 International Conference!

• Proceedings / Publisher:Springer LNCS

• Conferences:CSoNet 2026 (International Conference on Computational Data and Social Networks)

• Article title:
  GateSPINE: Gated Cross-View Fusion for Lumbar Spine MRI Report Generation

• Main contributions:

• Pioneering GateSPINE architecture: Proposing an AI model for the problem of automatically generating lumbar spine Magnetic Resonance Imaging (MRI) diagnostic reports, contributing to optimizing the clinical diagnosis process.

• Gated Cross-View Fusion: Flexibly combines image features from many different pulse sequences and cross-sections through an information control gate, helping to capture complex lesions across the entire spine.

• Incorporating Medical Context Language Models: Integrating merged visual features into Large Language Models to automatically generate detailed, highly accurate, and professionally meaningful textual medical reports.