MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202641073659 A) filed by G Ashwin Prabhu; Dr. K. Ramachandra Reddy; Mr. R. Mothish Kumar; and Mr. R. J. Mugilan on June 13, 2026, for A Dual-Path Yolo26 Framework With Micro-Segmentation And Displacement-Aware Correlation For Automated Lumbar Vertebral Analysis And Early Spondylolisthesis Detection.
Inventors include Dr. K. Ramachandra Reddy; Mr. R. Mothish Kumar; and Mr. R. J. Mugilan.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: Spondylolisthesis is a common spinal disorder characterized by anterior or posterior displacement of a vertebral body, frequently affecting the lumbar region and causing lower back pain, radiculopathy, and functional disability. Early and accurate detection is essential for clinical decision-making, treatment planning, and preventing disease progression. This study proposes a Dual-Path YOLO26 framework integrated with micro-segmentation and displacement-aware correlation for automated lumbar vertebral analysis and early spondylolisthesis detection from radiographic images. The proposed architecture employs two complementary detection pathways: one focused on global vertebral localization and alignment assessment, and the other dedicated to fine-grained anatomical feature extraction. A micro- segmentation module is incorporated to delineate vertebral boundaries, intervertebral spaces, and endplate regions with high precision, enabling improved structural interpretation even in subtle displacement cases. To enhance diagnostic sensitivity, a displacement-aware correlation mechanism is introduced to measure spatial relationships between adjacent vertebrae and quantify abnormal translational shifts associated with early-stage spondylolisthesis. The framework is designed to address major challenges in lumbar spine analysis, including overlapping anatomical structures, low contrast, variable patient positioning, and minor vertebral misalignment that may be overlooked in conventional visual assessment. By combining object detection, localized segmentation, and correlation-based displacement estimation, the model provides both vertebral-level localization and clinically meaningful displacement indicators. The automated workflow reduces dependency on manual measurements, improves consistency, and supports radiologists and orthopedic specialists in rapid screening. Experimental evaluation is expected to demonstrate improved accuracy, sensitivity, and localization performance compared with conventional single-path detection models. The proposed Dual-Path YOLO26 framework offers a robust and scalable solution for computer-aided lumbar spine assessment. Its ability to identify early vertebral slippage may contribute to timely intervention, better patient monitoring, and enhanced diagnostic reliability in routine spinal imaging practice.
Disclaimer: Curated by HT Syndication.
