PROTECT YOUR DNA WITH QUANTUM TECHNOLOGY
Orgo-Life the new way to the future Advertising by AdpathwayA hybrid deep learning framework combining specialized vertebrae localization and dual-network image classification demonstrated improved performance over conventional methods in identifying early-stage osteoporosis from spine Dual-Energy X-ray Absorptiometry (DEXA) scans, according to a study published in the International Journal of Computational Intelligence Systems.
The research evaluated real-time spine DEXA images obtained from Chennai Hospital to address longstanding challenges in early diagnosis. While DEXA scans serve as the standard method for calculating bone mineral density, conventional evaluation techniques frequently lack sufficient precision and computational efficiency, making it difficult for clinicians to reliably recognize initial signs of bone loss such as osteopenia.
In the first stage of the automated workflow, the authors introduced the You Only Look Once version 9-based Lumbar Vertebrae Detection (YOLOv9-LVD) model. This stage isolates and accurately localizes target anatomical segments within the spine imaging, explicitly identifying the L1, L2, L3, and L4 individual lumbar vertebrae regions for subsequent diagnostic evaluation.
Once localized, the isolated lumbar vertebrae regions undergo automated diagnostic assessment using the Hybrid Model for Lumbar Vertebrae Classification (HMLVC). This architecture combines Deep Convolutional Neural Network (DCNNet) and Deep DenseNet (DDNet) models to categorize each detected vertebral segment into either a normal or abnormal classification.
For vertebral segments flagged as abnormal, the framework applies detailed statistical feature analysis to differentiate between osteopenia and osteoporosis. The researchers evaluated specific statistical metrics extracted from the abnormal lumbar regions, including mean values, the coefficient of variation, entropy measurements, and skewness indicators.
The study reported that the integrated YOLOv9-LVD and HMLVC pipeline achieved high accuracy alongside enhanced sensitivity compared to traditional assessment methods on the Chennai Hospital clinical dataset. Because osteoporosis is characterized by reduced bone density and elevated fracture risks that heavily affect older adults and postmenopausal women, the authors concluded that this hybrid model represents a promising approach to support clinical diagnosis and patient management strategies.
Subject of Research: Automated detection and classification of early-stage osteoporosis in lumbar vertebrae using hybrid deep learning on DEXA scans
Article Title: Enhanced Osteoporosis Detection in Lumbar Vertebrae Using Hybrid Deep Learning Models on DEXA Images
Article References: Umamaheswari, R., & Lakshmi, D. (2026). Enhanced Osteoporosis Detection in Lumbar Vertebrae Using Hybrid Deep Learning Models on DEXA Images. International Journal of Computational Intelligence Systems. https://doi.org/10.1007/s44196-026-01667-4
Image Credits: AI Generated
DOI: 10.1007/s44196-026-01667-4
Keywords: osteoporosis, osteopenia, lumbar vertebrae, DEXA images, dual-energy X-ray absorptiometry, deep learning, YOLOv9-LVD, HMLVC, DCNNet, Deep DenseNet, bone density, fracture risk, postmenopausal women, Chennai Hospital, entropy, coefficient of variation, skewness


2 days ago
28




















English (US) ·
French (CA) ·