AI Driven Osteoporosis Diagnosis

Bol

This book presents an advanced artificial intelligence-driven framework for the accurate diagnosis of osteoporosis using Dual-Energy X-ray Absorptiometry (DEXA) images of the lumbar spine. Osteoporosis is a progressive skeletal disorder characterized by reduced bone mineral density and structural deterioration, leading to increased fracture risk and significant global health challenges. To overcome limitations of conventional diagnostic methods, an intelligent Osteoporosis Detection System (ODS) is proposed, integrating deep learning with metaheuristic optimization techniques.A novel real-world dataset, Medical Lumbar Spine Images (MLSI), collected from clinical practice in Mosul, Iraq, is utilized for training and evaluation. The dataset undergoes comprehensive preprocessing, including normalization, augmentation, and resizing, to enhance robustness and generalization. The core contribution lies in optimizing Convolutional Neural Network (CNN) hyperparameters using Genetic Algorithm (GA), Whale Optimization Algorithm (WOA), and Bobcat Optimization Algorithm (BOA), resulting in three optimized models: GOCNN, WOCNN, and BOCNN.Experimental results demonstrate high performance.

Amazon

Pagina's: 144, Paperback, KS OmniScriptum Publishing