Paper Key : IRJ************262
Author: Gokulakrishnan V,Aadhikesavan M,Apsar Mohamed A,Asathullah S,Balaguhan S
Date Published: 13 Apr 2025
Abstract
Bone osteoarthritis (OA) is a degenerative joint disorder that significantly impairs mobility and quality of life by affecting the bones and cartilage within joints. Traditional diagnostic approaches, such as visual inspection of X-ray or MRI scans by radiologists, are often time-intensive, subjective, and prone to inconsistencies. These limitations highlight the pressing need for a fast, accurate, and objective diagnostic tool. To address this, the present study proposes a deep learning-based automated system utilizing the YOLO (You Only Look Once) algorithm to detect and classify osteoarthritic regions in radiographic images. The model leverages annotated datasets to precisely localize affected areas in real-time, offering enhanced diagnostic support. By integrating preprocessing techniques like image normalization, augmentation, and noise reduction, the system boosts accuracy and generalization across varied imaging conditions. The proposed model is evaluated using key performance metrics including mean Average Precision (mAP), Intersection over Union (IoU), and classification accuracy, ensuring a thorough assessment of detection and classification capabilities. Moreover, the incorporation of Grad-CAM (Gradient-weighted Class Activation Mapping) visualization allows for better interpretability by highlighting the specific regions influencing the model's predictions, aiding medical professionals in understanding the decision-making process. Experimental results demonstrate high detection accuracy and robust performance, positioning the system as a promising solution for early diagnosis and clinical integration. By reducing diagnostic time and minimizing human error, this approach holds significant potential to enhance osteoarthritis detection, improve patient outcomes, and streamline the clinical workflow.
DOI Requested