article · Scientific Reports
Osteoporosis is a chronic condition that reduces bone density and creates substantial health risks, especially among older populations. Standard diagnostic approaches can be slow and imprecise. To tackle this, a system called FuzzyBoneNet was developed to predict osteoporosis from X-ray imagery. The method prepares images using advanced enhancement tools, specifically bilateral improvement alongside top-hat and bottom-hat filtering, with quality confirmed via standard signal and similarity metrics. The workflow then applies transfer learning models, including AlexNet, VGG-19, and Xception, integrated with a fuzzy rank-based fusion technique to boost classification performance. Class imbalance is handled through oversampling. In testing, the combined approach reached an overall accuracy of 98.68% across normal, osteopenic, and osteoporotic conditions, outperforming existing leading methods and showing the promise of merging deep learning with fuzzy logic.
Osteoporosis weakens bones and leads to serious health risks, but standard screening can be imprecise and slow. By pairing deep learning models with fuzzy logic and enhanced X-ray processing, this approach provides a highly accurate method for identifying bone loss early. Such automated analysis could support clinicians in making faster, more dependable diagnoses before severe fractures occur.
This work demonstrates an applied, tested machine learning pipeline for diagnostic medical software. The primary users would be healthcare providers, radiologists, and developers of clinical decision-support systems seeking automated X-ray analysis tools. Because the framework has been validated computationally against existing models, moving towards real-world adoption would require further integration into clinical imaging workflows and formal regulatory validation.
AI-generated from the published abstract. Always read the original work before citing.
Osteoporosis is a chronic condition affecting the bones, resulting in decreased bone density. It poses significant health risks, particularly for the elderly. Conventional diagnostic methods frequently lack precision and are time-consuming. This article presents FuzzyBoneNet, an innovative approach for predicting osteoporosis with transfer learning and enhanced medical imaging techniques. To improve X-ray images, we propose utilizing advanced image enhancement techniques, including top-hat/bottom-hat filtering and bilateral image improvement. We employ a set of transfer learning models like AlexNet, VGG-19, and Xception that coupled with a fuzzy rank-based fusion technique to enhance classification accuracy. Oversampling resolves class imbalance, while quantitative criteria such as Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM) assess image quality. Research demonstrates that FuzzyBoneNet significantly outperforms existing leading approaches, accurately recognizing 98.68% of instances of normal, osteopenic, and osteoporotic bone conditions. The integration of deep learning with fuzzy logic may enhance the accuracy of osteoporosis detection, as demonstrated by this work.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1038/s41598-025-25946-w
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
