AI-Powered Bone Fracture Detection and Advisory System: A Deep Learning Method Using GPT-Based Health Advice
DOI:
https://doi.org/10.70112/ajeat-2026.15.1.4342Keywords:
Bone Fractures, Healthcare, CNN, Orthopedic Diagnosis, Fracture Detection, Deep LearningAbstract
One of the most common musculoskeletal conditions that poses major challenges to healthcare systems across the globe is bone fractures. Traditional diagnostic methods are often prone to variability and human errors, although timely and accurate diagnosis is essential to avoid complications. The AI-Fracture Detection and Advisory System proposed in this research combines Convolutional Neural Networks (CNNs) for deep learning-based fracture detection with a Generative Pretrained Transformer (GPT)-based advisory module for health recommendations. CNNs utilize several convolutional layers activated by ReLU for scanning X-ray images. Pooling layers and fully connected layers with a Softmax function are used for classification. With 0.92 accuracy, 0.94 sensitivity, and 0.88 specificity, the model achieves 90% classification accuracy. The GPT-based advisory component provides recommendations based on fracture location, severity, and patient-specific information. It is built on a React.js-based front-end real-time interface, while image uploading and processing are handled by a scalable Flask/FastAPI backend. The system is trained on a balanced dataset of 10,580 X-ray images using preprocessing techniques such as pixel normalization, random rotation, flipping, brightness adjustment, and resizing to 180 × 180 pixels. By suggesting appropriate treatment options, such as surgery or immobilization based on fracture type, the system offers computer-aided decision support. Existing and future research aims to incorporate 3D imaging modalities and adopt a multiclass classification approach for detecting various types of fractures to further enhance the clinical applicability and robustness of AI-based diagnostic systems
References
[1] W. Hu and S. Razmjooy, “Combining an improved political optimizer with convolutional neural networks for accurate anterior cruciate ligament tear detection in sports injuries,” Scientific Reports, vol. 15, no. 1, Dec. 2025, doi: 10.1038/s41598-025-91242-2.
[2] Y. Zhou et al., “Rapid rib lesion diagnosis based on improved YOLOv8 and Byte track algorithms enhanced multi-planar reconstruction,” Neurocomputing, vol. 649, p. 130842, 2025, doi: 10.1016/j.neucom.2025.130842.
[3] L. Sun et al., “AI-assisted radiologists vs. standard double reading for rib fracture detection on CT images: A real-world clinical study,” PLoS One, vol. 20, no. 1, Jan. 2025, doi: 10.1371/journal.pone.0316732.
[4] Pacal and O. Attallah, “Hybrid deep learning model for automated colorectal cancer detection using local and global feature extraction,” Knowledge-Based Systems, vol. 319, p. 113625, 2025, doi: 10.1016/j.knosys.2025.113625.
[5] Y. Song, S. Ren, Y. Lu, X. Fu, and K. K. L. Wong, “Deep learning-based automatic segmentation of images in cardiac radiography: A promising challenge,” Computer Methods and Programs in Biomedicine, vol. 220, p. 106821, 2022, doi: 10.1016/j.cmpb.2022.106821.
[6] C. Pan, L. Lian, J. Chen, and R. Huang, “FemurTumorNet: Bone tumor classification in the proximal femur using DenseNet model based on radiographs,” Journal of Bone Oncology, vol. 42, p. 100504, 2023, doi: 10.1016/j.jbo.2023.100504.
[7] V. Hansen, J. Jensen, M. W. Kusk, O. Gerke, H. B. Tromborg, and S. Lysdahlgaard, “Deep learning performance compared to healthcare experts in detecting wrist fractures from radiographs: A systematic review and meta-analysis,” European Journal of Radiology, vol. 174, p. 111399, 2024, doi: 10.1016/j.ejrad.2024.111399.
[8] H. C. Ruitenbeek et al., “Cross-validation of an artificial intelligence tool for fracture classification and localization on conventional radiography in the Dutch population,” Insights into Imaging, vol. 16, no. 1, p. 150, 2025, doi: 10.1186/s13244-025-02034-1.
[9] C. Rainey, J. McConnell, C. Hughes, R. Bond, and S. McFadden, “Artificial intelligence for diagnosis of fractures on plain radiographs: A scoping review of current literature,” Intelligent-Based Medicine, vol. 5, p. 100033, 2021, doi: 10.1016/j.ibmed.2021.100033.
[10] S. Gernandt, R. Aymon, and P. Scolozzi, “Assessing the accuracy of artificial intelligence in the diagnosis and management of orbital fractures: Is this the future of surgical decision-making?,” JPRAS Open, vol. 42, pp. 275–283, 2024, doi: 10.1016/j.jpra.2024.09.014.
[11] M. Kutbi, “Artificial intelligence-based applications for bone fracture detection using medical images: A systematic review,” Diagnostics, Sep. 2024, doi: 10.3390/diagnostics14171879.
[12] M. A. Kassem, S. M. Naguib, H. M. Hamza, M. M. Fouda, M. K. Saleh, and K. M. Hosny, “Explainable transfer learning-based deep learning model for pelvis fracture detection,” International Journal of Intelligent Systems, vol. 2023, 2023, doi: 10.1155/2023/3281998.
[13] I. Matetić, I. Štajduhar, I. Wolf, and S. Ljubic, “A review of data-driven approaches and techniques for fault detection and diagnosis in HVAC systems,” Sensors, Jan. 2023, doi: 10.3390/s23010001.
[14] D. Wang et al., “‘Brilliant AI Doctor’ in rural clinics: Challenges in AI-powered clinical decision support system deployment,” in Proc. CHI Conf. Human Factors in Computing Systems (CHI ’21), New York, NY, USA: ACM, 2021, doi: 10.1145/3411764.3445432.
[15] M. Al Yaseen, H. Al Zahid, and S. Al-Haroon, “Amyloid deposits in the ligamentum flavum related to lumbar spinal canal stenosis and lumbar disc degeneration,” Cureus, vol. 14, no. 6, p. e26221, 2022, doi: 10.7759/cureus.26221.
[16] M. J. Leming et al., “Challenges of implementing computer-aided diagnostic models for neuroimages in a clinical setting,” npj Digital Medicine, Dec. 2023, doi: 10.1038/s41746-023-00868-x.
[17] B. Ozek, Z. Lu, F. Pouromran, S. Radhakrishnan, and S. Kamarthi, “Analysis of pain research literature through keyword co-occurrence networks,” PLOS Digital Health, vol. 2, no. 9, Sep. 2023, doi: 10.1371/journal.pdig.0000331.
[18] G. Litjens et al., “A survey on deep learning in medical image analysis,” Medical Image Analysis, vol. 42, pp. 60–88, 2017, doi: 10.1016/j.media.2017.07.005.
[19] P. Rajpurkar et al., “Deep learning for chest radiograph diagnosis: A retrospective comparison of the CheXNeXt algorithm to practicing radiologists,” PLoS Medicine, vol. 15, no. 11, Nov. 2018, doi: 10.1371/journal.pmed.1002686.
[20] A. Abd-alrazaq et al., “The performance of artificial intelligence-driven technologies in diagnosing mental disorders: An umbrella review,” npj Digital Medicine, Dec. 2022, doi: 10.1038/s41746-022-00631-8.
[21] Kaggle, “Bone fracture X-ray dataset.” [Online]. Available: https://www.kaggle.com
[22] M. I. Khan, J. Amin, M. A. Shehzad, and S. Iqbal, “An innovative hybrid model for elbow bone fracture detection: Integrating ViT and CNN,” 2025.
[23] J. P. Harper, G. R. Lee, I. Pan, X. V. Nguyen, N. Quails, and L. M. Prevedello, “External validation of a winning AI algorithm from the RSNA 2022 cervical spine fracture detection challenge,” American Journal of Neuroradiology, Feb. 2025, doi: 10.3174/ajnr.A8715.
[24] G. Vempati, T. Paladugu, J. Sama, R. T. Kamineni, and S. Kamepalli, “Evaluating CNN and deep learning models for bone fracture detection: A comparative study of VGG19, ResNet-50, LeNet, and DenseNet,” in Proc. Innovations in Communications, Electrical and Computer Engineering (ICICEC), Davangere, India: IEEE, Dec. 2024.
[25] J. Mao et al., “Automated detection and classification of mandibular fractures on multislice spiral computed tomography using modified convolutional neural networks,” Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology, Dec. 2024, doi: 10.1016/j.oooo.2024.07.010.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Centre for Research and Innovation
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
