Related Experiment Video
Updated: Sep 16, 2025
06:23
Diagnosis and Surgical Treatment of Human Brucellar Spondylodiscitis
Published on: May 23, 2021
4.9K
Development of a deep learning-based MRI diagnostic model for human Brucella spondylitis.
Binyang Wang1, Jinquan Wei2, Zhijun Wang3
1Ningxia Institute of Clinical Medicine, The Third Clinical Medicine College, People's Hospital of Ningxia Hui Autonomous Region, Ningxia Medical University, Zhengyuan Street 301, Yinchuan, 750002, China.
Biomedical Engineering Online
|July 9, 2025
Summary
Deep learning models can accurately differentiate Brucella spondylitis (BS) and tuberculous spondylitis (TS) using MRI scans. This AI approach offers a promising tool for precise diagnosis of spinal infections.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Infectious Diseases
Background:
- Brucella spondylitis (BS) and tuberculous spondylitis (TS) are distinct spinal infections requiring accurate differentiation for effective treatment.
- Current diagnostic methods using imaging and pathogen detection have limitations in speed and accuracy.
- Deep learning (DL) offers a potential solution for improving diagnostic capabilities.
Purpose of the Study:
- To investigate the feasibility of using deep learning models with conventional MRI to differentiate between BS and TS.
- To develop and evaluate a DL model for improved diagnostic accuracy in spinal infections.
Main Methods:
- A dataset of 310 patients (209 BS, 101 TS) was used, split into training, testing, and external validation sets.
- A ResNeXt-50 architecture integrated with a Convolutional Block Attention Module (CBAM) was trained on sagittal T2-weighted MRI images.
- Model performance was assessed using AUC, accuracy, precision, recall, and F1-score, compared against other DL models.
Main Results:
- The CBAM-ResNeXt model demonstrated superior performance compared to general DL models.
- Key performance metrics included accuracy (0.942), precision (0.940), recall (0.928), F1-score (0.934), and AUC (0.953).
Conclusions:
- The developed deep learning model shows significant potential for diagnosing BS and TS using conventional MRI.
- This AI tool can serve as a valuable aid in clinical practice for distinguishing between these two prevalent spinal infections.
