Cervical Spondylosis Diagnosis Based on Convolutional Neural Network with X-ray Images

Area of Science:

• Medical Imaging
• Artificial Intelligence in Healthcare
• Spinal Diagnostics

Background:

• Rising cases of Cervical Spondylosis, affecting younger demographics, increase the need for accurate X-ray screening.
• Variability in imaging technology, equipment, and clinician experience challenges diagnostic accuracy for Cervical Spondylosis.

Purpose of the Study:

• To develop and validate a deep learning model for improved Cervical Spondylosis diagnosis from X-ray images.
• To address diagnostic inconsistencies caused by diverse imaging parameters and clinician expertise.

Main Methods:

• A ResNet-34 convolutional neural network was trained on 1235 cervical spine X-ray images with varied projection angles.
• The model's performance was validated on an independent dataset of 136 X-ray images, also with diverse projection angles.

Main Results:

• The deep learning model achieved a classification accuracy of 89.7% for Cervical Spondylosis detection.
• This accuracy significantly surpasses the traditional manual diagnostic approach, which yielded 68.3% accuracy.

Conclusions:

• Deep learning models, like the developed ResNet-34, can effectively enhance diagnostic accuracy and efficiency for Cervical Spondylosis.
• This AI-driven approach offers a viable method to complement clinical expertise in identifying Cervical Spondylosis.