Scholarlink

Until the last few years, Cervical Spondylosis is an age-based degenerative spine disease. However, in a few years, it has become a problem for everyone, irrespective of the age group. We are choosing X-ray images since they are cost-friendly for every victim. Challenges include the difference in the various adapted scanning equipment, the ornaments the patients wear that were not kept aside during the scan, and the lack of patients in our demographical region affected by c-spine disorders. After a survey, we found that doctors manually inspect X-rays, and very few automatic diagnosis methods are available. In response, we are implementing a deep learning approach using the DenseNet-121 Convolutional Neural Network model along with the Generative Adversarial Network (GAN) for augmentation of the medical image dataset since the medical image data is minimal. Our model is used on a comprehensive dataset of 133 cervical spine X-ray images covering a wide range of projection angles to overcome these limitations and enhance diagnostic accuracy. Our study resulted in 95.13% accuracy. This progress shows that deep learning models can assist and improve clinicians’ ability to diagnose Cervical Spondylosis.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Intelligent Cervical Spondylosis Diagnosis Using Hybrid GAN-DenseNet Deep Learning Approach

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • U. D. Prasan,
  • PanduRanga Vital Terlapu,
  • Ruvva Pujitha,
  • Gorle Gayatri,
  • Savalapurapu Jhansi,
  • Morri Adithya

摘要

Until the last few years, Cervical Spondylosis is an age-based degenerative spine disease. However, in a few years, it has become a problem for everyone, irrespective of the age group. We are choosing X-ray images since they are cost-friendly for every victim. Challenges include the difference in the various adapted scanning equipment, the ornaments the patients wear that were not kept aside during the scan, and the lack of patients in our demographical region affected by c-spine disorders. After a survey, we found that doctors manually inspect X-rays, and very few automatic diagnosis methods are available. In response, we are implementing a deep learning approach using the DenseNet-121 Convolutional Neural Network model along with the Generative Adversarial Network (GAN) for augmentation of the medical image dataset since the medical image data is minimal. Our model is used on a comprehensive dataset of 133 cervical spine X-ray images covering a wide range of projection angles to overcome these limitations and enhance diagnostic accuracy. Our study resulted in 95.13% accuracy. This progress shows that deep learning models can assist and improve clinicians’ ability to diagnose Cervical Spondylosis.