The following is a summary of “LSS-VGG16: Diagnosis of Lumbar Spinal Stenosis With Deep Learning,” published in the June 2023 issue of Spinal Disorders and Techniques by Altun et al.
The design of this investigation was retrospective. Lumbar Spinal Stenosis (LSS) is a condition that causes chronic low back pain and is frequently mistaken for a herniated disk. This study proposes a classification model based on deep learning as a rapid and objective method for diagnosing LSS. LSS is a disease that causes low back pain, foot numbness, and discomfort, among other adverse effects. This disease is difficult to diagnose because it is often mistaken for a herniated disk and requires high expertise.
The form and severity of this stenosis play a crucial role in determining the surgery and surgical technique to be used on these patients. When the spinal canal becomes constricted due to compression on these nerves and pressure on the vessels supplying the nerves, inadequate nutrition of the nerves results in a loss of function and structure. Applying image processing techniques to biomedical images, such as MR and CT, results in successful classification. On this basis, computer-assisted diagnosis systems can be created to aid in diagnosing various diseases.
Various deep learning and conventional machine learning techniques have been investigated to demonstrate the proposed model’s efficacy. The VGG16 classification technique yielded the highest success rate, with 87.70%. The proposed LSS-VGG16 model demonstrates the feasibility of developing a computer-aided diagnosis system for diagnosing spinal canal stenosis. In addition, a higher classification success rate was observed compared to other investigations of a similar nature. This demonstrates that the proposed LSS-VGG16 model will be a valuable tool for scientists in this field.