Predicting the anterior slippage of vertebral lumbar spine using Densenet-201

Lumbar spine illnesses have become more widespread in recent years due to sedentary lifestyles, accidents, and lifestyle changes. Spondylolisthesis is the most common illness, accounting for 5% of all cases worldwide. Both abnormalities can occur in children while they are very young and, if disregarded, can progress to agonizing pain. As a result, early detection can facilitate the implementation of therapies and interventions and halt the disease's spread. This investigation uses radiography scans to determine if any anterior slippage of the lumbar spine's vertebrae exists. This paper suggests a sequential approach to processing data and classifying lumbar spine pictures using deep learning's DenseNet 201, which can efficiently distinguish between normal and Lumbar Spine Anterior Slippage radiography images. Lumbar spine radiograph images are processed with is proven by comparing with conventional preprocessing. The Shah Orthopedic Hospital in Miraj is where the datasets were gathered, which consisted of both the front and side radiological images of the lumbar spine. The result with 30 epochs demonstrates that the DenseNet201 model produces the highest accuracy. The method's Accuracy, Sensitivity, Specificity, & precision are 95.17%, 96.51%, 89.23%, and 93.51%, respectively. According to the comparative data, our technique got the maximum accuracy compared to the other methods.

Automated summary

Lumbar spine illnesses, such as spondylolisthesis, have become more common and early detection is crucial. This study proposes a sequential approach using deep learning's DenseNet 201 to process and classify lumbar spine images for detecting anterior slippage. The method achieves high accuracy and outperforms other techniques.