Enhanced Lumbar Disease Classification through Hybrid Deep Learning Methods

ABSTRACT

Accurate and efficient classification of lumbar diseases is crucial for maintaining patient health and preventing significant economic burdens on healthcare systems. This study proposes a novel approach to automate lumbar disease classification by integrating multiple deep learning architectures, including MobileNet, DenseNet, and a hybrid CNN-SVM model. This combination of advanced models leverages their strengths in feature extraction and classification. Additionally, an Involutional Neural Network-based VGG architecture is employed to further enhance the learning capability and performance of the system, particularly in handling complex and detailed features associated with lumbar conditions.

The proposed method was evaluated using a comprehensive dataset of medical images related to various lumbar diseases. Experimental results demonstrate a significant improvement in classification accuracy and computational efficiency compared to traditional CNN-based approaches. This system provides a promising solution for automated lumbar disease classification, with potential applications in healthcare for real-time diagnosis and monitoring.

Keywords: Lumber disease classification, MobileNet, DenseNet, CNN-SVM, Involutional Neural Networks, deep learning, image analysis.