Integrating Spatial and Channel Features for Multi-Disease Classification Using Attention-Based CNN

Abstract

The increasing adoption of deep learning in medical image analysis has enabled significant improvements in disease detection and classification. This paper presents a multiple disease classification model that integrates spatial and channel feature extraction with a Convolutional Neural Network (CNN) enhanced by an attention mechanism. The model is trained on a chest X-ray images to classify four different diseases, and its performance is evaluated using key metrics. Experimental results demonstrate that the proposed model achieves an accuracy of 94.0%, with an average precision of 94.4%, recall of 94.6%, and F1-score of 93.4%, outperforming conventional CNN architectures. An ablation study was conducted to show the effectiveness of the spatial and channel feature extraction components in improving classification performance.