A Multi-Task Attention-Driven SegNet for Lung Infection Segmentation and Classification From HRCT Images

Abstract

High-resolution computed tomography (HRCT) plays a critical role in diagnosing lung diseases by providing detailed visualization of infections, inflammation, fibrosis and tumors. Recent advances in deep learning have enabled automated detection and segmentation of infected regions, with attention mechanisms further enhancing feature discrimination. This study proposes a multi-task attention-driven SegNet model for automated analysis of HRCT Scans. The model integrates attention layers to emphasize key pathological regions, including consolidation and ground-glass opacities, and supports both binary and multi-class segmentation/classification tasks. Experimental evaluation demonstrates that the proposed model achieves 91.4% accuracy in binary and 84.5% accuracy in multi-class lung infection segmentation and classification, outperforming the attention-driven U-Net baseline. By providing improved precision and efficiency, the attention-driven SegNet model can accelerate computer-aided diagnosis (CAD), support clinical decision-making, and promote standardized approaches for managing lung infections in healthcare settings.

More by Dmitry Kaplun

• Codewithsayanjib/slm-systematic-review: v1.1Zenodo · 2026
• Microscopy Image Dataset of pulmonary vessels for Quantitative assessment of fibrosisScientific Data · 2026
• Microscopy Image Dataset of pulmonary vessels for Deep Learning-Based Quantitative assessment of fibrosisFigshare · 2026
• MA2MU-Net++: A Mahalanobis distance and Ackley function based multi-modal U-Net++ model for breast nuclei segmentation in histopathology imagesBiomedical Signal Processing and Control · 2026