Lesion Segmentation and CT Biomarkers for Tuberculosis Assessment and Follow-Up
Development of segmentation-based CT biomarkers for longitudinal tracking and assessment of tuberculosis, enabling quantitative disease follow-up.
Research
Peer-reviewed papers in Deep Learning and Medical Imaging. Filters apply across both year and tag simultaneously.
5 publications
Development of segmentation-based CT biomarkers for longitudinal tracking and assessment of tuberculosis, enabling quantitative disease follow-up.
A graph-based architecture that captures structural topology in 3D volumetric lung data for accurate airway tree segmentation using multiscale attention mechanisms.
A diffusion model for realistic synthesis of lung CT textures, enabling data augmentation for medical image analysis tasks with limited annotated data.
A geometry-constrained Cycle-GAN framework that generates synthetic pathological lung tissues while preserving anatomical shape, improving segmentation robustness.
A multi-modal approach using vascular topology for automatic lung lobe partitioning in ultra-short echo time MRI, bridging CT and MRI analysis.