3D Reconstruction of Human Body Organs

Abstract

MedReconstruct-Lite serves as a computational system which reconstructs human body organs in three dimensions by using multiple medical imaging techniques without applying artificial intelligence. The traditional organ reconstruction techniques of classical methods require either manual segmentation surface fitting or voxel modeling, which results in slow processing times and rare cases of producing geometry errors. The system operates through three steps which involve image preprocessing and image segmentation through thresholding and edge detection and volumetric reconstruction with marching cubes and surface meshing techniques. The process of point cloud registration together with mesh smoothing establishes consistent surface appearance while maintaining anatomical features. The generated three-dimensional surfaces enable researchers to examine and assess anatomical structures for medical research and diagnostic purposes and surgical planning. The MedReconstruct-Lite prototype we created establishes a rapid automatic system which produces anatomically accurate three-dimensional organ models through processing CT and MRI and ultrasound image slices. The future plans include developing real-time reconstruction capabilities together with enhanced mesh optimization and the integration of haptic technology and AR-based visualization systems for medical training and simulation purposes. Index Terms: 3D Reconstruction, Medical Imaging, Image Segmentation, Volumetric Modeling, Point Cloud Registration, Surface Meshing, Computational Anatomy, Visualization in Healthcare.