LOOC: Localizing Organs Using Occupancy Networks and Body Surface Depth Images

2024-06-18

IEEE Access

AI Summary

This study addresses the challenge of non-invasively and accurately localizing 67 deep anatomical structures from a single human surface depth image. Methodologically, it introduces, for the first time, a multi-class occupancy network for 3D organ localization under occlusion, trained with supervision from CT-derived organ segmentations, and incorporates a robust pose-aware sampling strategy tailored for densely packed organs to construct personalized 3D anatomical atlases. Key contributions include: (1) the first formulation of deep organ spatial occupancy via occupancy representation; (2) an organ-density-adaptive sampling mechanism; and (3) an end-to-end mapping from surface depth images to 3D anatomical coordinates. Experiments demonstrate significantly higher localization accuracy than conventional template-matching approaches and enable qualitative 3D reconstruction in real-world scenarios, establishing a novel paradigm for non-invasive, automated medical image analysis.

Technology Category

• Computer Vision: 3D Computer Vision
• Intelligent Robots: Localization, Mapping, and Navigation
• Machine Learning: Learning with Manifolds

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

We introduce a novel approach for the precise localization of 67 anatomical structures from single depth images captured from the exterior of the human body. Our method uses a multi-class occupancy network, trained using segmented CT scans augmented with body-pose changes, and incorporates a specialized sampling strategy to handle densely packed internal organs. Our contributions include the application of occupancy networks for occluded structure localization, a robust method for estimating anatomical positions from depth images, and the creation of detailed, individualized 3D anatomical atlases. We outperform localization using template matching and provide qualitative real-world reconstructions. This method promises improvements in automated medical imaging and diagnostic procedures by offering accurate, non-invasive localization of critical anatomical structures.

Problem

• Localizing 67 anatomical structures from single depth images
• Handling densely packed internal organs with specialized sampling
• Improving automated medical imaging via non-invasive structure localization

Innovation

• Multi-class occupancy network for organ localization
• Depth image-based anatomical position estimation
• Specialized sampling for dense organ handling