Lune

CVPR Findings2026

Two-Stage 3D Pulmonary Vessel Reconstruction via Trunk--Expansion Coupled Point Cloud Generation

Jie Zhang, Yu Xin, Guoqing Li

Abstract

This paper addresses the challenges in pulmonary vascular 3D reconstruction, where CT slice discreteness, boundary ambiguity, and stage-wise processing in traditional methods often lead to topological disconnections and loss of geometric details. We propose a two-stage point cloud generation algorithm based on "trunk-expansion" coupled reconstruction. To tackle the problem of topological discontinuity, the extraction of the vascular trunk is reformulated as an n-part graph modeling problem across n slices. A Transformer-based joint inference strategy is employed to simultaneously identify key points and their connections, effectively suppressing error accumulation and preserving the integrity of the tree-like topology. To address boundary ambiguity and incomplete expansion, we introduce a trunk-skeleton-guided Signed Distance Function (SDF) reconstruction mechanism, which integrates image features with geometric priors. By applying offset sampling and density weighting, the method achieves boundary-aware point cloud generation, significantly improving boundary accuracy and anatomical consistency. Furthermore, by coupling trunk topology reconstruction with expansion generation in a unified framework, the proposed method enhances the structural continuity and morphological plausibility of the reconstructed 3D point clouds.Extensive experiments on pulmonary CT datasets demonstrate that our method achieves superior topological integrity and geometric accuracy compared with state-of-the-art approaches.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • High-Resolution Image Synthesis with Latent Diffusion Models cites Denoising Diffusion Probabilistic Models
  • Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on17

  1. Denoising Diffusion Probabilistic ModelsJonathan Ho, Ajay Jain, Pieter AbbeelNeurIPS 2020 · 35,902 citations
  2. An Image is Worth 16x16 Words: Transformers for Image Recognition at ScaleAlexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn et al.ICLR 2021 · 21,477 citations
  3. High-Resolution Image Synthesis with Latent Diffusion ModelsRobin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser et al.CVPR 2022 · 13,123 citations
  4. 3D Gaussian Splatting for Real-Time Radiance Field RenderingBernhard Kerbl, Georgios Kopanas, Thomas Leimkühler, George DrettakisSIGGRAPH 2023 · 5,687 citations
  5. Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without ConvolutionsWenhai Wang, Enze Xie, Xiang Li, Deng-Ping Fan et al.ICCV 2021 · 4,909 citations

Related papers

  1. VesMamba: 3D Pulmonary Vessel Segmentation from CT images via Mamba with Structural Perception and Scale-aware FilteringZhipeng Liu, Guilian Chen, Zheng Jiang, Huisi Wu et al.CVPR 2026
  2. NETracer: A Topology-Aware Iterative Tracing Approach for Tubular Structure ExtractionChao Liu, Yangbo Jiang, Nenggan ZhengICCV 2025 · 2 citations
  3. Rethinking Multimodal Point Cloud Completion: A Completion-by-Correction PerspectiveWang Luo, Di Wu, Hengyuan Na, Yinlin Zhu et al.AAAI 2026
  4. CrossSDF: 3D Reconstruction of Thin Structures From Cross-SectionsThomas Walker, Salvatore Esposito, Daniel Rebain, Amir Vaxman et al.CVPR 2025
  5. Accurate Point Cloud Registration with Robust Optimal TransportZhengyang Shen, Jean Feydy, Peirong Liu, Ariel Hernán Curiale et al.NeurIPS 2021 · 81 citations
Dusk over the sea between two cliffs drawn in fine vertical lines