PulseAugur
EN
LIVE 07:39:30

Bi-PT pipeline reconstructs 3D heart meshes from sparse cardiac MRI data

By PulseAugur Editorial · [2 sources] ·

Researchers have developed Bi-PT, a novel pipeline for reconstructing 3D four-chamber heart meshes from sparse cardiac MRI data. This method utilizes bidirectional cross-attention point transformers to learn robust point features by comparing an atlas with sparse point clouds extracted from clinical scans. Bi-PT formulates the deformation field as a Neural Ordinary Differential Equation (NODE) to ensure diffeomorphic transformations, incorporating semantic label loss and smoothness regularization for improved accuracy and stability. AI

IMPACT This research could lead to more accurate and efficient 3D heart modeling from standard MRI scans, potentially improving diagnostic capabilities.

RANK_REASON The cluster contains a research paper detailing a new method for medical image reconstruction.

Read on arXiv cs.CV →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

Bi-PT pipeline reconstructs 3D heart meshes from sparse cardiac MRI data

COVERAGE [2]

  1. arXiv cs.CV TIER_1 English(EN) · Chenchuhui Hu, Shaoming Pan, Leon Axel, Meng Ye ·

    Bi-PT: Bidirectional Cross-Attention Point Transformers for Four-Chamber Heart Reconstruction from Sparse Cardiac MRI Data

    arXiv:2607.06923v1 Announce Type: new Abstract: We propose Bi-PT, a pipeline for reconstructing 3D four-chamber human heart meshes from clinical sparsely sampled cardiac magnetic resonance imaging (CMR) data. This work addresses the error-prone generation of 3D cardiac shape from…

  2. arXiv cs.CV TIER_1 English(EN) · Meng Ye ·

    Bi-PT: Bidirectional Cross-Attention Point Transformers for Four-Chamber Heart Reconstruction from Sparse Cardiac MRI Data

    We propose Bi-PT, a pipeline for reconstructing 3D four-chamber human heart meshes from clinical sparsely sampled cardiac magnetic resonance imaging (CMR) data. This work addresses the error-prone generation of 3D cardiac shape from a sparse point cloud (SPC) extracted from 2D lo…

RELATED ENTITIES

RELATED TOPICS