MRI method improves 3D heart shape reconstruction from limited image slices
MRI-Guided Reslice-Refined Cross-Slice SDF Reconstruction of the Left Ventricle from Cardiac MRI with Sparse Axial Supervision
Computer Vision and Pattern Recognition
Summary
Rebuilding a detailed 3D model of the heart’s left ventricle from MRI scans is hard when only a few cross-sectional images are clear. The authors developed a method that starts with rough guesses from the available images, then refines the heart’s shape using image boundaries and consistency checks, even when segmentation boundaries are imperfect. Their approach works well across different segmentation tools and various amounts of data, producing more accurate heart models from sparse information. This could help make heart imaging more reliable without needing many image slices.
What this means in practice
- •For medical imaging teams: Generate accurate 3D models of the left ventricle from cardiac MRI scans with few annotated slices to aid diagnosis and treatment planning.
- •For medical device developers: Incorporate a refinement algorithm for left ventricular shape reconstruction into cardiac imaging software to enhance performance with limited manual annotation.$Commercial implications: Enables more reliable 3D heart modeling from fewer MRI slices, which can be sold as an improved cardiac imaging tool feature.
Authors
Quanxin Zheng, Shuai Zhao
Abstract
Reconstructing a three-dimensional left-ventricular (LV) endocardial surface from cardiac magnetic resonance (CMR) data is challenging when supervision is available on only a small number of axial slices. Through-plane geometry is weakly constrained, and automatically generated two-dimensional masks can propagate segmentation errors into the recovered shape. We present MR-RS-SDFR, a per-case implicit signed distance field (SDF) framework that reconstructs a continuous LV surface from a CMR volume and sparse axial weak masks. The method first builds a cross-slice SDF initialization from axial and longitudinal geometric cues and then refines the field using two complementary signals: MRI edge-field normal alignment, which provides an image-derived boundary cue independent of the weak masks, and differentiable reslice Dice and contour consistency, which preserve agreement with the observed planes. We evaluate three weak-mask generators -- LOO TransUNet, LOO nnU-Net, and an off-the-shelf Medical SAM3 model used without MM-WHS-specific training or fine-tuning -- and five sparsity levels from 4 to 64 axial planes. In the sparse-16 setting, final MR-RS-SDFR reconstruction reaches 0.928 Dice and 3.80mm HD95 with Medical SAM3 masks. The upstream generators do not exhibit a single common ranking across 2D and dense 3D segmentation, and nnU-Net- and Medical-SAM3-driven sparse reconstruction achieve the same mean final Dice despite different upstream error profiles. Across all three sparse-16 mask sources, MR-RS-SDFR is numerically better than protocol-matched full GHD+DVS in both Dice and HD95. Final Dice improves markedly from sparse-4 to sparse-16 and then saturates at the reported precision through sparse-64. These results support MRI-guided per-case SDF refinement as a reconstruction strategy that remains effective across weak-mask generators and supervision densities.