PSCT-Net: Geometry-Aware Pediatric Skull CT Reconstruction via Differentiable Back-Projection and Attention-Guided Refinement
This work addresses the need for low-dose 3D imaging in pediatric craniofacial diagnosis, offering a geometry-aware solution that improves upon geometry-agnostic approaches.
PSCT-Net reconstructs 3D pediatric skull CT from sparse bi-planar X-rays using differentiable back-projection and attention-guided refinement, achieving high-quality osseous boundaries with reduced depth ambiguity. On a private pediatric dataset, it outperforms prior methods in reconstruction accuracy.
Computed Tomography (CT) is essential for diagnosing pediatric craniofacial abnormalities, yet poses radiation risks to developing anatomies. Reconstructing 3D CT from sparse bi-planar X-rays offers a low-dose alternative but is severely ill-posed. Existing methods employ geometry-agnostic feature lifting, naively projecting 2D features into 3D without explicit spatial modeling, causing depth ambiguity and degraded osseous boundaries. We present PSCT-Net, a geometry-aware framework with differentiable back-projection. Differentiable back-projection establishes a spatially faithful volumetric prior, alleviating depth ambiguity. An Attention-Guided Projection (AGP-3D) module then learns non-linear voxel-wise correspondences between 2D regions and 3D locations. A Bidirectional Mamba (BiM-3D) module captures long-range volumetric dependencies with linear complexity. We further curate a private institutional pediatric skull CT cohort, PedSkull-CT, comprising normal and pathological cases for internal evaluation, addressing the gap in adult-centric, trunk-focused datasets.