Abstract
Cardiac Magnetic Resonance (CMR) imaging is a critical tool for diagnosingand managing cardiovascular disease, yet its utility is often limited by thesparse acquisition of 2D short-axis slices, resulting in incomplete volumetricinformation. Accurate 3D reconstruction from these sparse slices is essentialfor comprehensive cardiac assessment, but existing methods face challenges,including reliance on predefined interpolation schemes (e.g., linear orspherical), computational inefficiency, and dependence on additional semanticinputs such as segmentation labels or motion data. To address theselimitations, we propose a novel \textbf{Ca}rdiac \textbf{L}atent\textbf{I}nterpolation \textbf{D}iffusion (CaLID) framework that introducesthree key innovations. First, we present a data-driven interpolation schemebased on diffusion models, which can capture complex, non-linear relationshipsbetween sparse slices and improves reconstruction accuracy. Second, we design acomputationally efficient method that operates in the latent space and speedsup 3D whole-heart upsampling time by a factor of 24, reducing computationaloverhead compared to previous methods. Third, with only sparse 2D CMR images asinput, our method achieves SOTA performance against baseline methods,eliminating the need for auxiliary input such as morphological guidance, thussimplifying workflows. We further extend our method to 2D+T data, enabling theeffective modeling of spatiotemporal dynamics and ensuring temporal coherence.Extensive volumetric evaluations and downstream segmentation tasks demonstratethat CaLID achieves superior reconstruction quality and efficiency. Byaddressing the fundamental limitations of existing approaches, our frameworkadvances the state of the art for spatio and spatiotemporal whole-heartreconstruction, offering a robust and clinically practical solution forcardiovascular imaging.