Sparse-View CT Reconstruction via Neural Signed Distance Functions and Volumetric Attenuation Fields
Published in Senior Thesis (targeting medical imaging venues), 2026
This senior thesis develops a novel approach to sparse-view CT reconstruction using neural signed distance fields (SDFs) and volumetric attenuation fields. The method aims to improve medical imaging quality while reducing radiation exposure through sparse sampling strategies. Primary contribution is unsupervised material grouping for multi-material CT reconstruction via SDFs.
Targeting publication in medical imaging venues.
Status: Manuscript in preparation
Co-authors: Daksh K. Shah, Emmanouil Nikolakakis, Razvan Marinescu
Recommended citation: Shah, D.K., Nikolakakis, E., Marinescu, R. (2026). Sparse-View CT Reconstruction via Neural Signed Distance Functions and Volumetric Attenuation Fields. Senior Thesis [In Preparation].