micro-ct

Library for performing multiphase simulations (based on the Shan-Chen model) in complicated geometries (i.e. porous media 3D images)

Deep-Learning for Tidemark Segmentation in Human Osteochondral Tissues Imaged with Micro-computed Tomography

Permeability tensors of 3D pore images on one GPU or on CPU cores: lattice Boltzmann with TRT collision and pore-only storage, verified against analytical and published benchmarks.

Image quality metrics for micro-CT reconstructions: MTF, NPS, NEQ, TTF and detectability index (d'), with automatic ROI finding and a one-command report

CoralLite: μCT reconstruction of coral colonies from individual corallites (CVBMC @ ICPR 2026)

Porosity REV (Representative Elementary Volume) from a segmented 3D image using random cube sampling.

Deep learning framework for micro-CT projection infilling - recovering missing projections to enable undersampled scanning with reduced radiation dose.

A particle tracking library

AI-powered TaxaMask workbench for morphology mask annotation, taxonomic evidence review, 2D/STL datasets, and TIF/CT micro-CT volume workflows with SAM/VLM.

3D surface segmentation of ancient Herculaneum papyrus scrolls from micro-CT scans. Dual-pipeline architecture: V1 (CE+Dice, clDice) and V2 (Focal Loss, Deep Supervision, Skeleton Recall). Built for the Vesuvius Challenge $200K Kaggle competition. CC BY-NC 4.0.

High-throughput maize kernel & embryo volume phenotyping from micro-CT and deep learning (VGG16-UNet); PyTorch + ONNX/DirectML builds. 基于显微 CT 与深度学习的玉米籽粒与胚体积高通量表型系统

This project is the codebase to: Impact of Gaussian Feathering on Diagnostic Metrics in Tile-Based Micro-CT Sinogram Infilling

Reusable pipeline for segmenting hair follicles, sweat glands, and blood vessels in bovine skin biopsies via micro-CT: Dragonfly protocols (manual + 2.5D U-Net DL), conversion utilities, and analysis scripts.

Pore and grain size distribution from a segmented 3D image using local thickness method.

Various reconstruction algorithms for the eXplore CT 120 scanner (Trifoil Imaging, USA) as a substitute/improvement to the GEHC specific FDK reconstruction software.

Extract ROIs (e.g. nerves, blood vessels) from multi-terabyte cloud-hosted medical scans.

From synchrotron micro-CT of tooth enamel to biomimetic lattices: parametric CAD, finite-element analysis, and closed-loop design optimization.