aylward / itkARGUS

Anatomic Reconstruction for Generalizable UltraSound (ARGUS) AI: A library for point-of-care ultrasound AI using MONAI and ITK

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

itkARGUS

License

Overview

Contains C++ methods used by Kitware's Anatomic Reconstruction for Generalized UltraSound project. These methods are implemented in C++ for speed, and then wrapped into python for use in the project.

Build Steps

Option 1: If you don't have a from-source build of ITK: 1. git clone --depth=1 https://github.com/InsightSoftwareConsortium/ITKPythonPackage 2. ./ITKPythonPackage/scripts/dockcross-manylinux-download-cache-and-build-module-wheels.sh

  • If the command errors with "curl: failed to write body", then chown -R the tools/ folder to be your user. Then, re-run the above command.
  1. To rebuild, run ./ITKPythonPackage/scripts/dockcross-manylinux-build-module-wheels.sh

Option 2: If oyu have a from-source build of ITK: 1. git clone https://github.com/KitwareMedical/AnatomicRecon-POCUS-AI 2. mkdir itkARGUS-Release 3. cd itkARGUS-Release 4. cmake-gui ../AnatomicRecon-POCUS-AI

  • Make certain to specify CMAKE_BUILD_TYPE=Release.
  • It should automatically find your ITK build directory
  1. Compile the project. It should then be wrapped for python and
available via import itk.

Usage

Sample usage can be found in usage/resample.py. It expects a US video in video.mp4 and outputs a output.nrrd file. Be sure to install the packages found in requirements.txt.

About

Anatomic Reconstruction for Generalizable UltraSound (ARGUS) AI: A library for point-of-care ultrasound AI using MONAI and ITK

License:Apache License 2.0


Languages

Language:Python 78.8%Language:C++ 12.5%Language:C 7.5%Language:CMake 1.2%