medcore 1.5.0

Medical imaging utilities based on SimpleITK

medcore

Medical imaging utilities for DICOM / NIfTI workflows based on SimpleITK.

Features

• IO
  • DICOM / NIfTI reading (ImageReader)
• Image utils
  • intensity array conversion / normalization
  • resampling / transform helpers
  • NIfTI write and label merge utilities
• Detection
  • coronal angle estimation
  • umbilicus detection helpers
• Segmentation
  • torso and abdomen segmentation classes
• Feature extraction
  • label area/volume summary
  • patch extraction around landmark points

Installation

From source

git clone https://github.com/yhchoi-data/medcore.git
cd medcore
pip install .

From source (developer mode)

git clone https://github.com/yhchoi-data/medcore.git
cd medcore
pip install -e ".[dev]"

From pypi [to be updated]

pip install medcore

Optional: DICOM to NIfTI conversion with dcm2niix

medcore.io.convert_dicom_to_nifti uses the external dcm2niix executable. Install it separately when you need dcm2niix-based DICOM to NIfTI conversion:

conda install -c conda-forge dcm2niix

After installation, make sure the executable is available in your shell or notebook kernel:

dcm2niix -h

Quick start

from medcore.io import ImageReader, convert_dicom_to_nifti
from medcore.utils import (
    sitk_get_array,
    sitk_write_nii,
    sitk_read_labelfiles,
)

from medcore.detect import UmbilicusPredictor, UmbilicusDetector
from medcore.segment import TorsoSegmenter, AbdomenSegmenter
from medcore.feature import compute_label_volumes, extract_patches_from_image

Package usage

# IO
from medcore.io import ImageReader

# Utility functions
from medcore.utils import sitk_resampler, figure_overlay_label_on_slices

# Detection / segmentation
from medcore.detect import UmbilicusPredictor
from medcore.segment import TorsoSegmenter

# Feature extraction
from medcore.feature import compute_label_areas, compute_label_volumes

Documentation

Detailed examples are in USAGE.md.

Developer workflow

pre-commit run --all-files

On the first run, some hooks may show Failed because they automatically fix files such as trailing whitespace or formatting. That is normal. Run pre-commit run --all-files once more and it should pass after the auto-fixes are applied.