Skill v1.0.0

currentAutomated scan100/100

mkurman/zorai/simpleitk

──Details

PublishedJuly 29, 2026 at 12:09 PM

Content Hashsha256:78e6bd7a2fa02496...

Git SHAd0acbfaf3d62

──Files

Files (1 file, 2.0 KB)

SKILL.md2.0 KBactive

SKILL.md · 64 lines · 2.0 KB

version: "1.0.0" name: simpleitk description: "Simplified interface to the Insight Toolkit (ITK) for medical image processing. Segmentation, registration, filtering, resampling, morphological operations. Supports DICOM, NIfTI, NRRD, dozens of formats." tags: [medical-image-processing, registration, dicom-workflows, itk-wrappers, simpleitk]

Overview

SimpleITK simplifies the Insight Toolkit (ITK) for medical image processing: segmentation, registration, filtering, resampling, and morphological operations. Supports DICOM, NIfTI, NRRD, and 50+ file formats.

Installation

bash

uv pip install SimpleITK

Basic Image Operations

python

import SimpleITK as sitk
import numpy as np
image = sitk.ReadImage("ct_scan.nii.gz")
print(image.GetSize(), image.GetSpacing(), image.GetOrigin())
array = sitk.GetArrayFromImage(image)
print(array.shape)  # (z, y, x)

Segmentation

python

binary = sitk.BinaryThreshold(image, lower=200, upper=500, insideValue=1, outsideValue=0)
cc = sitk.ConnectedComponent(binary)
stats = sitk.LabelIntensityStatisticsImageFilter()
stats.Execute(cc, image)
for label in stats.GetLabels():
    print(f"Label {label}: mean={stats.GetMean(label):.1f}")

Registration

python

fixed = sitk.ReadImage("template.nii.gz")
moving = sitk.ReadImage("moving.nii.gz")
R = sitk.ImageRegistrationMethod()
R.SetMetricAsMattesMutualInformation(numberOfHistogramBins=50)
R.SetOptimizerAsGradientDescent(learningRate=1.0, numberOfIterations=100)
R.SetInitialTransform(sitk.CenteredTransformInitializer(fixed, moving, sitk.Euler3DTransform()))
final_transform = R.Execute(fixed, moving)
resampled = sitk.Resample(moving, fixed, final_transform, sitk.sitkLinear)

Workflow

Read images with sitk.ReadImage() (auto-detects format)
Preprocess: BinaryThreshold, MedianFilter, ResampleImageFilter
Segment with thresholding, watershed, or connected components
Register with ImageRegistrationMethod + transform
Measure volumes with LabelStatisticsImageFilter
Write results with sitk.WriteImage()

All versions