OpenParable
Skip to content
← Back to skills

Simpleitk

Documents
simpleitk

Simpleitk is an Agent Skill for documents published in the mkurman/zorai repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Simplified interface to the Insight Toolkit (ITK) for medical image processing. Segmentation, registration, filtering, resampling, morphological operations. Supports DICOM, NIfTI, NRRD, dozens of formats.

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/mkurman/zorai/blob/HEAD/skills/scientific-skills/simpleitk/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/simpleitk, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

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

uv pip install SimpleITK

Basic Image Operations

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

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

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

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