AnaToPrint

Medical CT-Scan 3D Printing Web Application

Capstone project for BS in Computer Science, University of Maine

Quick Start

Web-app is now hosted using Vercel To access the web-app with no install follow the below link:

https://ana-to-print.vercel.app/

Manual Start and Codebase installation

1. Clone the repository

   git clone
   cd AnaToPrint

2. Install dependencies

   npm install

3. Start development server

   npm run dev

4. Open browser to http://localhost:5173

Running Tests

Unit & Component Tests (Vitest)

cd webapp
npx vitest run
npx vitest run --coverage

E2E Tests (Playwright)

# Install browsers (first time only)
npx playwright install --with-deps

# Run on all browsers
npm run test:e2e

Tech Stack

• Frontend: React + TypeScript
• Build Tool: Vite
• 3D Visualization: VTK.js
• DICOM Processing: ITK.WASM + jszip
• Testing: End-to-End Testing - Playwright + Unit & Component Testing - Vitest

Project Structure

• docs/ - Project documentation (proposal, SRS)
• webapp/ - React frontend application
• tests/ - End-to-end tests

Features

• Upload medical CT scans (DICOM format)
• View 3D reconstructed models
• Segment anatomical structures (bone, skin, muscle)
• Export STL files for 3D printing
• Export G-code for printer-specific tissue characteristics
• All processing done locally (no server uploads)

Team

• Israk Arafat
• Gregory Michaud
• Cooper Stepenkiw
• Bryan Sturdivant
• Ethan Wyman

License

University of Maine - Laboratory for Convergent Science

What This Project Is

This capstone project is a client-side web application that takes a medical CT scan (in DICOM format) and turns it into files ready for 3D printing. Specifically, it can produce:

• Polygonal .stl model of anatomy (e.g., skull/bone) suitable for 3D printing or further processing
• .gcode file that, when printed, mimics the X-ray attenuation properties (density/opacity) of the original tissue

All processing happens in the user’s web browser on their own machine: no servers, no cloud, no uploads. This is to protect medical privacy and make the tool easy to run on any common desktop platform.

Application Capabilities

Upload CT Data

• Accept a DICOM file or folder of CT slices from the local machine
• Validate and parse into a 3D voxel volume with associated metadata (spacing, orientation, etc.)

Select Anatomy to Focus On

• Choose between bone, skin, or muscle
• Apply appropriate Hounsfield Unit thresholds/presets to segment the chosen tissue

Preview the 3D Result

• Render an interactive 3D preview in the browser (rotate, zoom, inspect)
• Allow threshold tweaks/view adjustments before export

Export for 3D Printing

• Export a polygon mesh as .stl using a Marching Cubes–based surface extraction
• Export 3D-printer instructions as .gcode that encode geometry and tissue-mimicking properties
• Let the user pick the file name and local download location

Portability, Testing, and Safety

• Run fully on the client (offline after initial load)
• Work across Chrome/Firefox/Safari and major OSes
• Include an automated test suite (e.g., Selenium/Playwright)

Typical End-to-End Workflow

1. Launch app in browser (no install, just static web assets)
2. Upload CT scan (DICOM file or series)
3. Choose tissue type: bone/skin/muscle
4. Preview and adjust:
   • See a 3D rendering of the segmented anatomy
   • Adjust thresholds/modes until satisfied
5. Export:
   • Choose output format: .stl or .gcode
   • App generates the file and triggers a browser download
6. Use externally:
   • .stl → slicer → standard 3D printer
   • .gcode → printer or slicer, for density-mimicking prints