Status
Development

Tangible Neuroanatomy

A reproducible workflow converting T1 MRI slices into tactile anatomical models for spatial learning in medical education.

DCM
STL
Blender
AutoCAD
CNC Fabrication

Overview

This project bridges neuroimaging and physical fabrication to create an interactive, haptic learning tool for neuroanatomy. By converting a T1-weighted MRI slice into a 3D-embossed tactile map with conductive landmarks, the system enables game-based learning inspired by Operation, promoting spatial understanding through touch and immediate feedback.

Source data: T1-weighted sagittal MRI from a healthy 20-year-old female (case 37605)

T1-weighted sagittal MRI slice

Workflow Summary

1. DICOM/PNG to 3D Model

You can begin with a DICOM slice or a PNG image. If starting from DICOM, load the study into your prefered medical imaging software (eg, 3D Slicer) and export the desired slice as a PNG. Ensure 16-bit grayscale for maximum depth information.

Source options for free medical imaging:

  • The Cancer Imaging Archive (TCIA)
    Free, de-identified clinical datasets across CT, MRI, PET, and more.
    https://www.cancerimagingarchive.net

  • 3D Slicer Sample Data (QIICR)
    Small, high quality DICOM datasets ideal for testing and teaching.
    https://sampledata.slicer.org

  • Radiopaedia Cases (some with DICOM downloads)
    Not all cases include DICOM, but many do.
    https://radiopaedia.org

T1-weighted sagittal MRI png in blender

PNG to 3D Model (Blender):

  • Import PNG as image texture in Blender
  • Create plane mesh and apply UV mapping
  • Add material with displacement/bump node using PNG as input
  • Apply modifiers to mesh:
    • Subdivision Surface: Level 2 (Catmull-Clark) for smooth topology
    • Displace: Strength 1.5–2.0 mm, texture coordinate set to UV
    • Solidify: 10 mm thickness for structural integrity
  • Export as high-resolution STL for fabrication

2. CNC Fabrication

  • Material: ¾″ maple board (16″ × 16″)
  • Roughing pass: ¼″ flat-end bit, 3 mm stepdown
  • Finishing pass: ⅛″ ball-end bit for fine detail
  • Duration: ~12 hours
  • Post-processing: Sanding up to 220 grint, stain and oil finish

3. Interactive Electronics (In Progress)

  • Identify 32 key anatomical landmarks (e.g., corpus callosum, thalamus, cerebellar vermis)
  • Drill holes and insert 8-gauge copper wire at each landmark
  • Connect to Arduino Uno circuit for stylus-contact detection
  • Provide feedback via piezo buzzer and 16×2 LCD (displays structure name)

Contact

Questions about this workflow? Feel free to reach out.