NeuroView

A web-based 3D brain MRI viewer that converts NIfTI brain scans into interactive 3D models with adjustable depth visualization.

Features

• 3D Brain Visualization - Upload brain MRI scans and view them as interactive 3D models
• Depth Control - Adjust threshold to explore from outer cortex to deep internal structures
• Transparency - Control opacity to see through layers
• MRI Slices - View Axial, Sagittal, and Coronal slices with interactive sliders
• Slice Plane Overlay - Toggle slice images directly on the 3D model
• Wireframe Mode - Toggle mesh wireframe visualization with adjustable detail
• Auto-Rotate - Smooth automatic rotation for presentation
• Color Themes - Multiple professional color schemes
• Desktop App - Available as standalone application for Mac and Windows

Requirements

Backend

• Python 3.8+
• FastAPI
• nibabel
• numpy
• scipy
• scikit-image
• Pillow
• uvicorn

Frontend

• Modern web browser with WebGL support

Installation

1. Clone the repository:

git clone https://github.com/YOUR_USERNAME/neuroview.git
cd neuroview

2. Install Python dependencies:

pip install fastapi uvicorn nibabel numpy scipy scikit-image pillow

3. Start the backend server:

cd backend
python server.py

4. Open brain_viewer.html in your browser

Usage

1. Start the server (runs on http://localhost:8000)
2. Open brain_viewer.html in your browser
3. Upload a NIfTI brain scan (.nii or .nii.gz)
4. Use the controls to explore:
   • Depth (Threshold): Lower = outer surface, Higher = internal structures
   • Opacity: Make the brain transparent
   • Wireframe: See the mesh structure
   • Slice sliders: Navigate through MRI slices

Supported File Formats

• NIfTI (.nii)
• Compressed NIfTI (.nii.gz)

Getting Sample Brain Data

You can download skull-stripped brain MRI data using nilearn:

from nilearn import datasets
import nibabel as nib

mni = datasets.fetch_icbm152_2009()
t1_img = nib.load(mni['t1'])
mask_img = nib.load(mni['mask'])

# Apply mask for brain-only data
brain_only = t1_img.get_fdata() * mask_img.get_fdata()
brain_img = nib.Nifti1Image(brain_only, t1_img.affine)
nib.save(brain_img, 'brain_only.nii.gz')

Desktop Application

NeuroView can be built as a standalone desktop application for Mac and Windows.

Build for macOS

./build_mac.sh

This creates dist/NeuroView.app - a native macOS application.

Build for Windows

build_windows.bat

This creates dist/NeuroView/NeuroView.exe - a Windows executable.

Build Requirements

• Python 3.8+
• pywebview
• pyinstaller

All build dependencies are included in requirements.txt.

Tech Stack

• Frontend: HTML5, CSS3, JavaScript, Three.js
• Backend: Python, FastAPI, scikit-image (marching cubes)
• 3D Rendering: WebGL via Three.js
• Desktop: pywebview, PyInstaller

License

MIT License