DTI Analysis

Diffusion Tensor Imaging with Riemannian geometry. Curvature-based anisotropy (CA) complements standard FA/MD for white matter analysis.

📜 Patent Filed (#64/011,831)

Geometric features: 76.4% vs 69.2% raw tensors (+7.2pp). Benchmark: DOI 10.5281/zenodo.19192330

Why Geometric DTI?

The Clinical Problem

Diffusion Tensor Imaging (DTI) maps water diffusion in brain tissue to reveal white matter structure. It's critical for diagnosing multiple sclerosis, stroke, traumatic brain injury, and neurodegenerative diseases.

Standard DTI metrics (FA, MD) treat 3×3 diffusion tensors as flat vectors. This creates artifacts: arithmetic averaging produces eigenvalue swelling, linear interpolation can violate positive-definiteness, and standard metrics miss geometric tissue signatures.

The Geometric Approach

Diffusion tensors are symmetric positive-definite matrices living on GL+(3)/SO(3). The affine-invariant (geodesic) distance naturally respects this constraint. Geodesic interpolation preserves positive-definiteness at every intermediate point.

The novel curvature anisotropy (CA) metric uses Ricci curvature to weight shape sensitivity — detecting tissue changes in the deviatoric (fiber direction) sector that FA and MD cannot distinguish.

Clinical Applications

Geometric DTI is especially relevant for:

  • Multiple Sclerosis: Demyelination changes fiber orientation (V+) before affecting volume (V)
  • Traumatic Brain Injury: Diffuse axonal injury alters tract geometry subtly
  • Stroke Recovery: Monitoring white matter reorganization during rehabilitation
  • Neurodegenerative Disease: Early detection of Alzheimer's via tract integrity changes

Why Riemannian DTI?

📐

Affine-Invariant

Geodesic distances on GL+(3)/SO(3) respect the positive-definite constraint. No Euclidean artifacts like swelling.

🌀

Curvature Anisotropy

CA uses Ricci curvature to weight shape sensitivity — detecting tissue changes invisible to FA alone.

V+/V- Decomposition

Separates deviatoric (shape, 5D) from volumetric (size, 1D) using the DeWitt metric decomposition.

Tissue Types

Automatic classification from diffusion tensor eigenvalues.

White Matter

FA = 0.82

λ = [1.7, 0.3, 0.3] × 10-3

Highly anisotropic — coherent axonal bundles

Gray Matter

FA = 0.22

λ = [1.0, 0.7, 0.7] × 10-3

Mildly anisotropic — cell bodies

CSF

FA = 0.00

λ = [3.0, 3.0, 3.0] × 10-3

Isotropic — free water diffusion

Diffusion Tensor Ellipsoids

2D cross-sections showing tissue anisotropy and the Euclidean swelling artifact.

Diffusion Tensor Interpolation

Riemannian vs Euclidean: why geodesic interpolation eliminates the swelling artifact in DTI tractography.

0° parallel90° perpendicular
Low (isotropic)High (anisotropic)
Midpoint FA
0.00
Euclidean
vs
0.82
Riemannian
Determinant Ratio (mid / endpoint)
1.00
Euclidean
vs
1.00
Riemannian
Max Eigenvalue Ratio (mid / endpoint)
1.00
Euclidean
vs
1.00
Riemannian

DTI Features

API Playground

REST API