A Sobolev Norm Based Distance Measure for HARDI Clustering - A Feasibility Study on Phantom and Real Data

Ellen J. L. Brunenberg, Remco Duits, Bart M. ter Haar Romeny, Bram Platel

Abstract

Clustering in High Angular Resolution Diffusion Imaging (HARDI) is essential for segmenting complex brain structures. Traditional methods relying on the $L_2$ norm often fail to distinguish between voxels with similar diffusion amplitudes but different fiber orientations. This paper presents a novel dissimilarity measure based on the Sobolev norm. By weighting higher-order spherical harmonics, the Sobolev distance acts as a peak detector, significantly improving sensitivity to fiber geometry. Experimental results on synthetic phantoms and real human thalamic data demonstrate that the Sobolev norm outperforms the $L_2$ norm in both classification accuracy and anatomical segmentation.

Introduction

Diffusion MRI clustering is a primary tool for segmenting brain gray matter nuclei. While Diffusion Tensor Imaging (DTI) is widely used, its Gaussian assumption fails in regions with "crossing fibers." HARDI addresses this by estimating a complex Orientation Distribution Function (ODF). However, standard comparison metrics like the $L_2$ norm (Euclidean distance) are primarily sensitive to the overall amplitude of the ODF rather than the specific directions of its peaks. Consequently, voxels with distinct fiber architectures may be incorrectly clustered together if their total diffusion intensities are similar.

Methodology: The Sobolev Distance

The proposed approach utilizes a Sobolev norm-based distance to capture the geometric features of diffusion profiles. The distance is mathematically formulated to include a "blob-sensitive" component that rewards the alignment of extrema (fiber orientations).

To achieve computational efficiency, ODFs are represented using Spherical Harmonics (SH). In the SH domain, the Sobolev distance is calculated as a weighted $\ell_2$ inner product of the coefficients. The defining feature is the multiplier $\tilde{m}_l$, which applies increasing weights to higher-order harmonics ($l$).

The Sobolev norm weights higher-order spherical harmonics (left) to emphasize the fine angular details of the SH spectrum (middle), resulting in a weighted spectrum (right) that is more sensitive to fiber orientation.

By tuning a parameter $\gamma$, researchers can balance the influence of diffusion intensity versus geometric shape. This makes the Sobolev norm a "geometric" measure, whereas the $L_2$ norm remains an "intensity" measure.

Experimental Setup

The feasibility of the Sobolev measure was tested on two datasets:

  1. Synthetic HARDI Phantom: A multi-tensor model with 121 directions and Rician noise (SNR 30), featuring 18 distinct fiber profiles (crossings at various angles).
  2. Real Human Brain Data: 3T MRI (128 directions, b-value 2000 $s/mm^2$) focused on the right thalamus, validated against the Talairach atlas.

Results and Analysis

The Sobolev norm demonstrated superior sensitivity to angular changes compared to the $L_2$ norm. As the angle between crossing fibers increases, the Sobolev distance reaches its maximum more rapidly, providing better contrast for clustering algorithms.

Comparison of normalized distance measures as a function of the angle $\phi$ between fibers. The Sobolev-based distance (solid line) shows higher sensitivity to orientation changes than the standard $L_2$ norm (dashed line).

Quantitatively, the improvements were stark:

  • Classification: On noisy phantom data, 1-Nearest-Neighbor classification accuracy rose from 71.2% ($L_2$) to 100% (Sobolev).
  • Clustering: K-Means clustering accuracy on the phantom improved from 20.7% to 73.7%.
  • Anatomy: In real-world thalamic segmentation, the Sobolev norm successfully identified nuclei (e.g., pulvinar and medial nuclei) that were blurred or merged when using the $L_2$ norm.

Conclusion

The Sobolev norm provides a robust, geometry-aware distance measure for HARDI data. By emphasizing higher-order spherical harmonics, it effectively distinguishes complex fiber configurations that are invisible to standard Euclidean metrics. Future work will focus on integrating this measure into automated clustering frameworks like spectral clustering and enhancing fiber tractography pipelines.

Summary updated

Abstract

Dissimilarity measures for DTI clustering are abundant. However, for HARDI, the \(\mathbb{L}_{2}\) norm has up to now been one of only few practically feasible measures. In this paper we propose a new measure, that not only compares the amplitude of diffusion profiles, but also rewards coincidence of the extrema. We tested this on phantom and real brain data. In both cases, our measure significantly outperformed the \(\mathbb{L}_{2}\) norm.
BibTeX
@inproceedings{brunenberg2010sobolev,
  author = {Ellen J. L. Brunenberg and Remco Duits and Bart M. ter Haar Romeny and Bram Platel},
  title = {{A Sobolev Norm Based Distance Measure for HARDI Clustering - A Feasibility Study on Phantom and Real Data}},
  booktitle = {International Conference on Medical Image Computing and Computer-Assisted Intervention},
  year = {2010},
  doi = {10.1007/978-3-642-15705-9\_22},
}
Figure