Spherical Harmonics (SH) are functions defined on the sphere. A collection of SH can used as a basis function to represent and reconstruct any function on the surface of a unit sphere.
Spherical harmonics are ortho-normal functions defined by:
where \(l\) is the band index, \(m\) is the order, \(P_l^m\) is an associated \(l\)-th degree, \(m\)-th order Legendre polynomial, and \((\theta, \phi)\) is the representation of the direction vector in the spherical coordinate.
A function \(f(\theta, \phi)\) can be represented using a spherical harmonics basis using the spherical harmonics coefficients \(a_l^m\), which can be computed using the expression:
Once the coefficients are computed, the function \(f(\theta, \phi)\) can be approximately computed as:
In HARDI, the Orientation Distribution Function (ODF) is a function on the sphere.
Several Spherical Harmonics bases have been proposed in the diffusion imaging
literature for the computation of the ODF. DIPY implements two of these in the
shm module tool set:
In both cases, \(\Re\) denotes the real part of the spherical harmonic basis, and \(\Im\) denotes the imaginary part.
In practice, a maximum even order \(k\) is chosen such that \(k \leq l\). The choice of an even order is motivated by the symmetry of the diffusion process around the origin.
|||(1, 2) Descoteaux, M., Angelino, E., Fitzgibbons, S. and Deriche, R. Regularized, Fast, and Robust Analytical Q‐ball Imaging. Magn. Reson. Med. 2007;58:497-510.|
|||(1, 2) Tournier J.D., Calamante F. and Connelly A. Robust determination of the fibre orientation distribution in diffusion MRI: Non-negativity constrained super-resolved spherical deconvolution. NeuroImage. 2007;35(4):1459–1472.|