Radial q-space sampling for DSI.

Area of Science:

• Neuroimaging
• Biophysics
• Medical Physics

Background:

• Diffusion spectrum imaging (DSI) is crucial for noninvasively visualizing brain microstructure.
• Current DSI methods utilize a rectangular grid for sampling the diffusion encoding space.
• This limits the angular resolution and accuracy of reconstructed orientation distribution functions.

Purpose of the Study:

• To introduce a novel radially symmetric q-space sampling scheme for DSI.
• To enhance the angular resolution and accuracy of reconstructed orientation distribution functions.
• To present an alternative to traditional Cartesian grid-based DSI implementations.

Main Methods:

• Q-space sampling was performed along radial lines.
• The Fourier slice theorem was used to analytically connect radial q-space samples to orientation distribution functions.
• Computer simulations and in vivo human brain data were utilized for validation.

Main Results:

• Radial diffusion spectrum imaging accurately estimates orientation distribution functions.
• Effective estimation was achieved with high b-values (4000 s/mm2) and 236 q-space samples.
• Radial reconstruction minimizes truncation artifacts common in Cartesian methods.

Conclusions:

• Radial DSI offers improved angular resolution, dependent on the number of radial lines.
• The method shows promise for more accurate DSI reconstructions.
• Radial q-space acquisition is a favorable approach for DSI applications.