Skip to main content
. Author manuscript; available in PMC: 2023 Jan 16.
Published in final edited form as: Neuroimage. 2022 Mar 23;254:119138. doi: 10.1016/j.neuroimage.2022.119138

Table 1.

Summary of methods and reliability comparisons.

Method Linear Least Squares
(LLS)
Single Tensor
Bayesian (STB)
FSL’s BedpostX
(BPX)
Bayesian
Multi-Tensor
Model-Selection
(BaMM)
DSI’ Constant Solid
Angle Q-Ball
Imaging (CSA-QBI)
MRtrix3’s
Constrained
Spherical
Deconvolution (CSD)
Model Single tensor Single tensor with Biological Priors Sum isotropic and N fiber compartments, automatic relevance detection Sum isotropic and N fiber compartments, Bayesian model selection Diffusion orientation distribution function (dODF) with constant solid angle Fiber orientation distribution (FOD) through constrained spherical deconvolution
Simulated Single Tensor SNR 30, 50, 100 SNR 30, 50, 100 SNR 30, 50, 100 SNR 30, 50, 100 SNR 30, 50, 100 SNR 30, 50, 100
Simulated Two Crossing Tensors SNR 30, 50, 100
Fraction 50/50, 60/40, 80/20%
FA 0.6, 0.8
Angle 30°, 60°, 90°
SNR 30, 50, 100
Fraction 50/50, 60/40, 80/20%
FA 0.6, 0.8
Angle 30°, 60°, 90°
SNR 30, 50, 100
Fraction 50/50, 60/40, 80/20%
FA 0.6, 0.8
Angle 30°, 60°, 90°
SNR 30, 50, 100
Fraction 50/50, 60/40, 80/20%
FA 0.6, 0.8
Angle 30°, 60°, 90°
Simulated Three Crossing Tensors SNR 30, 50, 100
Fraction 33/33/33, 40/34/26, 53/34/13%
FA 0.6, 0.7, 0.8
Angle 30°, 60°, 90°
SNR 30, 50, 100
Fraction 33/33/33, 40/34/26, 53/34/13%
FA 0.6, 0.7, 0.8
Angle 30°, 60°, 90°
SNR 30, 50, 100
Fraction 33/33/33, 40/34/26, 53/34/13%
FA 0.6, 0.7, 0.8
Angle 30°, 60°, 90°
SNR 30, 50, 100
Fraction 33/33/33, 40/34/26, 53/34/13%
FA 0.6, 0.7, 0.8
Angle 30°, 60°, 90°
Human Brain Reliability Map All three subjects
Reliability Curve on Specific ROI All three subjects All three subjects All three subjects All three subjects All three subjects All three subjects