However, the results of model-based methods are limited by the model, and it is common that the diffusion pattern does not follows the assumption.A complicated model may also have overfitting problem.This categorization is similar to the classification of the parametric and non-parametric methods in statistics.The parametric or model-based approaches assume a known distribution/model (e.g.The down side of model-free methods is that they often need more diffusion samplings, at least 60 to get a more robust estimation (In comparison, DTI only needs 6 sampling in addition to b0).DSI and QBI offers only a numerical estimation of the diffusion ODF while GQI offers a direct analytical relation for the diffusion ODF (termed SDF here).This derived approach is called q-space diffeomorphic reconstruction (QSDR), a method that reconstructs GQI diffusion pattern directly in the MNI space.This makes group comparison and regression studies much easier.
The calculation does not require complicated optimization or fitting and thus is less affected by outliers in comparison with model-based method.
In DSI, the numerical estimation includes a Fourier transform, followed by a filter to remove noise and a radial integration with numerical interpolation.
In QBI, the numerical estimation includes re-sample diffusion at the radial directions before to convert diffusion signals to SDF.
Gaussian) to obtain inference, whereas non-parametric or model-free approaches assume no underlying distribution/model and obtain inference using empirical distribution.
Both model-based and model-free methods have their strength and weakness: Model-based methods include DTI, ball-and-sticks model, NODDI as well as more complicated model like CHARM and Ax Caliber.
For example, past studies have used bi-exponential model to fit intra-cellular and extra-cellular There are still other methods which are both model-based and model-free.