Here we present blind source separation (BSS) as a new tool to analyse multi-echo diffusion data. This technique is designed to separate mixed signals and is widely used in audio and image processing. Interestingly, when it is applied to diffusion MRI, we obtain the diffusion signal from each water compartment, what makes BSS optimal for partial volume effects correction. Besides, tissue characteristic parameters are also estimated. Here, we first state the theoretical framework; second, we optimise the acquisition protocol; third, we validate the method with a two compartments phantom; and finally, show an in-vivo application of partial volume correction.     «

Here we present blind source separation (BSS) as a new tool to analyse multi-echo diffusion data. This technique is designed to separate mixed signals and is widely used in audio and image processing. Interestingly, when it is applied to diffusion MRI, we obtain the diffusion signal from each water compartment, what makes BSS optimal for partial volume effects correction. Besides, tissue characteristic parameters are also estimated. Here, we first state the theoretical framework; second, we opt...     »