You can also do robust tensor fitting, such as RESTORE, which removes outlier data points on a voxel by voxel basis. This is implemented in TORTOISE and Camino, to my knowledge. Citation is: Chang, L.C., Jones, D.K., and Pierpaoli, C. (2005) RESTORE: Robust estimation of tensors by outlier rejection. Magn Reson Med. 53:1088-1095.
Lindsay.
________________________________________
From: Lythgoe, David [[log in to unmask]]
Sent: January 17, 2012 10:00 AM
To: [log in to unmask]
Subject: Re: [FSL] Spiking Artifacts in DTI-data
>Dear FSLexperts and -users,
>
>I would like to ask what your experience with Spiking-artifacts in
>DTI-data is. I have looked every single DICOM-image of my data (more than
>60 subjects) and there are unfortunately some spikes, which is of course
>no surprise with DTI, especially with Siemens Trio.
>
>What is the common way to deal with them? to ignore/ to erase single
>images / to erase the whole volume which has a single Artifact-image??
>
>Your help and comments are very much appreciated!
If you're a programmer, or have one to hand, take a look at Chavez, Storey
& Graham, Robust correction of spike noise: Application to diffusion
tensor imaging. Magn. Reson. Med., 2009, 62, 510-519. I always meant to
try it but never got round to it, so can't comment on its effectiveness.
Dave
>
>Best,
>
>Elif
|