error with batch_fitting

[Qianlan Chen]

Dear NODDI team,

I am new running NODDI and having some problems fitting my own dataset fitting to the NODDI model. In fact, I was able to successfully run the toolbox with the provided online dataset, and all the results are the same as the output file. However, this was not successful with my own dataset (3T scanner, Siemens). During batch_fitting, I got the following message, with all the output results wrong (see the attached image “_odi.nii”)

identifier: 'optimlib:optimfcnchk:checkfun:NaNFval'
       message: 'Supplied function 'fobj_rician_fix' returned NaN when evaluated;...'
         cause: {[1x1 MException]}
         stack: [6x1 struct]
    Correction: []

  MException with properties:

    identifier: 'optimlib:optimfcnchk:checkfun:NaNFval'
       message: 'Supplied function 'fobj_rician_fix' returned NaN when evaluated;...'
         cause: {[1x1 MException]}
         stack: [6x1 struct]
    Correction: []

Then, I compared the protocol between my dataset with the provided online dataset, found that there was a difference in the “b0_Indices” and the “numZeros” (my dataset protocol is attached below). 

protocol =

  struct with fields:

        pulseseq: 'PGSE'
      schemetype: 'multishellfixedG'
    teststrategy: 'fixed'
       totalmeas: 99
      b0_Indices: [0×1 double]
        numZeros: 0
               M: 13
               N: [7 1 14 10 15 6 3 5 8 9 11 9 1]
          udelta: [0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340]
       usmalldel: [0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340]
              uG: [0.0016 0.0281 0.0281 0.0282 0.0282 0.0283 0.0398 0.0398 0.0399 0.0399 0.0399 0.0400 0.0400]
           delta: [0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 … ]
        smalldel: [0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 0.0340 … ]
               G: [0.0016 0.0016 0.0281 0.0399 0.0282 0.0400 0.0282 0.0398 0.0282 0.0400 0.0281 0.0399 0.0281 0.0399 0.0282 0.0398 0.0282 0.0016 … ]
       grad_dirs: [99×3 double]

Is this the main reason why my dataset does not fit into the NODDI model? Or is there anything I can do to solve this problem?

Any help is appreciated, thanks in advance!

Qianlan

[Hui Zhang]

Hi Qianlan,

I think you found the problem.  Your b=0 images must have close to zero but not exactly zero b-values.  And your DWI volumes of a particular shell have similar b-values but not exactly the same.  What you will need to do is to take your bval file and edit it to look like the bval of my example.  You only have to do this once for any data acquired with the same protocol.  You can reuse this file.  Hope it helps.

Gary

--
You received this message because you are subscribed to the Google Groups "NODDI" group.
To unsubscribe from this group and stop receiving emails from it, send an email to noddi+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/noddi/fe9a34d1-f034-49b7-b94b-5fb1e0740073n%40googlegroups.com.

[陈]

Hi Gary,

Thank you so much for your explanation and your generosity for sharing the NODDI tutorial online. Indeed, as you thought, in our dataset the b-value of non-diffusion-weighted measurements was set very close to 0 but rather than 0 exactly. Besides, I found the solution from the other's discussion thread on your forum. By adding one extra argument "b0threshold" to the protocol command, then it worked fine with my dataset finally.

Thanks again for your prompt reply, it really helped a lot!

All the best,

Qianlan  

You received this message because you are subscribed to a topic in the Google Groups "NODDI" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/noddi/8NAUtX9YQFo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to noddi+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/noddi/CA%2BMh5Qd2hTqQ4KOahtxdVd-hQ6xtYq5pU8P6aH%2BgL1bU%3D0S%2BXg%40mail.gmail.com.

 

[Hui Zhang]

Glad to hear.

Gary

To view this discussion on the web visit https://groups.google.com/d/msgid/noddi/19d62ce1.64.17efa180415.Coremail.chenqianlan0823%40126.com.