Dear Jeff,
On Tue, 31 Jul 2018 11:10:57 -0400, Jeff Browndyke <[log in to unmask]> wrote:
>Hi, Christian.
>
>An earlier version of CAT12 identified and handled WMH, that was later dropped in new versions of the software. Does the newest version renew this capability? Maybe integration of the LST Toolbox LGA or LPA algorithm with CAT12 segmentation?
Yes, this option is back and largely updated because it was sometimes too aggressive for some data. We now use a larger weighting of the prior probability maps for lesions and some other improvements that should fairly work for most T1 data. However, keep in mind that this is just an alternative if you only have T1 data. In case you additionally have T2/FLAIR data the LST toolbox to fill the lesions is a better choice because the additional FLAIR allows to identify WMHs much better.
We are currently working on an additional integration of FLAIR images to utilize the WMH detection. In the developer mode there is an option for that available. However, this is still work in progress and more or less experimental.
Best,
Christian
>
>Warm regards,
>Jeff
>
>
>> On Jul 31, 2018, at 10:41 AM, Christian Gaser <[log in to unmask]> wrote:
>>
>> Dear Maria,
>>
>> the new version 12.4 of CAT12 can now correctly handle stroke lesions. There is an option for dealing with stroke lesions available if you cann cat12 in expert mode:
>> cat12('expert')
>>
>> This will also work for longitudinal data. The lesions should be set to "0" in your structural images before. If you only have a lesion map which indicates lesions with "1" there also exist a tool to create such a map using the lesion map:
>> SPM->Tools->CAT12->Tools->Manual images (lesion) masking
>>
>> Best,
>>
>> Christian
>>
>> On Thu, 19 Jul 2018 17:16:30 +0100, Maria Jose Torres <[log in to unmask]> wrote:
>>
>>> Dear expert,
>>>
>>> I`m running a VBM analysis in a single stroke patient to explore posibles grey matter changes associated to treatment.
>>> First, I draw the mask on time point 1 T1, then I corregister Time point 2 T1 to Time point 1 T1 and set, in both T1´s, lesion areas to 0 with mricron (draw--> mask image with VOI --> delete regions with VOI).
>>>
>>> After, I used the "segment longitudinal data" option on CAT12. For this I entered the masked images generated previously
>>> ¿This seems correct?
>>>
>>> Now, I`m not sure what statistical analysis should I use. The CAT12 manual gives and example for 1 group x 2 time points analysis and using a Flexible multifactorial analysis with 2 factors (subject and time). Are outcomes expected to be similar if using paired t test or within subject ANOVA (for more time point cases)?
>>>
>>> What would be the correct to use in this case? a paired t test or a flexible factorial analysis?
>>>
>>> When I tried to use the paired t-test the following error come out:
>>>
>>> The following modules did not run:
>>> Failed: Check design orthogonality and homogeneity
>>>
>>> Even though the flexible factorial analysis do not seems adequate to me in this case, I still run it setting 1 factor:time. I believe it does not make sense in this case to include a "subject" factor since I have just 1 subject.
>>> In this case, this error came out:
>>>
>>>
>>> No executable modules, but still unresolved dependencies or incomplete module inputs.
>>> The following modules did not run:
>>> Failed: Factorial design specification
>>> Skipped: Check design orthogonality and homogeneity
>>>
>>> ¿What is the best approach in this case?
>>>
>>> I would highly appreciate any help on this
>>>
>>> Maria Jose
