Dear HCP Team Members,
I’m in the process of revising a manuscript on statistical
modelling of DWI data. In the original version we used
the single subject unprocessed data from the Q1/Q2 Release (2013).
For the revision we have rerun our analysis on the corresponding
data for the same subject (subject ID 100307) in the 900 subjects
release (2015). Except from change in the scale of image intensities
between both datasets (by a factor of approx. 1.5)
we see significant changes in our analysis results, especially
the estimated SNR in the unprocessed data, that indicate qualitative
changes in data properties.
In my understanding, given what I can extract from the
reference manuals and literature, the unprocessed data
should be direct SENSE-1 reconstructions from K-space data,
i.e. the follow a Rice distribution.
- have any data processing steps been applied after SENSE 1 reconstruction
to get the “unprocessed” data ? If yes are there differences between
the Q1/Q2 release and the 900 subjects release ?
- did the SENSE 1 reconstruction parameters change between
the Q1/Q2 release and the 900 subjects release ?
- are there any other changes that can affect statistical properties,
especially SNR in the "unprocessed” data ?
Any comments would be highly appreciated.
Best regards,
Joerg Polzehl
---
Dr. Joerg Polzehl
WIAS, Mohrenstr. 39
D-10117 Berlin, Germany
email: [email protected]
Tel. +49 30/20372 481
Fax: +49 30/20372-303
Web: http://www.wias-berlin.de/people/polzehl/
_______________________________________________
HCP-Users mailing list
[email protected]
http://lists.humanconnectome.org/mailman/listinfo/hcp-users
