Hi Frank, Thanks for the catch. We found the problem with 103515 several weeks ago and posted this message to HCP-Users:
From: [email protected] [mailto:[email protected]] On Behalf Of Jennifer Elam Sent: Tuesday, October 13, 2015 4:19 PM To: [email protected] Subject: [HCP-Users] Preprocessed Diffusion Data for 103515 incomplete We have recently discovered that the preprocessed diffusion data for subject 103515, released in Q1 (June 2013) and available since then is incomplete/corrupted and should not be used. This is evidenced by the fact that the preprocessed diffusion package is <400MB, when it should be 1.22GB. It appears that the diffusion preprocessing for that subject skipped an important middle pipeline and that accounts for the missing data. Since 103515 is a Q1 Release subject, the preprocessed diffusion data for this subject has been incomplete since our first Q1 data release and affects all HCP distributions via ConnectomeDB, Connectome in a Box, and Amazon S3. We are reprocessing the subject and will rerelease the diffusion data for 103515 along with the upcoming 900 Subjects +MEG3 Release. At that time, Connectome in a Box and Amazon S3 distributions will be updated with the fix as well. Sorry for any inconvenience this has caused. We have added a note about this subject on our Data Issues and Fixes wiki page (https://wiki.humanconnectome.org/x/EIDDAQ). Best, Jenn For 173132, we posted to the list and put this entry in the Data Issues and Fixes wiki page (https://wiki.humanconnectome.org/x/EIDDAQ): . Diffusion data for subject 173132 corrupted A problem with the early stages of the v3.1 (500 Subjects) preprocessing of diffusion data released for subject 173132 has been found. Diffusion data for this subject downloaded from June 6, 2014-October 31, 2014 (or distributed in Connectome in a Box drives dated 20140805 or earlier) should not be used. Completed fix: Diffusion data for subject 173132 has been completely reprocessed and the fixed version is available in ConnectomeDB as of 10/31/14. These files will need to be replaced for 500 Subjects Connectome in a Box dated 20140805 or earlier. The reprocessed files are included in 500 Subjects Connectome in a Box versions dated 20141203 and later. For our QC process in HCP, we do have an extensive QC process, on the scan-level data especially, which is detailed in Marcus et al. 2013 <http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3845379/> and in the Post Session 1 section of the HCP Reference Manual Standard Operating Procedure Appendix 4 <http://humanconnectome.org/documentation/S500/HCP_S500+MEG2_Release_Appendi x_IV.pdf> . When we do find a problem, we work to fix it and update the database ASAP. If we are preparing for a data release, we typically roll the fix into the updated data in the release, as we will do with the preprocessed diffusion packages for 103515. Best, Jenn Jennifer Elam, Ph.D. Outreach Coordinator, Human Connectome Project Washington University School of Medicine Department of Anatomy and Neurobiology, Box 8108 660 South Euclid Avenue St. Louis, MO 63110 314-362-9387 [email protected] www.humanconnectome.org _______________________________________________ Jennifer Elam, Ph.D. Outreach Coordinator, Human Connectome Project Washington University School of Medicine Department of Anatomy and Neurobiology, Box 8108 660 South Euclid Avenue St. Louis, MO 63110 314-362-9387 [email protected] www.humanconnectome.org -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Fang-Cheng Yeh Sent: Friday, October 30, 2015 10:19 AM To: [email protected] Subject: [HCP-Users] Another corrupted diffusion MRI data set found (Subject #103515) Dear HCP experts, The preproc diffusion data in subject #103515 seems to contain empty values in some DWI acquisitions. It seems that the preproc failed in this subject? (The unproc data look okay) I tried to download and check the preproc data from the connectome db website, but got an error "Server aborted session: No such file". I am not sure whether this empty value problem happens only in my data source (I got the data from connectome in a box). I am also curious about what QC procedure had been applied to the HCP data? Last time I found subject #173132 has corrupted diffusion data. Is there any fix to the data set? This QC issue may become problematic for the next 1200 subject release. Best regards, Frank _______________________________________________ HCP-Users mailing list <mailto:[email protected]> [email protected] <http://lists.humanconnectome.org/mailman/listinfo/hcp-users> http://lists.humanconnectome.org/mailman/listinfo/hcp-users _______________________________________________ HCP-Users mailing list [email protected] http://lists.humanconnectome.org/mailman/listinfo/hcp-users
