Mathieu Dubois
I have noticed that some series ID are present in MRIMPRANK but not in IDA and vice-versa.
For images that are in IDA but not in MRIMPRANK, does it mean that those images have a low quality (and therefore could be ignored)?
I think I can ignore images that are in MRIMPRANK but not in IDA but how can that happen?
Thanks in advance,
Mathieu
Danielle J Harvey
The IDA should only include scans that passed QC and were released from quarantine – MRIMPRANK should be a complete record of the ranking of all ADNI-1 scans that went to Mayo – are the visits of the scans that you are seeing in the IDA but not in MRIMPRANK after m36 or are they within ADNI-1? Are they all ADNI-1 participants (RID<2000)?
Danielle
--
You received this message because you are subscribed to the Google Groups "Alzheimer's Disease Neuroimaging Initiative (ADNI) Data" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
adni-data+...@googlegroups.com.
To post to this group, send email to
adni...@googlegroups.com.
Visit this group at http://groups.google.com/group/adni-data.
For more options, visit https://groups.google.com/d/optout.
Mathieu Dubois
I only kept images in IDA for which visit is 'ADNI 1' or 'ADNI GO'. I got the series ID from merge with MPRAGEMETA (I remove series ID with only original or with more than one original).
There are 1365 series ID that are in the filtered IDA and not in MRIMPRANK.
The Visit column of the original image with those series ID are distributed like this:
ADNI1 Screening 10
ADNI1/GO Month 12 121
ADNI1/GO Month 18 53
ADNI1/GO Month 24 20
ADNI1/GO Month 36 20
ADNI1/GO Month 48 226
ADNI1/GO Month 6 250
ADNIGO Month 3 MRI 240
ADNIGO Month 60 119
ADNIGO Month 72 10
ADNIGO Screening MRI 296
This covers 382 different RID (235 of which are below 2000).
I guess that the "ADNI1 Screening" correspond to people which were not included in the study.
I haven't found anything particular about the subjects (they are in different protocols, from different phases, etc.).
Mathieu
Danielle J Harvey
MRIMPRANK only includes information from the ADNI-1 visits (which for most folks is through m36, though there do appear to be m48 visits in there, but it may not capture all m48). It will not include anything for RIDs>2000, so no newly recruited ADNI-GO (or ADNI-2) participants will be in the file. For ADNI-1 folks, it won't include anything past m48.
Danielle
Sent: Friday, April 24, 2015 9:41 AM
To: adni...@googlegroups.com
Subject: Re: [adni-data] MRIMPRANK
Maryam Kd
Hi Danielle,
May I ask which table has series id corresponding to ADNIGO or ADNI2?
THhanks,
maryam
Danielle J Harvey
Hi Maryam,
Several files include the variable for series ID across all phases of ADNI (for example, in MPRAGEMETA the variable is SeriesID - the corresponding variable in the IDA_MR_Metadata_Listing file is LONIUID).
Danielle
Sent: Sunday, December 04, 2016 11:44 PM
To: Alzheimer's Disease Neuroimaging Initiative (ADNI) Data
Subject: Re: [adni-data] MRIMPRANK
Maryam Kd
Thanks for the response. Now how we can tell which scan in ADNI2/ADNIGO is the preferred one?
For example we have two LONIUID for
| "116_S_4453" in MPRAGEMETA, how can I find the ranking for them? |
Maryam
Danielle J Harvey
Hi Maryam,
In ADNI-GO and ADNI-2, there we did a non-accelerated scan and an accelerated scan, so that is typically going to explain the 2 scans at a given visit for individuals with an RID>2000. In MPRAGEMETA with the variable Orig/Proc==”Original”, you will see a sequence descriptor MPRAGE for one of them and MPRAGE GRAPPA2 for the other (GRAPPA2 is the accelerated one). You will also see “Accelerated” in the Sequence Descriptors for IR-SPGR or IR-FSPGR. You can use SeriesID to identify the scan and then use ImageUID to identify the ID of either the original or the processed scan.
There shouldn’t be too many cases where there are “duplicates” not explained by accelerated vs non-accelerated (generally speaking, there should only be one processed scan for each accelerated and non-accelerated scan – and that is the preferred one, since Mayo only did the pre-processing on the preferred scan – I do think there are a few that slipped through, so there may be some that have more than one processed scan, but I’m not aware of how we can select the preferred one in those cases).
Danielle
Maryam Kd
Thanks for the response, it helped a lot.
1.I also have question about choosing between GradWarp-N3m and N3m?
Is it always a better choice to choose images which have been corrected for gradient non-linearity distortion?
2. For those that have more than one processed scans(same sequence), can I use all of them as FreeSurfer input or should I pick the one that is added last(bigger UID)?
Thanks,
Maryam
Danielle J Harvey
Hi Maryam,
I believe the MRI Core generally recommended the GradWarp. As for instances where there is more than one processed scan (same sequence), yes, the larger UID is probably the preferred one...
Danielle
Maryam Kd
some of the ADNI1/GO timepoints has 1.5T and 3T, which one is used by LONI as a freesurfer input?
Thanks,
Maryam
Danielle J Harvey
Hi Maryam,
The ADNI-1 study included a 3T MRI arm (so participants in that arm received both a 1.5T and 3T MRI). In ADNI-GO, the protocol (for newly recruited participants) was to scan at 3T (though continuing participants from ADNI-1 were continued at 1.5T unless there was a scanner change at the site). I'd potentially recommend using 1.5T since that was done on everyone in ADNI-1...
The lab that generated summary data using FreeSurfer processed both the 1.5T and 3T scans, so that data are available for both magnet strengths...
Danielle
Maryam Kd
I am a bit confused, you mean, we should use 1.5T for all RID<2000, right? or just for those that have both 1.5T and 3T?
Because I saw some time points like(
| "003_S_0908" |
| "ADNI2 Initial Visit-Cont Pt" |
Maryam
Danielle J Harvey
Hi Maryam,
I guess it depends on your research question. But, you will get more people if you stick to 1.5T in ADNI-1 (only ~25% of ADNI-1 participants received both 1.5T and 3T). If you see an ADNI-1 participant who has primarily 1.5T but then goes to 3T at some point and stays at 3T after that, that would likely be a scanner change. When I look at the person below (003_S_0908), I see 1.5T all the way through ADNI2 Year 1 Visit, but it switches to 3T for ADNI2 Year 2 Visit.