VISCODE for UCSFFSX51_11_08_19.csv

[Nalani Knight]

Good afternoon,

I'm comparing the UCSF cross-sectional Freesurfer MRI data sets UCSFFSX_11_02_15.csv and UCSFFSX51_11_08_19.csv. I understand that the former contains info on ADNI1, and the latter contains ADNIGO and ADNI2 data. However, I'm trying to understand the differences in the VISCODE in order to compare baseline values. For the ADNI 1 set, we have viscodes: (blank), "bl", "f", "m06",...,"m72", and "sc". For ADNIGO, the VISCODE is labeled "nv" only, while for ADNI2 we have "v02", "v04",...,"v51". My understanding is that a combination of "bl" and "sc" will yield the baseline/screening values for ADNI1. I was able to locate ADNI2_VISITID to discover that "v02" is the screening/baseline equivalent for ADNI2. Also, "v03" is month 3, and so on. 

However, I'm unable to find the meaning of "nv" for ADNIGO. How may I separate out the baseline/screening data for ADNIGO? I checked the USCF Freesurfer pdf as well as the data dictionaries for these data sets, and none offer an explanation for "nv". 

Thanks all! 

Nalani

[Danielle J Harvey]

Hi Nalani,

 

For the ADNI-1 data, VISCODE is, for the most part, the best thing to use, although pay attention to EXAMDATE, since “sc” should be the initial scan. Looking at some of the entries with VISCODE=”bl”, the EXAMDATES are much later, so these are likely visits that rolled over into ADNI-GO/2.

 

For the ADNI-GO/2 file, use VISCODE2 for more interpretable visit codes. “scmri” should be the VISCODE2 for the initial scan (for newly enrolled participants in ADNI-GO/2). “nv” usually refers to a scan in which the EXAMDATE does not match information that we have about visit dates or metadata for a scan being done, so it isn’t assigned a specific visit.

 

One thing to keep in mind, since I’m not sure how you will be using these data. ADNI-1 scans were mainly 1.5T (a subset also had 3T, but those data should be in a different file). ADNI-1 participants that rolled over into ADNI-GO/2 may have stayed on the 1.5T machine, but some may have switched to 3T. All newly enrolled participants in ADNI-GO/2 were scanned at 3T (and there’s non-accelerated and accelerated acquisitions). Generally, there are differences between measures derived from scans at 1.5T vs at 3T, so it is typically not recommended to combine these.

 

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 view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/707e2186-5ae6-4c7a-b01f-e5fbfc4b0c2a%40googlegroups.com.

[Nalani Knight]

Dear Danielle,

Thanks for your quick reply! 

Also, thank you for pointing out the EXAM-DATE, I may have overlooked this for the 17 "bl" values in the ADNI1 data. Since all of the 679 ADNIGO data (in UCSFFSX51_11_08_19) are assigned VISCODE="nv", does this mean I can not reliably use the ADNIGO MRI cross-sectional Freesurfer data in an analysis for baseline values? I don't want to include later data collection points in with my baseline analysis.

I'm glad you mentioned the scanner info, because I remember reading about this but it had slipped my mind. There is a variable FLDSTRENG for the ADNI1 data in which all are linked to the 1.5T scanner. However, there is no such variable for the ADNIGO/2 data. So my understanding is that if I eliminate those who carried over into ADNIGO/2 from ADNI1, it is safe to assume the remaining ADNI2 RIDs (2000<=RID<4000) used the 3T scanners. Is this correct? 

I appreciate your help!

Nalani

To unsubscribe from this group and stop receiving emails from it, send an email to adni...@googlegroups.com.

[Nalani Knight]

P.S. I have made note of the variable IMAGETYPE in the ADNIGO/2 data. You're also saying that accelerated and non-accelerated T1 should be separated out in an analysis as well? I'll do some more digging to understand what accelerated vs non-accelerated means. Thanks!

[Danielle J Harvey]

Hi Nalani,

 

Sorry for the confusion – I was referring to “nv” for VISCODE2 (there are only a few). If VISCODE2=”scmri”, that should be fine to use. For ADNI-GO/2, I actually ignore VISCODE, since I don’t know what all those codes mean – VISCODE2 was a variable created to help users interpret the tracking VISCODEs (such as v02 and v04).

 

I wouldn’t include accelerated and non-accelerated in the same analysis – we had added both acquisitions to see the difference in a shorter acquisition (accelerated) versus more standard acquisition (non-accelerated). Everyone got both in ADNI-GO/2 (though both may not always be available due to quality control issues).

To unsubscribe from this group and stop receiving emails from it, send an email to adni-data+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/d8b45d40-25ca-4f11-88b8-a6455617c3cf%40googlegroups.com.