ADNI MERGE data set
samara banno
Women Leading in AI steering committee @WLiAI
www.WLiAI.org
Danielle J Harvey
Hi Samara,
In ADNI, we have 2 “initial” visits – the screening visit, which is a subset of information collected to ensure the individual meets all the inclusion/exclusion criteria and then a baseline visit which includes a full neuropsychological battery along with additional information. DX_bl is the diagnosis at the screening visit (note, this diagnosis also dictated the visitation schedule) while DX contains the diagnosis at each visit. For most participants, DX_bl will be the same as DX when VISCODE==”bl”. However, as you’ve noticed, there are some individuals where the diagnosis was modified once the additional information was collected. The Clinical Core has generally recommended using the diagnosis at the baseline visit (rather than the screening visit), since it is based on a more complete picture rather than just the screening instruments.
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/CAHkq%2BeSY5QWX%2BF-qet4xHtCZC9vt1%2BapQ8cHguffmocbwpzkNA%40mail.gmail.com.
Dave E
Danielle J Harvey
Thanks Dave,
Yes, very important information that you have provided. SMC, EMCI and LMCI were added as recruitment groups to expand the normal and MCI groups, though again, clinically, SMC are still CN and EMCI and LMCI are both considered MCI, which is why those categories are not carried forward in the diagnosis at different visits.
In ADNIMERGE, the “sc” and “bl” information was all combined on a single row for each participant (VISCODE==”bl”).
Danielle
To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/ff019056-ab90-4f35-b999-0870d185b4bcn%40googlegroups.com.
samara banno
Statistical machine Learning and AI models
Women Leading in AI steering committee @WLiAI
www.WLiAI.org
To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/BY5PR08MB6230FF2CC3B5B0201A78E8F0C02F9%40BY5PR08MB6230.namprd08.prod.outlook.com.
Danielle J Harvey
Hi Samara,
My apologies for not being clear. DX_bl is the diagnosis at the screening visit. DX, when VISCODE==”bl” in ADNIMERGE, will correspond to the diagnosis at the baseline visit. So the participants that have different DX and DX_bl when VISCODE==”bl” are the ones whose screening diagnosis differed from the baseline diagnosis. The Clinical Core does not consider these disease progressions (or reversions), since they consider the diagnosis at the baseline visit to be the more accurate initial diagnosis (as I said before). Other information that was collected at the screening visit (such as MMSE or CDR or MRI) is combined with information collected at the baseline visit (such as PET or the RAVLT measures) on the row that has VISCODE==”bl”.
You can refer to the ADNI2 procedures manuals available for download (ADNI | Study Documents (usc.edu)) to see the definitions of EMCI and LMCI. I didn’t see anything specific for SMC, but there is information about the different groups on the Study design page of the LONI website (ADNI | Study Design (usc.edu)).
To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/CAHkq%2BeR2rY6QzvmJR%3DZhJfr9PdZ%3DAKKO%2BtRoNKo-_8Q_51gEXQ%40mail.gmail.com.
samara banno
Sent from my Huawei phone
To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/BY5PR08MB623052388BEEDDD24A696062C0309%40BY5PR08MB6230.namprd08.prod.outlook.com.
Danielle J Harvey
Hi Samara,
A diagnosis was needed at the screening visit because there were different visit schedules for the different diagnostic groups – the screening diagnosis ended up determining the visitation schedule and for most individuals the screening diagnosis was the same as for the baseline diagnosis. The baseline diagnosis (so DX when VISCODE==”bl” in ADNIMERGE) is deemed the more accurate reflection of the initial diagnosis, since the screening diagnosis was based on very limited information.
DX at other VISCODEs reflects the diagnosis at future visits, so can be used to determine whether or not there were any progressions (or reversions) in diagnosis.
So, I would say DX_bl is useful primarily to understand why some people have certain visits while others do not (for example, in ADNI-1, only those that were MCI at screening had a m18 visit and individuals that were AD at the screening visit were only supposed to be followed for 2 years).
samara banno
To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/BY5PR08MB6230248FCB5E7DF693C9C216C0309%40BY5PR08MB6230.namprd08.prod.outlook.com.