The DX doesn't correspond to the MRI group

[avid...@gmail.com]

Hello,

First of all, thank you so much for providing such an extensive data. So my question was that when I found the mri images corresponding to the DX changes from the tadpoole_D1_D2 data, it still shows it to be a part of the same group. But according to the DX's given in the data, it corresponds to a patient's diagnosis being changed to "MCI to dementia" or "NL to MCI".

As an example: the patientId's corresponding to those discrepancies are: 011_S_0022(NL to MCI) and 018_S_0057(MCI to Dementia

So is there some discrepancy with the labels in the MRI group or is it in the study data? I am unsure. Any help will be greatly appreciated thanks!

[Neil Oxtoby]

In the ADNI IDA, the "Research Group" corresponds to DX_bl (baseline diagnosis), not DX (current diagnosis).

This is how the situation you saw can arise:

PTID,       VISCODE, DX,              Research Group,  IMAGEUID_UCSFFSX_11_02_15_UCSFFSX51_08_01_16
018_S_0057, m18,     MCI to Dementia, MCI,             74240    
011_S_0022, m24,     NL to MCI,       CN,              81357

Cheers

[avid...@gmail.com]

Ah okay, thank you so much. Just one more question, whether does the DX(current diagnosis) come from? Are there specific data files from ADNI that I can get it from? Thanks once again. 

[Razvan Marinescu]

You can find the diagnosis and related information under Assessments -> Diagnosis -> Diagnostic Summary. They are provided as codes (1-9), and you need to convert them to the right label.

Raz

[Neil Oxtoby]

I recommend checking out the ADNI training slides, part 2:

https://adni.loni.usc.edu/wp-content/uploads/2012/08/slide_data_training_part2_reduced-size.pdf

And for a more basic introduction, part 1:

https://adni.loni.usc.edu/wp-content/uploads/2012/08/ADNI_data_training_slides_part1.pdf

[avid...@gmail.com]

Great, thank you so much! 

[avid...@gmail.com]

I will go through those. Thank you for the help Neil.