Two questions about the sample selection of ABIDE I dataset

[Erick Zhou]

Hi, everybody.

 

I am a new user of ABIDE I dataset from Tulane University.. I have a few questions about the sample selection. Would you please give me some suggestions?

 According to the paper  namely “The autism brain imaging data exchange: towards a large-scale evaluation of the intrinsic brain architecture in autism”, the selected sample should meet the requirements of  “each subject had an anatomical image” and “individuals with mean framewise displacement (FD) < 2 s.d. above the sample mean (0.23±0.27 mm)”. But I don't know the meaning of these two requirements.

 

1.      “each subject had an anatomical image”. How to confirm whether a subject data including an anatomical image? I have not found the information about the anatomical image in the ABIDE website. 

2.      “individuals with mean framewise displacement (FD) < 2 s.d. above the sample mean (0.23±0.27 mm)”. I have tried to calculated the mean and std of “func_mean_fd” in Phenotypic_V1_0b_preprocessed1.csv(the file is enclosed as the attachment). But I got the mean = 0.131, and the std=0.158. Can you tell me how to get the mean and std (0.23±0.27 mm)?

 

Would you please give me some suggestions at your convenient time? Thank you.

[Franklin Feingold]

Hi Erick, 

An example anatomical image can be a T1 weighted scan. In BIDS they are identified with these anatomical modality suffices - https://bids-specification.readthedocs.io/en/stable/04-modality-specific-files/01-magnetic-resonance-imaging-data.html#anatomy-imaging-data . 

The preprocessed ABIDE dataset also has a GitHub page - https://github.com/preprocessed-connectomes-project/abide . Perhaps that can help too

Thank you,

Franklin

[Erick Zhou]

Thank you

在 2020年7月14日星期二 UTC+8上午5:50:19，Franklin Feingold写道：