How to deal with the data set of different trial numbers

1 回表示 (過去 30 日間)
Takara Sumoto
Takara Sumoto 2021 年 2 月 5 日
回答済み: Walter Roberson 2021 年 2 月 5 日
I'm conducting a psychological experiment using EEG. I'm analyzing EEG data with Psychotoolbox on MATLAB (not BrainVIsionAnalyzer).
Each experiments consist of 60 trials but there is only one subject whose last 2 trials couldn't recorded so it has only 58 trials.
When I did preprocessing, it shows error and stop preprocessing.
How should I deal with this data set which has differnt trial numbers?

採用された回答

Walter Roberson
Walter Roberson 2021 年 2 月 5 日
As discussed https://www.mathworks.com/matlabcentral/answers/725387-how-to-deal-with-the-missing-information-on-vmrk-file?s_tid=srchtitle the file format used does not permit you to define a file marker as being present in some indefinite place on order but simultaneously marked as missing in a way that the analysis program would be likely to be able to work with.
As discussed there, you could try putting in definite locations but marked "bad interval". If your data file is short you might need to pad the data file. Unfortunately chances are that your analysis program will get confused or will error with that kind of data.
So what can you do instead? This:
Fix the analysis program so that it can deal sensibly with a marker not being present. Ideally even a missing marker that is not at the end.
What should you do in the analysis of such cases? Hard to say without knowing what calculations are being used. It can turn out to depend upon which data is missing: sometimes you can resynchronize enough and sometimes you can't.

その他の回答 (0 件)

カテゴリ

Help Center および File ExchangeEEG/MEG/ECoG についてさらに検索

製品


リリース

R2019a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by