I had a question about the newly added MSLT data in the WSC dataset. I was hoping to use the MSLT data for a classifier but I found that the data for all 4-5 naps comes in one .edf data file. This is a problem for me because I want to separate out the data from each nap. The start times of each nap are given in the dataset .csv file but not end times and there doesn't seem to be any other variables that can help.
The closest I got was adding up sleep latencies. For example, I found one file with a total duration of 107.5 minutes. The start time according to the .edf file and what was recorded in the datasets .csv was different by ~1.5 minutes so I took that into account. By definition, the naps (in clinical MSLT's) are supposed to last for 20 minutes or 15 minutes after the first epoch of sleep. If I add up the sleep latencies + 15 minutes (unless it's a 20 minute no-sleep nap) and everything else I get 95.5; off by 12 minutes. Maybe since the reported sleep latencies are the average of 2 scorers, it could be that the 15 minutes cutoff was set to one of the two and thus the extra 12 minutes maybe could come from that? In any case it doesn't matter because I was hoping for (and in order to use this data, need) precise indicies to segment the data.
I also tried plotting and visually inspecting some signals for clues, combing through the dataset variables, checking all the attributes of the .edf file object, checking the forum on this site for related posts, etc. and I have yet to find an answer.
Has anyone else worked with the MSLT data? Am I missing something? Was the data that could segment the signals not retained or was it mistakenly not included? Something else? Much appreciation to anyone who might be able to point me in the right direction.
Hey Snowman,
I heard back from WSC today - they have log files that provide information about lights out/on for each of the naps. I think these will provide the "precise indices" you need to segment the data as you discussed.
The WSC team offered to prepare these log files for the MSLTs posted on the site, however it will take them some time (at least a few months) to fit it into their schedule. I will post here again when I have further updates.