Note that the false negative data is much lower (eg. the one with only 2 percent false negative rate). Again, it is important to note that EEG is another detector of microsleep just like CRAM and does not necessarily mean that it is correct. So these figures must be weighed based on their merits.