Lie detector systems (polygraphs) generally work by looking at the physiological responses to the own subjects baseline, and do this in real-time. You could use a file that represents the stored "baseline" for the subject, but the overall environment, circumstances, general state of the subject, etc. may not be similar enough to easily see subtle changes.
From the stored file, you might determine a set of parameters or thresholds that indicate "normal, relaxed" for each signal. You could read the file into your LabVIEW program, process each channel, and save these parameters. Then when you have the new data, you can read the new data file in and process it in a similar way and then determine whether the parameters are outside the limits that you established for a "lie".
You may get more responses to this question if posted in the Biomedical User Group
