Measurement Spectral Edge following FFT

[Babmi]

For EEG studies, I need to measure and determine in real time the frequency below which a given percent (e.g., 90%) of the power resides.  After FFT analysis, what to do?  Thanks for any help. 

[DjDaveNI]

Hello Babmi,

Can you elaborate more on what/how you plan to do? Are you using a LabVIEW Real-Time system? If so, which hardware and software do you have? If you have already developed a Real-Time VI and need help with that, please take a screenshot and post it so everyone can better understand your issue.

Just to clarify: Real-Time operating systems are deterministic, not instant. They allow the timed execution of a loop within a finite amount of time and minimal jitter. What acquisition rate are you looking for? What kind of data are you performing a FFT on?

Please post back with more specific details about your system, and exactly what you would like to do. Thanks 🙂

[Babmi]

I am running labView 8.5.1 on an Intel Mac with data acqusition through an USB-6009.    Data originating from a rat EEG (ie, "brain waves") is an analog signal from a Grass Instruments amplifier, band pass filtered roughly between 0.5 and 30 Hz; acquistion rate will be 128 Hz or faster.  My plan is/was to analyze, possibly, individual 4 sec epochs to determine spectral edge frequency at baseline and following/during application and titration of an anesthetic drug.  Anesthetics cause the EEG to shift to slower frequencies and this measurement is used to quantify anesthetic depth.  Thanks. 

[DjDaveNI]

Hey Babmi,

These tutorials may help you in the development of your application. I also found this article on FFT analysis, as well as the help file for FFT in LabVIEW.

Still not sure what you are having trouble with… could you post a screenshot of your code so far? What exactly is your question?

[Babmi]

My question was how to use the FFT ouput to determine the spectral edge, ie, the frequency below which a 90% of the power resides.  I was hoping there was a vi for this.  I think I figured it out, however.  I summed up the all power values in the array provided by the Express FFT.vi (Add Array Elements), and then summing up increasing subsets of the array (using Array Subset and Add Array Elements) in a While Loop (using >= for the Stop if True, ie, greater or = to 0.9), I was able to compute the frequency below which 90% of the power resides.  Does this sound reasonable?  Thank you for pointing out the FFT information. 

[Babmi]

Here's my vi.