A C-language program for the computation of power spectra on a laboratory microcomputer

Area of Science:

• * Computational Neuroscience
• * Biomedical Engineering
• * Signal Processing

Background:

• * Time-domain data analysis often requires specialized tools for spectral analysis.
• * Detecting high-frequency oscillations in physiological signals is crucial for understanding neural pathways.
• * Existing methods may lack portability or efficiency for complex datasets.

Purpose of the Study:

• * To develop and describe a portable computer program for power-spectral analysis.
• * To apply the program for identifying high-frequency oscillations in the phrenic neurogram.
• * To demonstrate the program's capability in revealing spectral characteristics of neural activity.

Main Methods:

• * Implementation of a fast Fourier transform (FFT) algorithm in C language.
• * Processing sequential segments of time-domain data.
• * Ensemble averaging of frequency-domain transformed data to generate smoothed power spectra.

Main Results:

• * The C program successfully performed power-spectral analyses on time-domain data.
• * Application to cat phrenic neurograms revealed high-frequency oscillations.
• * 100 successive 512-point FFTs accurately identified the power and frequency of multiple peaks.

Conclusions:

• * The developed C program provides an efficient and transportable method for power-spectral analysis.
• * It is effective in detecting and characterizing high-frequency oscillations in physiological signals.
• * The program's portability ensures broad applicability across different computing environments.