Digital Signal Processing

Catalog description

The course is designed for an audience that does digital signal processing (e.g., people who do neuroscience) but that do not have a strong math or engineering background. The goal of the course is that after you have completed it you'll have a fairly sophisticated understanding of how to apply several digital signal processing techniques, including better understanding to what is really happening when you push certain buttons in packaged neuroimaging software (e.g., filter settings). After completing the course you'll also better understand how to collect neuroimaging data (e.g., data sampling rate). Digital Signal Processing contains four sections: Basics, Tutorial, Try It, and Literacy. Part 1: Introduction to sine/cosine functions, discussion of time series and spatial data, discussion of amplitude, frequency, and phase, and a section on adding sine waves. There's also a brief introduction to complex numbers and the Euler Identities. Students also read in time and spatial data (grayscale images). Part 2: Detailed discussion of the Nyquist Theorem and aliasing (time and spatial domain), a section on multiplying sine waves, and a brief discussion of plotting complex numbers and determining the magnitude and phase of complex numbers. Part 3: Convolution, and via convolution, filtering. Ideas are explored in the time domain. In this process, students are introduced to high- and low-pass filters and gain functions. Students use convolution to filter several time domain datasets. Part 4: Generally the same as Chapter 3, but now examining spatial data. Students use convolution methods to filter grayscale and color images. Normal distributions and random noise are also discussed. Part 5: Using sine and cosine to compute the magnitude and phase of activity at different frequencies: time and spatial data. Students also see that magnitude and phase information can be…