|Curricular Unit (UC)||
Biomedical Digital Signal Processing
Course category: B - Basic; C - Core Engineering; E - Specialization; P - Complementary.
|Year:3rd||Semester: 1st||ECTS: 5.5||Total Hours: 140|
|Contact Hours||T: 22.5||TP: 45||PL:||S:||OT:2|
|Professor in charge||
T - Lectures; TP - Theory and practice; PL - Lab Work; S - Seminar; OT - Tutorial Guidance.
- Learning outcomes of the curricular unit
On successful completion of this course, students will be able to:
- Understand concepts related to signals and their multiple representations, applying them to the synthesis of real signals;
- Analyze a spectrum, interpreting its fundamental characteristics (bandwidth, dominant frequencies, etc ...), correlating it with the temporal representation of the signal under analysis;
- Understand the process of sampling, the Nyquist frequency and the necessary steps to convert analog and digital signals;
- Know the basic SLIT related concepts, convolution, impulse response and frequency;
- Implement FIR and IIR filters, and establish criteria for evaluating their performance.
- Analyze FIR and IIR filters using the Z transform;
- Apply algorithms for synthesis of FIR (window method) filters;
- Understand the stationary and transient regime of a filter;
- Know and understand the tools used in the spectral analysis, such as DFT / FFT and STFT.
I. Revisions about continuous time signals and Fourier representation. Examples with real signals (analysis of electrocardiographic signals - synthesis and reconstruction).
II. Description of the sampling and A/D and D/A conversion process.
III. Acquisition of biomedical signals. Spectrum of interest (ECG, EMG, accelerometry)
IV. Digital Signal processing - Discrete LTIs. Block Diagrams.
V. FIR filters. Convolution and impulse response.
VI. Fourier transform of discrete signals. Frequency response. Systems analysis - filtering concept.
VII. Ideal Filtering - low pass, high band. Typical noise removal.
VIII. Systematic analysis of filters using the transformed Z. Relationship with frequency response. Poles and Zeros.
IX. IIR filters. Direct Forms I and II.
X. Spectral analysis - Discrete Fourier Transform (DFT) and its implementation using Fast Fourier Transform (FFT); spectrogram and relation with the short time Fourier transform (STFT).
- Demonstration of the syllabus coherence with the curricular unit's objectives
This course covers fundamentals of digital signals, filters and their application on biomedical signals. The teaching / learning process is supported by the realization of a set of laboratory work using the Python (using numpy libraries, matplotlib and scipy) and using practical examples of physiological signals collected in situ with the help of sensors and opensource platforms (Arduino and BITalino).
- Teaching methodologies (including evaluation)
The teaching methodology is developed in several components:
T – 21 theoretical teaching contact hours - Presentation and discussion of theoretical concepts, interactivity and asking questions are encouraged;
PL - 42 laboratory practice contact hours: Theoretical concepts are further developed through the implementation of practical examples, performed in groups.
The individual final results are assessed with a final examination given at the end of the semester, with Python-based assignments during the semester.
- Demonstration of the coherence between the teaching methodologies and the learning outcomes
In theoretical classes, syllabus content is presented, which match the learning outcomes 1 to 10.
In laboratorial classes students practice in the Python the techniques associated with these learning outcomes.
- Main Bibliography:
- McClellan, Schafer and Yoder, DSP FIRST: A Multimedia Approach. Prentice Hall, Upper Saddle River, New Jersey, 1998. Prentice Hall
- Alan V. Oppenheim, Ronald W. Schafer, Discrete-Time Signal Processing, Pearson, 3rd Ed., 2013
- Steven W. Smith, The Scientist and Engineer's Guide to Digital Signal Processing, California Tech. Pub., 1st Ed., 1997 - Download gratuito e legal em http://www.dspguide.com