Course program
EDSB1 constitutes the second module (6 CFU) of the integrated course DETERMINISTIC AND STOCASTIC SIGNALS AND BIOMEDICAL DATA AND SIGNAL PROCESSING I (12 CFU).
The first module of this course (Deterministic and Stochastic Signals - SDS) of 6 CFU is delivered by Prof. Lorenzo Piazzo.
1. Measurement chain of a biosignal
1.1. Introduction: biosignals, measurements and noise
1.2. Measurement systems
1.3. Transducers
1.4. Analogue signal processing
1.5. Noise
1.6. ADC conversion
1.7. Fourier analysis
2. Power Density Spectrum Estimation
2.1. Discrete signals
2.2. Discrete Fourier Transform
2.3. Fast Fourier Transform
2.4. Non-parametric spectral estimation
2.4.1. Intro
2.4.2. Stochastic series
2.4.3. Estimation theory
2.4.4. Windowing
2.4.5. Classical periodogram
2.4.6. Modified periodogram
2.4.7. Bartlett's periodogram
2.4.8. Welch periodogram
2.4.9. Blackman-Tukey periodogram
2.5. Parametric spectral estimation
2.5.1. Parametric methods
2.5.2. Autoregressive models
2.5.3. Yule-Walker equations
3. Electroencephalographic signal
3.1. Physiology behind the EEG signal
3.2. Instrumentation for EEG signal acquisition
3.3. EEG signal processing
4. Electromyographic signal
4.1. Physiology behind the EEG signal
4.2. Instrumentation for EEG signal acquisition
4.3. EEG signal processing
Prerequisites
The main prerequisites to understand the topics addressed during the course are:
1. Basis principles of signals theory (fundamental):
1.1 Deterministic and stochastic signals
1.2 Energy and Power signals
1.3 Nyquist theorem
1.4 Fourier transform
2. Statistics and probability (fundamental):
2.1 Distributions
2.2 Descriptive statistics
2.3 Stationery and ergodicity of a random process
3. Estimation theory (important):
3.1 Estimator definition
3.2 Main performance parameter of an estimator
4. Mathematical analysis (useful):
4.1 Integrals
4.2 Correlation
4.3 Systems of equations
5. Basic principles of human anatomy and physiology (important):
5.1 cells functioning mechanisms
5.2 anatomy of nervous system
5.2 anatomy of musculoskeletal system
Books
Semmlow and Griffel, “Biosignal and Medical Image Processing”, CRC Press, 2014
Smith, “The scientist and Engineer’s Guide to Digital Signal Processing”, Second Edition, California Technical Publishing, San Diego, California, 1999
Prabhu, “Window functions and their application in Signal Processing”, CRC Press, 2014
Mecarelli, “Clinical Electroencephalograpy”, Springer, California, 2019
Kamen and Gabriel, “Essentials of Electromiography”, Human Kinetics, 2010
Rangayyan, “Biomedical Signal Analysis”, Second Edition, Wiley IEEE Press, 2015
Frequency
The course is completely delivered according to the in-person modality (info about time and place could be find on the university web site). Students can attend lessons freely since any site attendance is recorded.
Exam mode
The examination consists of a written test organised as follows
33 closed-answer questions (true/false) on 11 different topics covered throughout the course (3 per topic) --> max 33 points
2 exercises relating to spectral analysis --> max 5 points
1 short open-ended question --> max 10 points
For each of the 33 sentences in the assignment, the student must indicate whether the statement is true or false, or may choose not to answer. Will be awarded: 1pt for each correct answer, -0.8pt for each incorrect answer and 0pt for each answer not given. Each of the two exercises will be given a mark of between 0 and 3 points.
The final mark for the paper will be the weighted sum of the marks taken in the three sections: 70% section 1, 10% section 2, 20% section 3.
The test is passed with a mark >18. Those who pass the examination with a mark above 28 may request an oral test (1 open question on the course syllabus) to which the lecturer may award or deduct a maximum of 3 points (in relation to the written mark).
The oral mark will be the arithmetic mean of the marks obtained in the two modules (SDS+EDSB1).
The candidate must pass both modules in the same calendar year, otherwise they must re-sit the module they have passed.
Lesson mode
The course will be delivered for 60 hours (6CFU) organized as follows:
1. 48 hours (80%) of traditional lessons
2. 9 hours of practical lessons (exercises)
3. 3 hours dedicated to seminars about the main research trends in the analysis of EEG and EMG signals
