Biological Signals
Code  Completion  Credits  Range  Language 

17ABBBLS  Z,ZK  4  2P+2C  English 
 Garant předmětu:
 Lecturer:
 Tutor:
 Supervisor:
 Department of Biomedical Technology
 Synopsis:

The subject deals with origins and description of the most important electric and nonelectric biological signals. The principles of generation, recording and basic properties are studied in all the signals. The studied signals involve native and evoked biosignals, including biological signals of the heart, brain, muscles, nervous system, auditory signals, visual system, signals from the gastrointestinal system etc. Advanced methods of digital biosignal processing,spectrum analysis, modern methods of artificial intelligence, features extraction, automatic classification, graphic presentation of results. Adaptive segmentation, artificial neural networks for signal procesing.
 Requirements:

Basic knowledge of MATLAB, C++ or other objectoriented language is an advantage, but it is not strictly required.
Exercises:
Exam:
A. You can not take the exam without getting a credit (exercises) and enrollment in the KOS.
B. The exam consists of a written test, where ABC (one correct one)  1 point and a 5point answer (very important questions) is combined. It is possible to get a maximum of 70 points from the test. The student must earn at least 50 % of the exam points (35 points).
Overall rating of the course:
A. See the ECTS grading scale. 100 points are divided between the parts as follows: maximum 30 % for the obtained credit and maximum 70 % for the passed test/exam.
B. The minimum score is 50. The student must gain at least 15 points from the exercise and a minimum of 35 points per exam.
C. Bonus for successful solving of optional tasks  10 points / task = max. 40 points.
The overall assessment is carried out according to the ECTS table given in the CTU:
10090 points: A, great
8980 points: B, very good
7970 points: C, okay
6960 points: D, satisfactory
5950 points: E, enough
less than 50 points: F, insufficient
 Syllabus of lectures:

1. Introduction to digital biosignal processing. Motivation. Basic characteristics of EEG, EKG, EOG, EMG. Basic graphoelements in EEG, polysomnography, hypnogram. Polysomnography. Artefacts.
2. Statistic and probabilistic signal properties. Probability distribution. Stochastic processes and time series analysis. Convolution, impulse characteristics. Mean, standard deviation, correlation analysis. Crosscorrelation function. The nonstationary behaviour of EEG. Frequency bands.
3. Biological signals recording and preprocessing. Digital EEG devices. Basic sequence of signal transfer into computer. A/D converter, differential amplifiers. Analog and digital filters. Problems of sampling and quantization, Nyquist theorem and sampling frequency. Errors during signal conversion. Signal conditioning, aliasing in the time and frequency domains. Digital and frequency aliasing. Denoising a detrending. EEG machine calibration.
4. ECG, method of measurement and basic signal characteristics. EOG, method of measurement and basic signal characteristics.
5. EMG, method of measurement and basic signal characteristics. Multimodal monitoring.
6. Evoked potentials, VEP, AEP, SEP, BAEP, MEP.
7. Fourier transformation. Discrete FT. Fast FT (FFT). Principles of computing. Decimation in time and frequency. FFT butterfly. Special algorithms of computing. Inverse transform. Signal analysis and synthesis. Spectrum estimation. Filtering using FFT. Digital filters for biosignal analysis. FIR and IIR filters, properties. Linear and nonlinear phase characteristics. Types of filters, band pass, low pass, high pass, notch filters. Simple methods of design. Example of design using FFT (window method). Examples of application to real and simulated signal.
8. Spectrum analysis. Power spectral density. Periodogram. Parametric and nonparametric methods of spectral analysis. Practical problems of spectrum estimation. CSA
9. Multichannel adaptive segmentation. Motivation. Nonstationarity of biosignals. Basic methods. Multichannel online adaptive segmentation. Extraction of symptoms. The parameter settings. Advantages and limitations of methods. Other segmentation algorithms.
10. Methods of automatic classification. Basic algorithms of cluster analysis. Kmeans algorithm. Optimal number of classes. Limits and constraints of cluster analysis. Fuzzy cluster analysis.
11. Densitybased classification methods. Instancebased learning methods. KNN classification. Fuzzy kNN. Practical examples of classification methods for biological signals.
12. Simple methods for automatic epileptic spikes detection.
13. Topographic mapping of electrophysiological activity. Visualization. Principle of brain mapping. Amplitude and frequency brain mapping. Interpolation. Direct and inverse task. Use in clinical diagnostics.
14. Metrics. Data normalization. Statistical data processing.
 Syllabus of tutorials:

1. Artefacts in biosignal recording. Measurement of the electrical properties of the recording electrodes.
2. Measurement on ECG.
3. Measurement on EMG.
4. Measurement on EEG.
5. Measurement of tendon jerks.
6. Recording of evoked EEG potentials.
7. Audiometric measurements.
Conditions for credit
A. Participation in the exercises, max. 1 unexcused absence
B. Submission of the measurement protocols
C. Presentation of the selected topic (510min, PowerPoint)
D. Passing the test at the end of the semester with questions from the practical measurements
 Study Objective:

The aim of the subject is to get to know the basic biological signals of electric and nonelectric origin, methods of their recording protecting their diagnostic values.
The basic and advanced methods of biosignal processing will be discussed.
 Study materials:

[1] Sormno L, Laguna P, Bioelectrical Signal Processing in nurological and cardiological applications, Elsevier,2005
[2] Bruce, E.N. Biomedical Signal Processing and Signal Modelling.New York, J.Willey & sons 2001.
[3] Baura G.D. System Theory and Practical Applications of Biomedical Signals.Piscataway, IEEE Press 2002.
[4] Krajca V., Mohylova J. Biologicke signely. elearning www.skolicka.fbmi.cvut.cz, password signaly
[5] MIKE X. COHEN. Analyzing neural time series data: theory and practice. 2014. ISBN 0262019876.
 Note:
 Further information:
 www.skolicka.fbmi.cvut.cz
 No timetable has been prepared for this course
 The course is a part of the following study plans:

 Biomedical Technician  full time study in English (compulsory course)