450-4032/01 – Biological Signals Processing (ZBS)
| Gurantor department | Department of Cybernetics and Biomedical Engineering | Credits | 4 |
| Subject guarantor | Ing. Jan Kubíček, Ph.D. | Subject version guarantor | Ing. Jan Kubíček, Ph.D. |
| Study level | undergraduate or graduate | Requirement | Optional |
| Year | 2 | Semester | winter |
| | Study language | Czech |
| Year of introduction | 2010/2011 | Year of cancellation | |
| Intended for the faculties | FEI | Intended for study types | Follow-up Master |
Subject aims expressed by acquired skills and competences
The goal of this course is to give an information about several biomedical signals and their digital signal processing - spectral analysis (frequency and phase spectrum, autocorrelation and cross-correlation analysis), segmentation, classification, fuzzy analysis and methods of display. Acquired knowledge and skill in this subject and the whole alignment forms the basic presumption of knowledge of biomedical engineering.
Graduates will have an appropriate knowledge of techniques, equipments, systems and facilities for acquiring, processing and transferring information and how to use such information in practice. They will be able to apply this knowledge in order to interpret, describe and solve engineering problems.
Teaching methods
Lectures
Individual consultations
Tutorials
Experimental work in labs
Project work
Summary
Characteristics of biological signals, coding data structures, digital signal processing - time domain and frequency domain - spectral analysis (Fourier analysis random (biological) signal and parametric spectrum analysis of random (biological) signal. Adaptive segmentation, automatic classifying biological
signal - learning operations, cluster analysis. Neural networks.Applications of signal processing EKG, EEG.
Compulsory literature:
Recommended literature:
Cohen A., Biomedical signal processing, CRC Press, Boca Raton, Florida
Remond, A.: (Editor-in-chief): Handbook of electroencephalograph and clinical neuro-physiology, vol. 5. Elsevier, 1972
Dumermuth G., Fundamentals of spectral analysis in electroencephalography, In: A. Rémond (Ed.), EEG Informatics : A Didactic Review of Methods and Applications of EEG data Processing. Elsevier, Amsterdam,1977, pp. 83-105
Additional study materials
Way of continuous check of knowledge in the course of semester
Evaluation criteria are oriented on outputs allowing:
• Continuous verifying of student knowledge in the numerical exercises in a form of debate and inquiries to achieve student active participations in study process. Identify, deduce and search of problem solving and their interpretation by students.
• Tests from numerical exercises, eventually from chosen theoretical circuits
• Term work and projects on a given theme on the basis of selection, investigation, ordering and final compilation of facts and their processing into final form of given theme.
E-learning
Other requirements
Any additional requirements aren't for student.
Prerequisities
Subject has no prerequisities.
Co-requisities
Subject has no co-requisities.
Subject syllabus:
Lectures:
Signals in medicine - origin, character and common principles processing biological signal, view of methods and algorithms processing biological signal, EEG, EMG, ECG, EOG. Origin, resources, diagnostics. Chances of exercise bioengineer.
Processing biological signal in real-time and off line. Statistical properties, probability distribution, stochastic processes, analysis of signals in time domain, analysis of signals in frequency domain
Data about patient, identification files. Collection and preprocessing biological data, A/D inverter, aliasing. Filtering. Trends.
Spectral analysis I. - fundamental method. Power spectral density, parametric and non-parametric method. Practical problems estimation of spectra. Cross spectrum, coherency and phase.
Spectral analysis II. - FFT. Application. Method compressed spectral array (CSA). Extraction of the hidden information from signals - local and interhemispheric coherence, phase, measurement of small time differences between EEG channels, time delay
Topographic mapping of brain activity - principle. Use in clinical diagnostics. Dynamic mapping
Adaptive segmentation - Adaptive segmentation with fixed and moving window, Segmentation using the two connected windows. Multichannel adaptive segmentation. Extraction symptoms.
Method automatic classification I - learning without teacher. Structure of data, classes, cluster analysis, fuzzy analysis. K-means algorithm. Limits and limitation fuzzy analyses.
Neural networks, Automatic classification II. - learning classifier, Kohonen layer, classification, classical set theory, fuzzy set theory. Compare with neural net.
Long-term EEG processing, automatic epileptic spike detection. Arithmetical detector, median detector, spike detector based on combination of classical filtering and median filtering
ECG signal, digital processing, characteristics. - frequency analysis, filtration, adaptive filtration. Data reduction, Holter's techniques patient identification.
Respirometry, description signal data. Demand on digital processing and graphic presence.
Video signal - image processing, Presentation in discrete form.
Computer labs:
Introduction into processing biosignal. Practical examples of EEG, EMG, ECG activities, epileptic graphoelements, artefacts.
Statistical characteristics of biosignals. Software. User's interface. Data format.
I Semester work - reading and displaying real signal, term: 1 week
Collection and preprocessing biological data. Data reading in classical and paperless apparatus. A/D inverter Nyquist theorem. Mistakes at transmission.
Spectral analysis I. Fundamental method. Spectral analysis and synthesis - FFT. Filtration, windowing.
II Semester work. Spectral analysis and synthesis of signal term: 2 weeks
Topographic mapping. Demonstration topographic of brain activity - spectral, phase, time delay and coherence mapping - iterative generation map. Animation.
Spectral analysis II - application CSA. Coherence analysis
III Semester work Topographic (brain) mapping - net with 20 point term: 2 weeks
Adaptive segmentation - setting parameter, preference and limitation, algorithm.
Method automatic classification I - learning without teacher. Fundamental algorithm of cluster analyses on simulated data. Examples classification EEG data. Using fuzzy set.
Analysis long-term signal. Summary information.
Extraction compressed information from long-term signal. Applications on real data, further method, programme WaveFinder.
Automatic classification II - learning classifier. Demonstration fundamental algorithm learning classifier on simulated and data. Using fuzzy set in to-NN classifier
IV semester work: 3-NN learning classifier for simulated data. Term: 2 for weeks
Automatic epileptic spike detection - Demonstration commercial programme (Gotman, Scherg, FOCUS).
Preprocessing ECG signal using of wavelet transformation: compression, filtration, artefacts. Calculation of frequency, amplitude and phase spectrum.
One detector algorithm of QRS complex - calculation and comparison, using method detection R-R internals depending on morbid state and variability heart rhythm. Practical demonstration fully automatic evaluative system signal ECG on regional hygienic station in Ostrava.
(Excurse in computerized EEG laboratory Neurological department FN Bulovka. Consultation record semester washing.)
Conditions for subject completion
Occurrence in study plans
Occurrence in special blocks
Assessment of instruction