Lectures:
- Introduction to biomedical signals and images like: electrocardiographic signals (ECG), neurophysiological signals (EEG, EMG, EHG), medical images (CT, MRI, OCT) and introduction to modern computer technologies in selected clinical settings.
- International standardized reference databases of medical samples (MIT/BIH BD, LTST DB, TPEHG DB, TPEHGT DS, ICEHG DS, EEGMMI DS) and Physionet repository.
- Feature extraction (time domain, Fourier transform, wavelets, principal components – Karhunen-Loeve transform, feature representations).
- Noise extraction (linear procedures in time domain, feature space procedures, weighted averaging, robust approaches).
- Spectral analysis and characterization of samples and features (time-frequency representations, spaces of diagnostic and morphologic features).
- Analysis of time series and nonstationary signals.
- Modelling (linear stochastic and non-linear models, autoregressive modelling).
- Event detection, clustering and event classification (techniques in time domain and in feature space), machine learning approaches
- Processing of images and 3-D and images with the aim of noise reduction, conture extraction, and segmentation and visualization of anatomical structures.
- Research challenges in the area: automated detection of tansient heart ischaemia (ECG); automated prediction of preterm birth (EHG); brain computer communication (EEG); automated detection and evaluation of atrophy in retinal images (OCT images).
- Performance evaluation of biomedical computer systems (metrics, protocols, predictioning performance in real world, assessing robustness, standards).
Laboratory work: Practical work will be performed in the form of project work in suitable equipped student laboratories. Students in the scope of projects independently implement procedures. Obligatory work on projects allows deepen and critical understanding of the subject topics and stimulates to independence and creativity.