Publications

sort by:  author date

BodyInNumbers - Software Prototype for Rapid Collection and Storage of Heterogeneous Health Related Data

BRŮHA, P., MOUČEK, R., ŠNEJDAR, P., VACEK, V., KRAFT, V., BOHMANN, D., VAŘEKA, L., ČERNÁ, K., ŘEHOŘ, P. BodyInNumbers - Software Prototype for Rapid Collection and Storage of Heterogeneous Health Related Data. 2017.
Abstract PDF BibTeX

BodyInNumbers is a software prototype for rapid collection, storage, management and visualization of heterogeneous health related data (reaction time data, P300 event-related component data, color vision, spirometry, electrocardiography, blood pressure, blood glucose, body proportions and flexibility) together with corresponding metadata (for example, a summary of the participant\'s current lifestyle and health). After data evaluation the user can view relevant information related to his/her health and fitness. The software was supported by the UWB grant SGS-2016-018 Data and Software Engineering for Advanced Applications, the project LO1506 of the Czech Ministry of Education, Youth and Sports under the program NPU I and the 2nd Internal grant scheme of UWB School of Computing, 2016. The project repository is available at https://gitlab.com/bodyinnumbers-public/bodyinnumbers-public.git. Information about the project is available at http://bodyinnumbers.kiv.zcu.cz/. The software prototype has been tested on 470 people in real environment (mainly during the Days of Science and Technology 2016 and 2017) and continuously improved according to operation difficulties. Published in: BRUHA, Petr, et al. Exercise and Wellness Health Strategy Framework. BIOSTEC 2017, 2017, 477.

Heart rate and sentiment experimental data with common timeline

SALAMON, J., MOUČEK, R. Heart rate and sentiment experimental data with common timeline. Data in Brief, 2017, roč. 15, č. December 2017, s. 851-861. ISSN: 2352-3409
Abstract PDF BibTeX

Sentiment extraction and analysis using spoken utterances or written corpora as well as collection and analysis of human heart rate data using sensors are commonly used techniques and methods. On the other hand, these have been not combined yet. The collected data can be used e.g. to investigate the mutual dependence of human physical and emotional activity. The paper describes the procedure of parallel acquisition of heart rate sensor data and tweets expressing sentiment and difficulties related to this procedure. The obtained datasets are described in detail and further discussed to provide as much information as possible for subsequent analyses and conclusions. Analyses and conclusions are not included in this paper. The presented experiment and provided datasets serve as the first basis for further studies where all four presented data sources can be used independently, combined in a reasonable way or used all together. For instance, when the data is used all together, performing studies comparing human sensor data, acquired noninvasively from the surface of the human body and considered as more objective, and human written data expressing the sentiment, which is at least partly cognitively interpreted and thus considered as more subjective, could be beneficial.

Archetype-based approach for modelling of electroencephalographic/event-related potentials data and metadata

PAPEŽ, V. Archetype-based approach for modelling of electroencephalographic/event-related potentials data and metadata. 1. vyd. Plzeň : neuveden, 2017, 170 s. ISBN: neuvedeno
Abstract PDF BibTeX

Currently, there is no common data standard in the experimental electro-encephalography/event. related potential (EEG/ERP) domain. Existing standardization efforts are mainly based on the conventional approaches and use generic data formats and containers (e.g. HDF5, odML) popular in the research community. This work draws on the medical/health characteristics of EEG/ERP data and investigates the feasibility of applying openEHR (an archetype-based approach for electronic health records representation) to modelling data stored in EEGBase, a portal for experimental EEG/ERP data management. The work evaluates re-usage of existing openEHR archetypes and proposes a set of new archetypes together with the openEHR templates covering the domain. The main goals of the work are to (i) link existing EEGBase data/metadata and openEHR archetype structures; (ii) propose a new openEHR archetype set describing the EEG/ERP domain since this set of archetypes currently does not exist in public repositories. Apart from that, the work describes common data models (e.g. relational, object-oriented) and compares their expressive power in order to (i) determine the elements, which these models have in common; (ii) build a data model hierarchy according to their expressive power. The work uses the proposed archetypes and their reference models as semantic schemata to derive a specific data model for each level of the hierarchy. Finally, the work describes a~newly proposed personal electronic health records system for research purposes, which serves as a~first use-case of obtained results.

Stacked Autoencoders for the P300 Component Detection

VAŘEKA, L., MAUTNER, P. Stacked Autoencoders for the P300 Component Detection. Frontiers in Neuroscience, 2017, roč. 11, č. 302, s. 1-9. ISSN: 1662-453X
Abstract PDF BibTeX

Using a combination of unsupervised pre-training and subsequent fine-tuning, deep neural networks have become one of the most reliable classification methods. The aim of the experiments subsequently presented was to verify if deep learning-based models can also perform well for single trial P300 classification with possible application to P300-based brain-computer interfaces. The P300 data used were recorded in the EEG/ERP laboratory at the Department of Computer Science and Engineering, University of West Bohemia, and are publicly available. Stacked autoencoders were implemented and compared with some of the currently most reliable state-of-the-art methods, such as LDA and multi-layer perceptron. The parameters of stacked autoencoders were optimized empirically. Subsequently, fine-tuning using backpropagation was performed. The architecture of the neural network was 209-130-100-50-20-2. The classifiers were trained on a dataset merged from four subjects and subsequently tested on different 11 subjects without further training. The trained SAE achieved 69.2% accuracy that was higher (p < 0.01) than the accuracy of MLP (64.9%) and LDA (65.9%). The recall of 58.8% was slightly higher when compared with MLP (56.2%) and LDA (58.4%).

Text, Speech and Dialogue, 20 th International Conference, TSD 2017

Ekštein, K., Konopík, M., Matoušek, V., Mouček, R. Text, Speech and Dialogue, 20 th International Conference, TSD 2017. Praha, 27.08.2017 - 31.08.2017.
Abstract PDF BibTeX

The TSD 2017 conference was held on August 27 - 31, 2017 in Prague, Czech Republic. It was focused primarely on the following topics language corpora and their utilizaion, speech recognition, natural language processing, speech synthesis, automatic dialogtue systems and related.

Our partners