Brain-computer interface (BCI) is an emerging technology that receives, processes, and converts brain signals into commands sent to output devices to perform desired tasks. Motor imagery (MI) based on electroencephalograms (EEGs) is one of the most widely used BCI paradigms, and it has demonstrated potential as an effective tool for neurorehabilitation. Recently, neural networks-in particular, deep architectures-have received substantial attention for the analysis of EEG signals (BCI applications). This paper proposes a new classification algorithm called EEG-ITT to increase the accuracy of classification motor imagery EEG signals using a non-invasive brain-computer interface. Utilizing a motor imagery dataset of 29 healthy subjects, including males aged 2126 and females aged 18-23, the proposed model demonstrated the highest accuracy, at 79.53 %. The noise injection method has also been implemented for data augmentation.

The brain-computer interface (BCI) is a promising technology that could bring about a significant revolution in various fields, including healthcare and human enhancement. One commonly used BCI method in healthcare, particularly in rehabilitation, is the analysis of motor imagery (MI) through an electroencephalogram (EEG). Our study introduces a hybrid deep learning model called CNN-Transformer-LSTM, which utilizes multi-channel EEG signals to classify MI binary and multiclass automatically. Our experiments have shown that this proposed method is more effective than previous state-of-the-art studies at accurately classifying MI using multi-channel EEG signals.

The relatively small size of EEG datasets impacts the use of traditional and spiking neural networks as EEG data classifiers. Since getting a larger number of EEG recordings requires much laborious laboratory work, using data augmentation methods and techniques seems beneficial. This paper deals with the experiments with, in particular, spiking neural networks over the augmented P300 dataset. Augmentation methods for EEG data are shortly presented; generative adversarial network models and sliding windows of various sizes are used to augment the original P300 dataset. The classification results over the original and augmented P300 datasets are compared, noting that classification accuracy increased by almost 27%

Sleep stage classification plays a crucial role in diagnosing sleep disorders and understanding sleep physiology. In recent years, automated models based on machine learning and deep learning have gained attention for sleep stage classification. This paper uses the single-channel EEG signal to present an automatic sleep stage classification system using a combination of Convolutional Neural Network (CNN), Transformer, and Long Short-Term Memory (LSTM) models. Experimental evaluation of the ISRUC sleep datasets S1 and S3 demonstrates the effectiveness of the proposed model. It achieves accuracies of 80.37% and 82.40%, respectively, achieving competitive performance compared to state-of-the-art models.

Současný model neurorehabilitace poskytuje komplexní péči (neurolog, fyzioterapeut, ergoterapeut, psycholog, logoped) v akutní fázi získaných poškození mozku při hospitalizaci v některých zemích. Následná péče je však nedostatečná nebo zcela chybí, zejména u znevýhodněných osob. Zvýšení dostupnosti a efektivity neurorehabilitační péče je přínosné jak společensky, tak ekonomicky. Projekt BrainIn si klade za cíl zlepšit kvalitu, účinnost a efektivitu neurorehabilitačních postupů jak v akutní fázi, tak ve fázi dlouhodobé domácí rehabilitace. Jeho cílem je navíc usnadnit a urychlit návrat postižených lidí do jejich rodin, společenského a pracovního života. Projekt BrainIn je volně dostupný jako webová aplikace (https://brainin.kiv.zcu.cz/). Je přizpůsobena potřebám terapeutů a pacientů, kteří prodělali získaná poškození mozku.

Spiking neural networks are a promising concept not only in terms of better simulation of biological neural networks but also in terms of overcoming the current disadvantages of artificial neural networks, such as high energy consumption or slow response time. The paper focuses on the potential benefits of spiking neural networks in the classification of event-related components processed in many traditional brain-computer interface experiments. Experiments with various spiking network architectures and optimization approaches over specific brain computer interface and image datasets are presented, and their results are provided and discussed. The best accuracy achieved was 64.86% for the event-related component dataset and 97.09% for the image dataset.

Increasing requirements on data sharing in the domain of electrophysiology lead to proposing new terminologies and data models. A current trend is to describe data by ontologies and semantic web resources. However, classic technologies and models cannot be replaced in a short time. Due to this, dependencies between various data models should be explicitly described and properties, which the models have in common, should be unified. This work summarizes the current state in data modeling. It describes various ways to model and store data and transformation mechanisms between data models. It deals with well-known concepts (relational and object oriented model) as well as with emerging concepts (ontologies). Finally, the hierarchical metadata model consisting of levels with different expressive power is introduced.

Most data management systems include a database in the backend to store data and the associated metadata and a web-based user interface to access and modify the data/metadata. User interfaces are specifically tailored for representing a unique database structure and cannot be easily reused for other database structure. Furthermore the generated web-based layouts are often not compatible for other platform such as desktop applications or mobile devices. We are proposing here a general framework for designing a graphical layout compatible with different platforms including mobile devices and independent of the database structure. This framework is based on a model-driven approach using annotations of database entities that will be used to create desired layouts. A use case study is presented on a database designed for neuroscience experiments.