Correlation dimension based lossless compression of EEG signals

A Personalized Compression Method for Steady-State Visual Evoked Potential EEG Signals

As an informative electroencephalogram (EEG) signal, steady-state visual evoked potential (SSVEP) stands out from many paradigms for application in wireless wearable devices. However, its data are usually enormous, occupy too many bandwidth sources and require immense power when transmitted in the raw data form, so it is necessary to compress the signal. This paper proposes a personalized EEG compression and reconstruction algorithm for the SSVEP application. In the algorithm, to realize personalization, a primary artificial neural network (ANN) model is first pre-trained with the open benchmark database towards BCI application (BETA). Then, an adaptive ANN model is generated with incremental learning for each subject to compress their individual data. Additionally, a personalized, non-uniform quantization method is proposed to reduce the errors caused by compression. The recognition accuracy only decreases by 3.79% when the compression rate is 12.7 times, and is tested on BETA. The proposed algorithm can reduce signal loss by from 50.43% to 81.08% in the accuracy test compared to the case without ANN and uniform quantization.

The effect of photic stimulation alone and in combination with sleep deprivation after a seizure-like event - reappraisal by using linear and nonlinear EEG methods

ObjectivesThe present study aimed to compare the effectiveness of different provocation tests used for the study of the 'susceptibility to seizure' by quantitative electroencephalography (EEG) analysis.MethodsEight subjects with a history of a seizure-like disturbed consciousness participated in this preliminary study. A routine EEG was carried out with photic stimulation (eyes closed and after eyes open) at the beginning of the investigation. Some days later, a sleep-deprived EEG was recorded with the same protocol. Selected epochs (in eyes closed condition) after the stimulations were analysed with Point(wise) Correlation Dimension (PD2i) and Synchronization Likelihood (SL) methods. The results were compared to those obtained by similar analysis of the resting state (control) epochs with Wilcoxon Signed Rank Test (p ≤ 0.05).ResultsIn our study, significantly lower grand mean PD2i and higher delta SL values were found in sleep-deprived state after stimulation with eyes closed compared to the control. Our results indicated a lower-dimensional, hypersynchronous state of the brain as a consequence of these combined provocations.DiscussionThis may correspond to a possible 'preictal' state of the brain. Accordingly, it is suggested that photic stimulation together with sleep deprivation seems to be more effective in provocation - especially when the stimulation was made with eyes closed.

Complexity Analysis of EEG, MEG, and fMRI in Mild Cognitive Impairment and Alzheimer's Disease: A Review

Alzheimer's disease (AD) is a degenerative brain disease with a high and irreversible incidence. In recent years, because brain signals have complex nonlinear dynamics, there has been growing interest in studying complex changes in the time series of brain signals in patients with AD. We reviewed studies of complexity analyses of single-channel time series from electroencephalogram (EEG), magnetoencephalogram (MEG), and functional magnetic resonance imaging (fMRI) in AD and determined future research directions. A systematic literature search for 2000-2019 was performed in the Web of Science and PubMed databases, resulting in 126 identified studies. Compared to healthy individuals, the signals from AD patients have less complexity and more predictable oscillations, which are found mainly in the left parietal, occipital, right frontal, and temporal regions. This complexity is considered a potential biomarker for accurately responding to the functional lesion in AD. The current review helps to reveal the patterns of dysfunction in the brains of patients with AD and to investigate whether signal complexity can be used as a biomarker to accurately respond to the functional lesion in AD. We proposed further studies in the signal complexities of AD patients, including investigating the reliability of complexity algorithms and the spatial patterns of signal complexity. In conclusion, the current review helps to better understand the complexity of abnormalities in the AD brain and provide useful information for AD diagnosis.

N-WRETS: Near-Lossless Wireless Real-time Efficient Electroencephalogram Transmission Solution to Support Sleep Disorder Monitoring Platforms

BACKGROUND

Sleep disorders lead to many adverse complications and chronic diseases. Sleep disorder-related healthcare costs are tens of billions of dollars worldwide. Sleep monitoring solutions have thus been the focus of research and industrial interest. However, the problem of limited bandwidth and battery consumption hinders the accuracy and practical use of sleep monitoring aids.

INTRODUCTION

The aim of this study is to propose Near-Lossless Wireless Real-time Efficient electroencephalogram Transmission Solution (N-WRETS) solution that solves the issue of limited bandwidth and battery consumption, thereby supporting platforms dedicated to sleep disorder monitoring.

MATERIALS AND METHODS

Electroencephalography (EEG) data materials were obtained from the Physionet PhysioBank database. The CAP Sleep Database was used. C programming was used for development.

RESULTS

To evaluate transmission efficiency, the compression ratio (CR) was compared to prior studies. The N-WRETS CR of 11.34 exceeded other reported values.

DISCUSSION

Compared to prior related research, N-WRETS showed the highest compression performance for EEG, but showed the lowest stability, which was a trade-off for its high efficiency. This article opens a possibility for future research to improve the performance of EEG compression algorithms according to sleep disease type. N-WRETS is also near-lossless, which is fit for priceless EEG data that contain important information on the patient's health. The proposed solution also supported wireless real-time transmission, which was another distinctive characteristic compared to related studies.

CONCLUSIONS

N-WRETS may provide a platform in which sleep disorder patients may be properly monitored in real time. The system could overcome the problems of limited bandwidth and battery consumption.

Highly Efficient Compression Algorithms for Multichannel EEG

The difficulty associated with processing and understanding the high dimensionality of electroencephalogram (EEG) data requires developing efficient and robust compression algorithms. In this paper, different lossless compression techniques of single and multichannel EEG data, including Huffman coding, arithmetic coding, Markov predictor, linear predictor, context-based error modeling, multivariate autoregression (MVAR), and a low complexity bivariate model have been examined and their performances have been compared. Furthermore, a high compression algorithm named general MVAR and a modified context-based error modeling for multichannel EEG have been proposed. The resulting compression algorithm produces a higher relative compression ratio of 70.64% on average compared with the existing methods, and in some cases, it goes up to 83.06%. The proposed methods are designed to compress a large amount of multichannel EEG data efficiently so that the data storage and transmission bandwidth can be effectively used. These methods have been validated using several experimental multichannel EEG recordings of different subjects and publicly available standard databases. The satisfactory parametric measures of these methods, namely percent-root-mean square distortion, peak signal-to-noise ratio, root-mean-square error, and cross correlation, show their superiority over the state-of-the-art compression methods.

A simple and efficient algorithm operating with linear time for MCEEG data compression

Popularisation of electroencephalograph (EEG) signals in diversified fields have increased the need for devices capable of operating at lower power and storage requirements. This has led to a great deal of research in data compression, that can address (a) low latency in the coding of the signal, (b) reduced hardware and software dependencies, (c) quantify the system anomalies, and (d) effectively reconstruct the compressed signal. This paper proposes a computationally simple and novel coding scheme named spatial pseudo codec (SPC), to achieve lossy to near lossless compression of multichannel EEG (MCEEG). In the proposed system, MCEEG signals are initially normalized, followed by two parallel processes: one operating on integer part and the other, on fractional part of the normalized data. The redundancies in integer part are exploited using spatial domain encoder, and the fractional part is coded as pseudo integers. The proposed method has been tested on a wide range of databases having variable sampling rates and resolutions. Results indicate that the algorithm has a good recovery performance with an average percentage root mean square deviation (PRD) of 2.72 for an average compression ratio (CR) of 3.16. Furthermore, the algorithm has a complexity of only O(n) with an average encoding and decoding time per sample of 0.3 ms and 0.04 ms respectively. The performance of the algorithm is comparable with recent methods like fast discrete cosine transform (fDCT) and tensor decomposition methods. The results validated the feasibility of the proposed compression scheme for practical MCEEG recording, archiving and brain computer interfacing systems.

A new near-lossless EEG compression method using ANN-based reconstruction technique

Compression algorithm is an essential part of Telemedicine systems, to store and transmit large amount of medical signals. Most of existing compression methods utilize fixed transforms such as discrete cosine transform (DCT) and wavelet and usually cannot efficiently extract signal redundancy especially for non-stationary signals such as electroencephalogram (EEG). In this paper, we first propose learning-based adaptive transform using combination of DCT and artificial neural network (ANN) reconstruction technique. This adaptive ANN-based transform is applied to the DCT coefficients of EEG data to reduce its dimensionality and also to estimate the original DCT coefficients of EEG in the reconstruction phase. To develop a new near lossless compression method, the difference between the original DCT coefficients and estimated ones are also quantized. The quantized error is coded using Arithmetic coding and sent along with the estimated DCT coefficients as compressed data. The proposed method was applied to various datasets and the results show higher compression rate compared to the state-of-the-art methods.

EEG compression of scalp recordings based on dipole fitting

A novel technique for electroencephalogram (EEG) compression is proposed in this paper. This technique models the intrinsic dependence inherent between the different EEG channels. It is based on methods borrowed from dipole fitting that is usually used in order to find a solution to the classic problems in EEG analysis: inverse and forward problems. To compress the EEG signals, the forward model based on approximated source dipoles is first used to provide an approximation of the recorded signals. Then, (based on a smoothness factor) appropriate coding techniques are suggested to compress the residuals of the fitting process. Results show that this technique works well for different recordings and for different patients, and is even able to provide near-lossless compression for certain types of recordings.