A Convolutional Autoencoder-based Explainable Clustering Approach for Resting-State EEG Analysis. ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY. IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY. ANNUAL INTERNATIONAL CONFERENCE 2023;2023:1-4. [PMID: 38083554 DOI: 10.1109/embc40787.2023.10340375] [Citation(s) in RCA: 0] [Impact Index Per Article: 0]

Machine learning methods have frequently been applied to electroencephalography (EEG) data. However, while supervised EEG classification is well-developed, relatively few studies have clustered EEG, which is problematic given the potential for clustering EEG to identify novel subtypes or patterns of dynamics that could improve our understanding of neuropsychiatric disorders. There are established methods for clustering EEG using manually extracted features that reduce the richness of the feature space for clustering, but only a couple studies have sought to use deep learning-based approaches with automated feature learning to cluster EEG. Those studies involve separately training an autoencoder and then performing clustering on the extracted features, and the separation of those steps can lead to poor quality clustering. In this study, we propose an explainable convolutional autoencoder-based approach that combines model training with clustering to yield high quality clusters. We apply the approach within the context of schizophrenia (SZ), identifying 8 EEG states characterized by varying levels of δ activity. We also find that individuals who spend more time outside of the dominant state tend to have increased negative symptom severity. Our approach represents a significant step forward for clustering resting-state EEG data and has the potential to lead to novel findings across a variety of neurological and neuropsychological disorders in future years.

Systems, Subjects, Sessions: To What Extent Do These Factors Influence EEG Data?

Lab-based electroencephalography (EEG) techniques have matured over decades of research and can produce high-quality scientific data. It is often assumed that the specific choice of EEG system has limited impact on the data and does not add variance to the results. However, many low cost and mobile EEG systems are now available, and there is some doubt as to the how EEG data vary across these newer systems. We sought to determine how variance across systems compares to variance across subjects or repeated sessions. We tested four EEG systems: two standard research-grade systems, one system designed for mobile use with dry electrodes, and an affordable mobile system with a lower channel count. We recorded four subjects three times with each of the four EEG systems. This setup allowed us to assess the influence of all three factors on the variance of data. Subjects performed a battery of six short standard EEG paradigms based on event-related potentials (ERPs) and steady-state visually evoked potential (SSVEP). Results demonstrated that subjects account for 32% of the variance, systems for 9% of the variance, and repeated sessions for each subject-system combination for 1% of the variance. In most lab-based EEG research, the number of subjects per study typically ranges from 10 to 20, and error of uncertainty in estimates of the mean (like ERP) will improve by the square root of the number of subjects. As a result, the variance due to EEG system (9%) is of the same order of magnitude as variance due to subjects (32%/sqrt(16) = 8%) with a pool of 16 subjects. The two standard research-grade EEG systems had no significantly different means from each other across all paradigms. However, the two other EEG systems demonstrated different mean values from one or both of the two standard research-grade EEG systems in at least half of the paradigms. In addition to providing specific estimates of the variability across EEG systems, subjects, and repeated sessions, we also propose a benchmark to evaluate new mobile EEG systems by means of ERP responses.

Simulation of an optimized technique based on DS-CDMA for simultaneous transmission of multichannel biosignals

Telemedicine is becoming increasingly, with applications in many areas of healthcare, such as home telecare of the elderly, diagnosis at a distance and robotic surgery. The simultaneous transmission of several leads of biomedical signals should be considered in telemedicine, given the many benefits it brings. Code division multiple access (CDMA) is a multiple access technique that enables users to transmit independent information simultaneously within the same bandwidth. The direct sequence CDMA (DS-CDMA) is a variant of the CDMA technique in which a pseudorandom sequence having a higher bandwidth than the information signal is used to modulate the information signal directly. Biomedical signals are confidential; thus, their transmission must be secured. In this paper we propose a protocol similar to DS-CDMA for the simultaneous transmission of all of the leads of some multichannel biomedical signals. We assigned orthogonal codes to different leads of a signal. The convolution of each lead with the code gives a signal spread over a broad frequency band. All of the spread signals are then mixed to produce a single composite signal. This composite signal is frequency modulated, amplified and transmitted. At the reception, inverse functions to the previous are developed to perform demodulation, demultiplexing and extraction of the physiological signals transmitted. We used the discrete Walsh functions as codes. The results obtained are satisfactory, even in situations where the noise disturbances are significant.

A programmable laboratory testbed in support of evaluation of functional brain activation and connectivity

An important determinant of the value of quantitative neuroimaging studies is the reliability of the derived information, which is a function of the data collection conditions. Near infrared spectroscopy (NIRS) and electroencelphalography are independent sensing domains that are well suited to explore principal elements of the brain's response to neuroactivation, and whose integration supports development of compact, even wearable, systems suitable for use in open environments. In an effort to maximize the translatability and utility of such resources, we have established an experimental laboratory testbed that supports measures and analysis of simulated macroscopic bioelectric and hemodynamic responses of the brain. Principal elements of the testbed include 1) a programmable anthropomorphic head phantom containing a multisignal source array embedded within a matrix that approximates the background optical and bioelectric properties of the brain, 2) integrated translatable headgear that support multimodal studies, and 3) an integrated data analysis environment that supports anatomically based mapping of experiment-derived measures that are directly and not directly observable. Here, we present a description of system components and fabrication, an overview of the analysis environment, and findings from a representative study that document the ability to experimentally validate effective connectivity models based on NIRS tomography.