2The decoding of selective auditory attention from noninvasive electroencephalogram (EEG) 3 data is of interest in brain computer interface and auditory perception research. The current 4 state-of-the-art approaches for decoding the attentional selection of listeners are based on 5 temporal response functions (TRFs). In the current context, a TRF is a function that facilitates a 6 mapping between features of sound streams and EEG responses. It has been shown that when 7 the envelope of attended speech and EEG responses are used to derive TRF mapping functions, 8 the TRF model predictions can be used to discriminate between attended and unattended talkers. 9 However, the predictive performance of the TRF models is dependent on how the TRF model 10 parameters are estimated. There exist a number of TRF estimation methods that have been 11 published, along with a variety of datasets. It is currently unclear if any of these methods perform 12 better than others, as they have not yet been compared side by side on a single standardized 13 dataset in a controlled fashion. Here, we present a comparative study of the ability of different TRF 14 estimation methods to classify attended speakers from multi-channel EEG data. The performance 15 of the TRF estimation methods is evaluated using different performance metrics on a set of 16 labeled EEG data from 18 subjects listening to mixtures of two speech streams. 17 Keywords: temporal response function, speech decoding, electroencephalography, selective auditory attention, attention decoding 18 21 1 Wong et al. Auditory Attention Decoding Method Comparisonsuccession of repeated short stimuli. More recently, these methods have been extended to continuous stimuli 22 such as speech by using linear stimulus-reponse models, broadly termed 'temporal response functions' 23 (TRFs). The TRF characterizes how a unit impulse in an input feature corresponds to a change in the 24 M/EEG data. TRFs can be used to generate continuous predictions about M/EEG responses or stimulus 25 features, as opposed to characterizing the response (ERP) to repetitions of the same stimuli. Importantly, it 26 has been demonstrated that the stimulus-response models can be extracted both from EEG responses to 27 artificial sound stimuli (16) but also from EEG responses to naturalistic speech (17). A number of studies 28 have considered mappings between the slowly varying temporal envelope of a speech sound signal (<10 29 Hz) and the corresponding filtered M/EEG response (16, 28, 11, 12). However, TRFs are not just limited to 30 the broadband envelope, but can also be obtained with the speech spectrogram (9, 10), phonemes (8), or 31 semantic features (4). This has opened new avenues of research into cortical responses to speech, advancing 32 the field beyond examining responses to repeated isolated segments of speech. 33TRF decoding methods have proven particularly apt for studying how the cortical processing of speech 34 features are modulated by selective auditory attention. A number of st...
Brain signals recorded with electroencephalography (EEG), magnetoencephalography (MEG) and related techniques often have poor signal-to-noise ratio due to the presence of multiple competing sources and artifacts. A common remedy is to average over repeats of the same stimulus, but this is not applicable for temporally extended stimuli that are presented only once (speech, music, movies, natural sound). An alternative is to average responses over multiple subjects that were presented with the same identical stimuli, but differences in geometry of brain sources and sensors reduce the effectiveness of this solution. Multiway canonical correlation analysis (MCCA) brings a solution to this problem by allowing data from multiple subjects to be fused in such a way as to extract components common to all. This paper reviews the method, offers application examples that illustrate its effectiveness, and outlines the caveats and risks entailed by the method.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.