Interpretable many-class decoding for MEG
Multivariate pattern analysis (MVPA) of Magnetoencephalography (MEG) and Electroencephalography (EEG) data is a valuable tool for understanding how the brain represents and discriminates between different stimuli. Identifying the spatial and temporal signatures of stimuli is typically a crucial outp...
| Main Authors: | , , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
Elsevier
2023
|
| _version_ | 1797111916309512192 |
|---|---|
| author | Csaky, R van Es, MWJ Parker Jones, O Woolrich, M |
| author_facet | Csaky, R van Es, MWJ Parker Jones, O Woolrich, M |
| author_sort | Csaky, R |
| collection | OXFORD |
| description | Multivariate pattern analysis (MVPA) of Magnetoencephalography (MEG) and Electroencephalography (EEG) data is a valuable tool for understanding how the brain represents and discriminates between different stimuli. Identifying the spatial and temporal signatures of stimuli is typically a crucial output of these analyses. Such analyses are mainly performed using linear, pairwise, sliding window decoding models. These allow for relative ease of interpretation, e.g. by estimating a time-course of decoding accuracy, but have limited decoding performance. On the other hand, full epoch multiclass decoding models, commonly used for brain–computer interface (BCI) applications, can provide better decoding performance. However interpretation methods for such models have been designed with a low number of classes in mind. In this paper, we propose an approach that combines a multiclass, full epoch decoding model with supervised dimensionality reduction, while still being able to reveal the contributions of spatiotemporal and spectral features using permutation feature importance. Crucially, we introduce a way of doing supervised dimensionality reduction of input features within a neural network optimised for the classification task, improving performance substantially. We demonstrate the approach on 3 different many-class task-MEG datasets using image presentations. Our results demonstrate that this approach consistently achieves higher accuracy than the peak accuracy of a sliding window decoder while estimating the relevant spatiotemporal features in the MEG signal. |
| first_indexed | 2024-03-07T08:17:05Z |
| format | Journal article |
| id | oxford-uuid:5be12907-1dd9-4183-b7dd-980148a94306 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T08:17:05Z |
| publishDate | 2023 |
| publisher | Elsevier |
| record_format | dspace |
| spelling | oxford-uuid:5be12907-1dd9-4183-b7dd-980148a943062024-01-11T09:38:02ZInterpretable many-class decoding for MEGJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:5be12907-1dd9-4183-b7dd-980148a94306EnglishSymplectic ElementsElsevier2023Csaky, Rvan Es, MWJParker Jones, OWoolrich, MMultivariate pattern analysis (MVPA) of Magnetoencephalography (MEG) and Electroencephalography (EEG) data is a valuable tool for understanding how the brain represents and discriminates between different stimuli. Identifying the spatial and temporal signatures of stimuli is typically a crucial output of these analyses. Such analyses are mainly performed using linear, pairwise, sliding window decoding models. These allow for relative ease of interpretation, e.g. by estimating a time-course of decoding accuracy, but have limited decoding performance. On the other hand, full epoch multiclass decoding models, commonly used for brain–computer interface (BCI) applications, can provide better decoding performance. However interpretation methods for such models have been designed with a low number of classes in mind. In this paper, we propose an approach that combines a multiclass, full epoch decoding model with supervised dimensionality reduction, while still being able to reveal the contributions of spatiotemporal and spectral features using permutation feature importance. Crucially, we introduce a way of doing supervised dimensionality reduction of input features within a neural network optimised for the classification task, improving performance substantially. We demonstrate the approach on 3 different many-class task-MEG datasets using image presentations. Our results demonstrate that this approach consistently achieves higher accuracy than the peak accuracy of a sliding window decoder while estimating the relevant spatiotemporal features in the MEG signal. |
| spellingShingle | Csaky, R van Es, MWJ Parker Jones, O Woolrich, M Interpretable many-class decoding for MEG |
| title | Interpretable many-class decoding for MEG |
| title_full | Interpretable many-class decoding for MEG |
| title_fullStr | Interpretable many-class decoding for MEG |
| title_full_unstemmed | Interpretable many-class decoding for MEG |
| title_short | Interpretable many-class decoding for MEG |
| title_sort | interpretable many class decoding for meg |
| work_keys_str_mv | AT csakyr interpretablemanyclassdecodingformeg AT vanesmwj interpretablemanyclassdecodingformeg AT parkerjoneso interpretablemanyclassdecodingformeg AT woolrichm interpretablemanyclassdecodingformeg |