Utilization of temporal autoencoder for semi-supervised intracranial EEG clustering and classification
Abstract Manual visual review, annotation and categorization of electroencephalography (EEG) is a time-consuming task that is often associated with human bias and requires trained electrophysiology experts with specific domain knowledge. This challenge is now compounded by development of measurement...
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-01-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-27978-6 |
| _version_ | 1828063765327773696 |
|---|---|
| author | Petr Nejedly Vaclav Kremen Kamila Lepkova Filip Mivalt Vladimir Sladky Tereza Pridalova Filip Plesinger Pavel Jurak Martin Pail Milan Brazdil Petr Klimes Gregory Worrell |
| author_facet | Petr Nejedly Vaclav Kremen Kamila Lepkova Filip Mivalt Vladimir Sladky Tereza Pridalova Filip Plesinger Pavel Jurak Martin Pail Milan Brazdil Petr Klimes Gregory Worrell |
| author_sort | Petr Nejedly |
| collection | DOAJ |
| description | Abstract Manual visual review, annotation and categorization of electroencephalography (EEG) is a time-consuming task that is often associated with human bias and requires trained electrophysiology experts with specific domain knowledge. This challenge is now compounded by development of measurement technologies and devices allowing large-scale heterogeneous, multi-channel recordings spanning multiple brain regions over days, weeks. Currently, supervised deep-learning techniques were shown to be an effective tool for analyzing big data sets, including EEG. However, the most significant caveat in training the supervised deep-learning models in a clinical research setting is the lack of adequate gold-standard annotations created by electrophysiology experts. Here, we propose a semi-supervised machine learning technique that utilizes deep-learning methods with a minimal amount of gold-standard labels. The method utilizes a temporal autoencoder for dimensionality reduction and a small number of the expert-provided gold-standard labels used for kernel density estimating (KDE) maps. We used data from electrophysiological intracranial EEG (iEEG) recordings acquired in two hospitals with different recording systems across 39 patients to validate the method. The method achieved iEEG classification (Pathologic vs. Normal vs. Artifacts) results with an area under the receiver operating characteristic (AUROC) scores of 0.862 ± 0.037, 0.879 ± 0.042, and area under the precision-recall curve (AUPRC) scores of 0.740 ± 0.740, 0.714 ± 0.042. This demonstrates that semi-supervised methods can provide acceptable results while requiring only 100 gold-standard data samples in each classification category. Subsequently, we deployed the technique to 12 novel patients in a pseudo-prospective framework for detecting Interictal epileptiform discharges (IEDs). We show that the proposed temporal autoencoder was able to generalize to novel patients while achieving AUROC of 0.877 ± 0.067 and AUPRC of 0.705 ± 0.154. |
| first_indexed | 2024-04-10T22:49:29Z |
| format | Article |
| id | doaj.art-82a61424d728449dae660fdcc140490f |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-10T22:49:29Z |
| publishDate | 2023-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-82a61424d728449dae660fdcc140490f2023-01-15T12:09:39ZengNature PortfolioScientific Reports2045-23222023-01-0113111310.1038/s41598-023-27978-6Utilization of temporal autoencoder for semi-supervised intracranial EEG clustering and classificationPetr Nejedly0Vaclav Kremen1Kamila Lepkova2Filip Mivalt3Vladimir Sladky4Tereza Pridalova5Filip Plesinger6Pavel Jurak7Martin Pail8Milan Brazdil9Petr Klimes10Gregory Worrell111St Department of Neurology, Faculty of Medicine, Masaryk UniversityDepartment of Neurology, Mayo Clinic, Mayo Systems Electrophysiology LaboratoryDepartment of Neurology, Mayo Clinic, Mayo Systems Electrophysiology LaboratoryDepartment of Neurology, Mayo Clinic, Mayo Systems Electrophysiology LaboratoryDepartment of Neurology, Mayo Clinic, Mayo Systems Electrophysiology LaboratoryInstitute of Scientific Instruments, The Czech Academy of SciencesInstitute of Scientific Instruments, The Czech Academy of SciencesInstitute of Scientific Instruments, The Czech Academy of Sciences1St Department of Neurology, Faculty of Medicine, Masaryk University1St Department of Neurology, Faculty of Medicine, Masaryk UniversityInstitute of Scientific Instruments, The Czech Academy of SciencesDepartment of Neurology, Mayo Clinic, Mayo Systems Electrophysiology LaboratoryAbstract Manual visual review, annotation and categorization of electroencephalography (EEG) is a time-consuming task that is often associated with human bias and requires trained electrophysiology experts with specific domain knowledge. This challenge is now compounded by development of measurement technologies and devices allowing large-scale heterogeneous, multi-channel recordings spanning multiple brain regions over days, weeks. Currently, supervised deep-learning techniques were shown to be an effective tool for analyzing big data sets, including EEG. However, the most significant caveat in training the supervised deep-learning models in a clinical research setting is the lack of adequate gold-standard annotations created by electrophysiology experts. Here, we propose a semi-supervised machine learning technique that utilizes deep-learning methods with a minimal amount of gold-standard labels. The method utilizes a temporal autoencoder for dimensionality reduction and a small number of the expert-provided gold-standard labels used for kernel density estimating (KDE) maps. We used data from electrophysiological intracranial EEG (iEEG) recordings acquired in two hospitals with different recording systems across 39 patients to validate the method. The method achieved iEEG classification (Pathologic vs. Normal vs. Artifacts) results with an area under the receiver operating characteristic (AUROC) scores of 0.862 ± 0.037, 0.879 ± 0.042, and area under the precision-recall curve (AUPRC) scores of 0.740 ± 0.740, 0.714 ± 0.042. This demonstrates that semi-supervised methods can provide acceptable results while requiring only 100 gold-standard data samples in each classification category. Subsequently, we deployed the technique to 12 novel patients in a pseudo-prospective framework for detecting Interictal epileptiform discharges (IEDs). We show that the proposed temporal autoencoder was able to generalize to novel patients while achieving AUROC of 0.877 ± 0.067 and AUPRC of 0.705 ± 0.154.https://doi.org/10.1038/s41598-023-27978-6 |
| spellingShingle | Petr Nejedly Vaclav Kremen Kamila Lepkova Filip Mivalt Vladimir Sladky Tereza Pridalova Filip Plesinger Pavel Jurak Martin Pail Milan Brazdil Petr Klimes Gregory Worrell Utilization of temporal autoencoder for semi-supervised intracranial EEG clustering and classification Scientific Reports |
| title | Utilization of temporal autoencoder for semi-supervised intracranial EEG clustering and classification |
| title_full | Utilization of temporal autoencoder for semi-supervised intracranial EEG clustering and classification |
| title_fullStr | Utilization of temporal autoencoder for semi-supervised intracranial EEG clustering and classification |
| title_full_unstemmed | Utilization of temporal autoencoder for semi-supervised intracranial EEG clustering and classification |
| title_short | Utilization of temporal autoencoder for semi-supervised intracranial EEG clustering and classification |
| title_sort | utilization of temporal autoencoder for semi supervised intracranial eeg clustering and classification |
| url | https://doi.org/10.1038/s41598-023-27978-6 |
| work_keys_str_mv | AT petrnejedly utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT vaclavkremen utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT kamilalepkova utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT filipmivalt utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT vladimirsladky utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT terezapridalova utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT filipplesinger utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT paveljurak utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT martinpail utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT milanbrazdil utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT petrklimes utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification AT gregoryworrell utilizationoftemporalautoencoderforsemisupervisedintracranialeegclusteringandclassification |