Permutation Entropy-Based Interpretability of Convolutional Neural Network Models for Interictal EEG Discrimination of Subjects with Epileptic Seizures vs. Psychogenic Non-Epileptic Seizures
The differential diagnosis of epileptic seizures (ES) and psychogenic non-epileptic seizures (PNES) may be difficult, due to the lack of distinctive clinical features. The interictal electroencephalographic (EEG) signal may also be normal in patients with ES. Innovative diagnostic tools that exploit...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-01-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1099-4300/24/1/102 |
| _version_ | 1797494205118939136 |
|---|---|
| author | Michele Lo Giudice Giuseppe Varone Cosimo Ieracitano Nadia Mammone Giovanbattista Gaspare Tripodi Edoardo Ferlazzo Sara Gasparini Umberto Aguglia Francesco Carlo Morabito |
| author_facet | Michele Lo Giudice Giuseppe Varone Cosimo Ieracitano Nadia Mammone Giovanbattista Gaspare Tripodi Edoardo Ferlazzo Sara Gasparini Umberto Aguglia Francesco Carlo Morabito |
| author_sort | Michele Lo Giudice |
| collection | DOAJ |
| description | The differential diagnosis of epileptic seizures (ES) and psychogenic non-epileptic seizures (PNES) may be difficult, due to the lack of distinctive clinical features. The interictal electroencephalographic (EEG) signal may also be normal in patients with ES. Innovative diagnostic tools that exploit non-linear EEG analysis and deep learning (DL) could provide important support to physicians for clinical diagnosis. In this work, 18 patients with new-onset ES (12 males, 6 females) and 18 patients with video-recorded PNES (2 males, 16 females) with normal interictal EEG at visual inspection were enrolled. None of them was taking psychotropic drugs. A convolutional neural network (CNN) scheme using DL classification was designed to classify the two categories of subjects (ES vs. PNES). The proposed architecture performs an EEG time-frequency transformation and a classification step with a CNN. The CNN was able to classify the EEG recordings of subjects with ES vs. subjects with PNES with 94.4% accuracy. CNN provided high performance in the assigned binary classification when compared to standard learning algorithms (multi-layer perceptron, support vector machine, linear discriminant analysis and quadratic discriminant analysis). In order to interpret how the CNN achieved this performance, information theoretical analysis was carried out. Specifically, the permutation entropy (PE) of the feature maps was evaluated and compared in the two classes. The achieved results, although preliminary, encourage the use of these innovative techniques to support neurologists in early diagnoses. |
| first_indexed | 2024-03-10T01:31:00Z |
| format | Article |
| id | doaj.art-c979813dbd724706a30649a87a42da0c |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-03-10T01:31:00Z |
| publishDate | 2022-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-c979813dbd724706a30649a87a42da0c2023-11-23T13:41:56ZengMDPI AGEntropy1099-43002022-01-0124110210.3390/e24010102Permutation Entropy-Based Interpretability of Convolutional Neural Network Models for Interictal EEG Discrimination of Subjects with Epileptic Seizures vs. Psychogenic Non-Epileptic SeizuresMichele Lo Giudice0Giuseppe Varone1Cosimo Ieracitano2Nadia Mammone3Giovanbattista Gaspare Tripodi4Edoardo Ferlazzo5Sara Gasparini6Umberto Aguglia7Francesco Carlo Morabito8Department of Science Medical and Surgery, University of Catanzaro, 88100 Catanzaro, ItalyDepartment of Neuroscience & Imaging, University G. d’Annunzio Chieti e Pescara, 66100 Chieti, ItalyDICEAM Department, University “Mediterranea” of Reggio Calabria, 89100 Reggio Calabria, ItalyDICEAM Department, University “Mediterranea” of Reggio Calabria, 89100 Reggio Calabria, ItalyRegional Epilepsy Center, Great Metropolitan Hospital “Bianchi-Melacrino-Morelli” of Reggio Calabria, 89124 Reggio Calabria, ItalyDepartment of Science Medical and Surgery, University of Catanzaro, 88100 Catanzaro, ItalyDepartment of Science Medical and Surgery, University of Catanzaro, 88100 Catanzaro, ItalyDepartment of Science Medical and Surgery, University of Catanzaro, 88100 Catanzaro, ItalyDICEAM Department, University “Mediterranea” of Reggio Calabria, 89100 Reggio Calabria, ItalyThe differential diagnosis of epileptic seizures (ES) and psychogenic non-epileptic seizures (PNES) may be difficult, due to the lack of distinctive clinical features. The interictal electroencephalographic (EEG) signal may also be normal in patients with ES. Innovative diagnostic tools that exploit non-linear EEG analysis and deep learning (DL) could provide important support to physicians for clinical diagnosis. In this work, 18 patients with new-onset ES (12 males, 6 females) and 18 patients with video-recorded PNES (2 males, 16 females) with normal interictal EEG at visual inspection were enrolled. None of them was taking psychotropic drugs. A convolutional neural network (CNN) scheme using DL classification was designed to classify the two categories of subjects (ES vs. PNES). The proposed architecture performs an EEG time-frequency transformation and a classification step with a CNN. The CNN was able to classify the EEG recordings of subjects with ES vs. subjects with PNES with 94.4% accuracy. CNN provided high performance in the assigned binary classification when compared to standard learning algorithms (multi-layer perceptron, support vector machine, linear discriminant analysis and quadratic discriminant analysis). In order to interpret how the CNN achieved this performance, information theoretical analysis was carried out. Specifically, the permutation entropy (PE) of the feature maps was evaluated and compared in the two classes. The achieved results, although preliminary, encourage the use of these innovative techniques to support neurologists in early diagnoses.https://www.mdpi.com/1099-4300/24/1/102EEGPNESepilepsymachine learningdeep learningconvolutional neural network |
| spellingShingle | Michele Lo Giudice Giuseppe Varone Cosimo Ieracitano Nadia Mammone Giovanbattista Gaspare Tripodi Edoardo Ferlazzo Sara Gasparini Umberto Aguglia Francesco Carlo Morabito Permutation Entropy-Based Interpretability of Convolutional Neural Network Models for Interictal EEG Discrimination of Subjects with Epileptic Seizures vs. Psychogenic Non-Epileptic Seizures Entropy EEG PNES epilepsy machine learning deep learning convolutional neural network |
| title | Permutation Entropy-Based Interpretability of Convolutional Neural Network Models for Interictal EEG Discrimination of Subjects with Epileptic Seizures vs. Psychogenic Non-Epileptic Seizures |
| title_full | Permutation Entropy-Based Interpretability of Convolutional Neural Network Models for Interictal EEG Discrimination of Subjects with Epileptic Seizures vs. Psychogenic Non-Epileptic Seizures |
| title_fullStr | Permutation Entropy-Based Interpretability of Convolutional Neural Network Models for Interictal EEG Discrimination of Subjects with Epileptic Seizures vs. Psychogenic Non-Epileptic Seizures |
| title_full_unstemmed | Permutation Entropy-Based Interpretability of Convolutional Neural Network Models for Interictal EEG Discrimination of Subjects with Epileptic Seizures vs. Psychogenic Non-Epileptic Seizures |
| title_short | Permutation Entropy-Based Interpretability of Convolutional Neural Network Models for Interictal EEG Discrimination of Subjects with Epileptic Seizures vs. Psychogenic Non-Epileptic Seizures |
| title_sort | permutation entropy based interpretability of convolutional neural network models for interictal eeg discrimination of subjects with epileptic seizures vs psychogenic non epileptic seizures |
| topic | EEG PNES epilepsy machine learning deep learning convolutional neural network |
| url | https://www.mdpi.com/1099-4300/24/1/102 |
| work_keys_str_mv | AT michelelogiudice permutationentropybasedinterpretabilityofconvolutionalneuralnetworkmodelsforinterictaleegdiscriminationofsubjectswithepilepticseizuresvspsychogenicnonepilepticseizures AT giuseppevarone permutationentropybasedinterpretabilityofconvolutionalneuralnetworkmodelsforinterictaleegdiscriminationofsubjectswithepilepticseizuresvspsychogenicnonepilepticseizures AT cosimoieracitano permutationentropybasedinterpretabilityofconvolutionalneuralnetworkmodelsforinterictaleegdiscriminationofsubjectswithepilepticseizuresvspsychogenicnonepilepticseizures AT nadiamammone permutationentropybasedinterpretabilityofconvolutionalneuralnetworkmodelsforinterictaleegdiscriminationofsubjectswithepilepticseizuresvspsychogenicnonepilepticseizures AT giovanbattistagasparetripodi permutationentropybasedinterpretabilityofconvolutionalneuralnetworkmodelsforinterictaleegdiscriminationofsubjectswithepilepticseizuresvspsychogenicnonepilepticseizures AT edoardoferlazzo permutationentropybasedinterpretabilityofconvolutionalneuralnetworkmodelsforinterictaleegdiscriminationofsubjectswithepilepticseizuresvspsychogenicnonepilepticseizures AT saragasparini permutationentropybasedinterpretabilityofconvolutionalneuralnetworkmodelsforinterictaleegdiscriminationofsubjectswithepilepticseizuresvspsychogenicnonepilepticseizures AT umbertoaguglia permutationentropybasedinterpretabilityofconvolutionalneuralnetworkmodelsforinterictaleegdiscriminationofsubjectswithepilepticseizuresvspsychogenicnonepilepticseizures AT francescocarlomorabito permutationentropybasedinterpretabilityofconvolutionalneuralnetworkmodelsforinterictaleegdiscriminationofsubjectswithepilepticseizuresvspsychogenicnonepilepticseizures |