A Deep Learning Architecture for Epileptic Seizure Classification Based on Object and Action Recognition

Karácsony, Tamás; Loesch-Biffar, Anna Mira; Vollmar, Christian; Noachtar, Soheyl; Cunha, João Paulo Silva

doi:10.1109/icassp40776.2020.9054649

Cited by 19 publications

(17 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Machine learning methods with bags of visual words help develop human action recognition applications [34]. Besides these popular techniques, Human action recognition uses other techniques like the LSTM network, Epileptic seizure classification, deep transfer learning approach, and hybrid transfer learning model [35][36][37][38]. Some researchers use the hybrid approach by merging the old techniques with the proposed ones.…”

Section: Literature Reviewmentioning

confidence: 99%

Suspicious Activity Recognition Using Proposed Deep L4-Branched-Actionnet With Entropy Coded Ant Colony System Optimization

et al. 2021

View full text Add to dashboard Cite

Intelligent visual surveillance systems are attracting much attention from research and industry. The invention of smart surveillance cameras with greater processing power has now been the leading stakeholder, making it conceivable to design intelligent visual surveillance systems. It is possible to assure the safety of people in both homes and public places. This work aims to distinguish the suspicious activities for surveillance environments. For this, a 63 layers deep CNN model is suggested and named "L4-Branched-ActionNet". The suggested CNN structure is centered on the alteration of AlexNet with added four blanched sub-structures. The developed framework is first transformed into a pre-trained framework by conducting its training on an object detection dataset called CIFAR-100 with the SoftMax function. The dataset for suspicious activity recognition is then forwarded to this pretrained model for feature acquisition. The acquired deep features are subjected to feature subset optimization. These extracted features are first coded by applying entropy and then an ant colony system (ACS) is utilized on the entropy-based coded features for optimization. The configured features are then fed into numerous SVM and KNN based classification models. The cubic SVM has the highest efficiency scores, with a performance of 0.9924 in terms of accuracy. The proposed model is also validated on the Weizmann action dataset and attained an accuracy of 0.9796. The successful findings indicate the suggested work's soundness.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Suspicious Activity Recognition Using Proposed Deep L4-Branched-Actionnet With Entropy Coded Ant Colony System Optimization

et al. 2021

View full text Add to dashboard Cite

show abstract

“…In this work, we take advantage of recent deep learning frameworks in computer vision for directly analyzing patients' semiology, focusing particularly on the body pose and face regions. Several related works have been proposed recently [2,4,17]. In [2], the authors use semiological signs from face, body, and hands to classify epilepsy with convolution neural networks (CNNs) and recurrent neural networks (RNNs).…”

Section: Introductionmentioning

confidence: 99%

“…The work in [4] also utilized similar strategy with pre-trained CNN features combined with RNNs for analyzing and fusing the information from face and body pose. The method proposed in [17] used a I3D [7] backbone to extract spatio-temporal features followed by RNNs as the classifier. Rather than using the standard combination framework like CNN-RNN architectures, in this work, we propose to leverage the recent powerful graph convolutional networks (GCNs) for seizure classification.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Multi-Stream Approach for Seizure Classification with Knowledge Distillation

Hou

McGonigal

Bartolomei

et al. 2021

2021 17th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS)

View full text Add to dashboard Cite

In this work, we propose a multi-stream approach with knowledge distillation to classify epileptic seizures and psychogenic non-epileptic seizures. The proposed framework utilizes multi-stream information from keypoints and appearance from both body and face. We take the detected keypoints through time as spatio-temporal graph and train it with an adaptive graph convolutional networks to model the spatio-temporal dynamics throughout the seizure event. Besides, we regularize the keypoint features with complementary information from the appearance stream by imposing a knowledge distillation mechanism. We demonstrate the effectiveness of our approach by conducting experiments on real-world seizure videos. The experiments are conducted by both seizure-wise cross validation and leaveone-subject-out validation, and with the proposed model, the performances of the F1-score/accuracy are 0.89/0.87 for seizure-wise cross validation, and 0.75/0.72 for leaveone-subject-out validation.

show abstract

“…Neural networks can overcome these challenges by automatically learning features from the training data that are more robust to variations in the data distribution. Most related works using neural networks focus on classifying the epilepsy type by predicting the location of the epileptogenic zone (EZ), e.g., "temporal lobe epilepsy" vs. "extratemporal lobe epilepsy", from short (≤ 2 s) snippets extracted from videos of one or more seizures [3,2,1,16,13]. Typically, this is done as follows.…”

Section: Introductionmentioning

confidence: 99%

Transfer Learning of Deep Spatiotemporal Networks to Model Arbitrarily Long Videos of Seizures

Pérez-García,

Scott,

Sparks

et al. 2021

Preprint

View full text Add to dashboard Cite

Detailed analysis of seizure semiology, the symptoms and signs which occur during a seizure, is critical for management of epilepsy patients. Inter-rater reliability using qualitative visual analysis is often poor for semiological features. Therefore, automatic and quantitative analysis of video-recorded seizures is needed for objective assessment. We present GESTURES, a novel architecture combining convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to learn deep representations of arbitrarily long videos of epileptic seizures. We use a spatiotemporal CNN (STCNN) pre-trained on large human action recognition (HAR) datasets to extract features from short snippets (≈ 0.5 s) sampled from seizure videos. We then train an RNN to learn seizure-level representations from the sequence of features. We curated a dataset of seizure videos from 68 patients and evaluated GESTURES on its ability to classify seizures into focal onset seizures (FOSs) (N = 106) vs. focal to bilateral tonic-clonic seizures (TCSs) (N = 77), obtaining an accuracy of 98.9% using bidirectional long short-term memory (BLSTM) units. We demonstrate that an STCNN trained on a HAR dataset can be used in combination with an RNN to accurately represent arbitrarily long videos of seizures. GESTURES can provide accurate seizure classification by modeling sequences of semiologies.

show abstract

A Deep Learning Architecture for Epileptic Seizure Classification Based on Object and Action Recognition

Cited by 19 publications

References 23 publications

Suspicious Activity Recognition Using Proposed Deep L4-Branched-Actionnet With Entropy Coded Ant Colony System Optimization

Suspicious Activity Recognition Using Proposed Deep L4-Branched-Actionnet With Entropy Coded Ant Colony System Optimization

A Multi-Stream Approach for Seizure Classification with Knowledge Distillation

Transfer Learning of Deep Spatiotemporal Networks to Model Arbitrarily Long Videos of Seizures

Contact Info

Product

Resources

About