Ngram-Derived Pattern Recognition for the Detection and Prediction of Epileptic Seizures

Eftekhar, Amir; Juffali, Walid; El-Imad, Jamil; Constandinou, Timothy G.; Toumazou, C.

doi:10.1371/journal.pone.0096235

Cited by 41 publications

(11 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Likewise, research conducted by Zhang & Parhi (2016) and Parvez & Paul (2017) also implied the use of zero SPH, which will not be compared directly with our results. Among the rest of the studies listed in Table 5, Eftekhar et al (2014) had a very good prediction sensitivity of 90.95%…”

Section: Resultsmentioning

confidence: 98%

“…Aarabi & He (2017) recently extracted six univariate and bivariate features, including correlation dimension, correlation entropy, noise level, Lempel-Ziv complexity, largest Lyapunov exponent, and nonlinear interdependence, and achieved a comparable sensitivity of 86.7% and lower FPR of 0.126/h. On the basis of the assumption that future events depend on a number of previous events, a multiresolution N -gram on amplitude patterns was used as features (Eftekhar et al, 2014). After optimization of the feature set per patient, this method yielded a high sensitivity of 90.95% and a low FPR of 0.06/h on the Freiburg Hospital dataset.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Convolutional neural networks for seizure prediction using intracranial and scalp electroencephalogram

et al. 2018

View full text Add to dashboard Cite

Seizure prediction has attracted growing attention as one of the most challenging predictive data analysis efforts to improve the life of patients with drug-resistant epilepsy and tonic seizures. Many outstanding studies have reported great results in providing sensible indirect (warning systems) or direct (interactive neural stimulation) control over refractory seizures, some of which achieved high performance. However, to achieve high sensitivity and a low false prediction rate, many of these studies relied on handcraft feature extraction and/or tailored feature extraction, which is performed for each patient independently. This approach, however, is not generalizable, and requires significant modifications for each new patient within a new dataset. In this article, we apply convolutional neural networks to different intracranial and scalp electroencephalogram (EEG) datasets and propose a generalized retrospective and patient-specific seizure prediction method. We use the short-time Fourier transform on 30-s EEG windows to extract information in both the frequency domain and the time domain. The algorithm automatically generates optimized features for each patient to best classify preictal and interictal segments. The method can be applied to any other patient from any dataset without the need for manual feature extraction. The proposed approach achieves sensitivity of 81.4%, 81.2%, and 75% and a false prediction rate of 0.06/h, 0.16/h, and 0.21/h on the Freiburg Hospital intracranial EEG dataset, the Boston Children's Hospital-MIT scalp EEG dataset, and the American Epilepsy Society Seizure Prediction Challenge dataset, respectively. Our prediction method is also statistically better than an unspecific random predictor for most of the patients in all three datasets.

show abstract

Section: Resultsmentioning

confidence: 98%

Section: Introductionmentioning

confidence: 99%

Convolutional neural networks for seizure prediction using intracranial and scalp electroencephalogram

et al. 2018

View full text Add to dashboard Cite

show abstract

“…We calculated co-occurrences of terms within each section by using the word level n-grams. N-gram method is a contiguous sequence of n items (syllables, letters, and words) from a given sequence of text and it helps to reduce the problems which arise from identifying entities presented by groups of words (Larynx Cancer, Mass of Neck) [28]. For each "n", where "n" is number of words in the entity, our algorithm passed through the data collection once and measured the frequencies of unigrams, 2-grams, 3-grams, and 4-grams).…”

Section: Information Extractionmentioning

confidence: 99%

Kulak Burun Boğaz Taburcu Notlarından Birliktelik Kurallarının Çıkartılması

Yolcular¹,

Bilge²,

Samur³

2018

Bilişim Teknolojileri Dergisi

View full text Add to dashboard Cite

The objectives of this study were to structure otorhinolaryngology discharge summaries with text mining methods and analyze structured data and extract relational rules using Association Rule Mining (ARM). In this study, we used otorhinolaryngology discharge notes. We first developed a dictionary-based information extraction (IE) module in order to annotate medical entities. Later we extracted the annotated entities, and transformed all documents into a data table. We applied ARM Apriori algorithm to the final dataset, and identified interesting patterns and relationships between the entities as association rules for predicting the treatment procedure for patients. The IE module's precision, recall, and f-measure were 95.1%, 84.5%, and 89.2%, respectively. A total of fifteen association rules were found by selecting the top ranking rules obtained from the ARM analysis. These fifteen rules were reviewed by a domain expert, and the validity of these rules was examined in the PubMed literature. The results showed that the association rules are mostly endorsed by the literature. Although our system focuses on the domain of otorhinolaryngology, we believe the same methodology can be applied to other medical domains and extracted rules can be used for clinical decision support systems and in patient care.

show abstract

“…The SPH is 0 in some studies 10 , 17 , 18 , and ranges from 10 s to 4 h in other studies 8 , 9 , 19 . SOP also varies from 5 min to 1 h in several studies 8 , 20 , 21 , 22 . The SPH mentioned in Park et al 23 is actually the SOP definition used in this study, which easily leads to misunderstandings and confusion.…”

Section: Introductionmentioning

confidence: 99%

Power efficient refined seizure prediction algorithm based on an enhanced benchmarking

Wang

Yang

2021

Sci Rep

View full text Add to dashboard Cite

Deep learning techniques have led to significant advancements in seizure prediction research. However, corresponding used benchmarks are not uniform in published results. Moreover, inappropriate training and evaluation processes used in various work create overfitted models, making prediction performance fluctuate or unreliable. In this study, we analyzed the various data preparation methods, dataset partition methods in related works, and explained the corresponding impacts to the prediction algorithms. Then we applied a robust processing procedure that considers the appropriate sampling parameters and the leave-one-out cross-validation method to avoid possible overfitting and provide prerequisites for ease benchmarking. Moreover, a deep learning architecture takes advantage of a one-dimension convolutional neural network and a bi-directional long short-term memory network is proposed for seizure prediction. The architecture achieves 77.6% accuracy, 82.7% sensitivity, and 72.4% specificity, and it outperforms the indicators of other prior-art works. The proposed model is also hardware friendly; it has 6.274 k parameters and requires only 12.825 M floating-point operations, which is advantageous for memory and power constrained device implementations.

show abstract

Ngram-Derived Pattern Recognition for the Detection and Prediction of Epileptic Seizures

Cited by 41 publications

References 39 publications

Convolutional neural networks for seizure prediction using intracranial and scalp electroencephalogram

Convolutional neural networks for seizure prediction using intracranial and scalp electroencephalogram

Kulak Burun Boğaz Taburcu Notlarından Birliktelik Kurallarının Çıkartılması

Power efficient refined seizure prediction algorithm based on an enhanced benchmarking

Contact Info

Product

Resources

About