A DWT-entropy-ANN based architecture for epilepsy diagnosis using EEG signals

AlSharabi, Khalil; Ibrahim, Sutrisno; Djemal, Ridha; Alsuwailem, Abdullah

doi:10.1109/atsip.2016.7523093

Cited by 22 publications

(18 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Detection of Alzheimer's also achieved a classification rate of 100% by using a feed forward neural network and DCT [13]. The usage of DWT, Shannon entropy and FFNN [15] provided an accuracy of 100% for the diagnosis of Epilepsy. An accuracy of 94% was obtained for the detection of Parkinson's disease using SPECT images and sequential grass-fire algorithm [16].…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Diagnosis of Brain Diseases using Neural Networks

2019

IJITEE

View full text Add to dashboard Cite

Intensification in the occurrence of brain diseases and the need for the initial diagnosis for ailments like Tumor, Alzheimer’s, Epilepsy and Parkinson’s has riveted the attention of researchers. Machine learning practices, specifically deep learning, is considered as a beneficial diagnostic tool. Deep learning approaches to neuroimaging will assist computer-aided analysis of neurological diseases. Feature extraction of neuroimages carried out using Artificial Neural Networks leads to better diagnoses. In this study, all the brain diseases are revisited to consolidate the methodologies carried out by various authors in the literature.

show abstract

Section: Discussionmentioning

confidence: 99%

“…However, wavelet transform requires a lot of storage and computational cost is high. Khalil et al [15] proposed a technique which made use of Discrete wavelet transform (DWT) for splitting the EEG signal. The signal comprised of various frequency sub-band which are: alpha, beta, gamma, delta and theta.…”

Section: Diagnosis Of Epilepsymentioning

confidence: 99%

Diagnosis of Brain Diseases using Neural Networks

2019

IJITEE

View full text Add to dashboard Cite

show abstract

“…In details, its algorithm initializes the value of K from 1 (setting as initial iteration value). After loading data, iteration from initial K =1 (generally) to the total number of training data point while distances specifically Euclidean distance between test data and each row of training data is measured and sorted in ascending order to get topmost k rows from the sorted array and the most frequent class is returned as the predicted class [22]. The value of K was tuned, and the k for best efficiency was chosen in the classifier model in this research to reduce overfitting.…”

Section: K-nearest Neighbours (Knn)mentioning

confidence: 99%

“…Ensemble learning models, such as Random forests are made of individual decision trees with a logic of group of weak learners to finally make a strong learner while the decision trees operate as divided or conquer. A class is predicted from every decision tree and a final class is predicted by model depending on their vote [22]. Two parameters were tuned in the RFC models in this study, namely, 'n_estimate', which implies the number of trees in the forest and 'max-depth' which signifies the depth of each tree.…”

Section: Random Forest Classifier (Rfc)mentioning

confidence: 99%

Diagnosis of Tobacco Addiction using Medical Signal: An EEG-based Time-Frequency Domain Analysis Using Machine Learning

Hasan¹,

Hasan²,

Alsubaie³

et al. 2021

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

Addiction such as tobacco smoking affects the human brain and thus causes significant changes in the brainwaves. The changes in brain wave due to smoking can be identified by focusing on changes in electroencephalogram pattern, extracting different time-frequency domain features. In this aspect, a laboratory-based study has been presented in this paper, for assessing the brain signal changes due to the tobacco addiction. Four classifier models, namely, Logistic Regression (LR), K-Nearest Neighbor (KNN), Support Vector Machine (SVM) and Random Forest Classifier (RFC) were trained and tested for assessing the performance of the time domain, frequency domain and fusion of time-frequency domain features, with a five-fold cross-validation. Four different performance measures (sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve) were used to measure the overall performance, and the results suggested that the classifiers based on time-frequency domain features perform the best while using combinedly. Using the utilized fusion of the time-frequency domain features, the classification models can identify the smoker group with an accuracy ranged from (86.5-91.3%), where the RFC shows the best accuracy of 91.3%, which is higher than the three other classifiers models.

show abstract

“…There are several studies that use different entropy measures for EEG signals such as Rényi entropy, 25 diffusion entropy, 26 and Kraskov entropy. 27 EEG complexity measures as feature vectors are also used for discrimination of different kinds of disorders [28][29][30][31] via multiple classification methods such as SVM 32 and ANN. 33 The literature includes several studies on SUD that are classified via different ML approaches.…”

Section: Introductionmentioning

confidence: 99%

Entropy: A Promising EEG Biomarker Dichotomizing Subjects With Opioid Use Disorder and Healthy Controls

et al. 2020

View full text Add to dashboard Cite

Electroencephalography (EEG) signals are known to be nonstationary and often multicomponential signals containing information about the condition of the brain. Since the EEG signal has complex, nonlinear, nonstationary, and highly random behaviour, numerous linear feature extraction methods related to the short-time windowing technique do not satisfy higher classification accuracy. Since biosignals are highly subjective, the symptoms may appear at random in the time scale and very small variations in EEG signals may depict a definite type of brain abnormality it is valuable and vital to extract and analyze the EEG signal parameters using computers. The challenge is to design and develop signal processing algorithms that extract this subtle information and use it for diagnosis, monitoring, and treatment of subjects suffering from psychiatric disorders. For this purpose, finite impulse response–based filtering process was employed rather than traditional time and frequency domain methods. Finite impulse response subbands were analyzed further to obtain feature vectors of different entropy markers and these features were fed into a classifier namely multilayer perceptron. The performances of the classifiers were finally compared considering overall classification accuracies, area under receiver operating characteristic curve scores. Our results underline the potential benefit of the introduced methodology is promising and is to be treated as a clinical interface in dichotomizing substance use disorders subjects and for other medical data analysis studies. The results also indicate that entropy estimators can distinguish normal and opioid use disorder subjects. EEG data and theta frequency band have distinctive capability for almost all types of entropies while nonextensive Tsallis entropy outperforms compared with other types of entropies.

show abstract

A DWT-entropy-ANN based architecture for epilepsy diagnosis using EEG signals

Cited by 22 publications

References 15 publications

Diagnosis of Brain Diseases using Neural Networks

Diagnosis of Brain Diseases using Neural Networks

Diagnosis of Tobacco Addiction using Medical Signal: An EEG-based Time-Frequency Domain Analysis Using Machine Learning

Entropy: A Promising EEG Biomarker Dichotomizing Subjects With Opioid Use Disorder and Healthy Controls

Contact Info

Product

Resources

About