ROC Analysis for detection of Epileptical Seizures using Haralick features of Gamma band

In this study, gamma band (30–60 Hz) is used for detection of epileptical seizures using Haralick features. Most of the previous methods are based on the whole frequency spectrum for detection. This work use only high frequency electroencephalogram (EEG) subband for seizure detection using image descriptors. To convert one dimensional EEG data into image Short-time Fourier transform (STFT) has been used. Gamma band is cut from the time frequency (t-f) plane and Haralick features is used as image descriptors to fed in the decision tree classifier. The results have been evaluated using receiver operating characteristic (ROC) analysis. Maximum area under curve (AUC) of 0.96 is obtained to classify between seizures and healthy. Advantage of this work is rather using whole frequency band it utilizes only a particular band which reduces computational load. It also shows the utility of gamma band in seizure detection.

[1]  Andrew P. Bradley,et al.  The use of the area under the ROC curve in the evaluation of machine learning algorithms , 1997, Pattern Recognit..

[2]  Ahmed Bouridane,et al.  Haralick feature extraction from time-frequency images for epileptic seizure detection and classification of EEG data , 2014, 2014 26th International Conference on Microelectronics (ICM).

[3]  Osman Erogul,et al.  Epileptic EEG detection using the linear prediction error energy , 2010, Expert Syst. Appl..

[4]  P. Geethanjali,et al.  DWT Based Detection of Epileptic Seizure From EEG Signals Using Naive Bayes and k-NN Classifiers , 2016, IEEE Access.

[5]  Robert M. Haralick,et al.  Textural Features for Image Classification , 1973, IEEE Trans. Syst. Man Cybern..

[6]  Rajeev Kumar,et al.  Receiver operating characteristic (ROC) curve for medical researchers , 2011, Indian pediatrics.

[7]  David A. Landgrebe,et al.  A survey of decision tree classifier methodology , 1991, IEEE Trans. Syst. Man Cybern..

[8]  Boualem Boashash,et al.  Principles of time-frequency feature extraction for change detection in non-stationary signals: Applications to newborn EEG abnormality detection , 2015, Pattern Recognit..

[9]  Ke Li,et al.  Epileptic seizure detection in EEG signals using sparse multiscale radial basis function networks and the Fisher vector approach , 2019, Knowl. Based Syst..

[10]  Schreiber,et al.  Quantification of depth of anesthesia by nonlinear time series analysis of brain electrical activity , 2000, Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics.

[11]  L. Cohen,et al.  Time-frequency distributions-a review , 1989, Proc. IEEE.

[12]  K Lehnertz,et al.  Indications of nonlinear deterministic and finite-dimensional structures in time series of brain electrical activity: dependence on recording region and brain state. , 2001, Physical review. E, Statistical, nonlinear, and soft matter physics.

[13]  M. Kemal Kiymik,et al.  Comparison of STFT and wavelet transform methods in determining epileptic seizure activity in EEG signals for real-time application , 2005, Comput. Biol. Medicine.

[14]  S. Moshé,et al.  Scalp EEG Ictal gamma and beta activity during infantile spasms: Evidence of focality , 2017, Epilepsia.

[15]  J. Gotman.  Automatic seizure detection: improvements and evaluation. , 1990, Electroencephalography and clinical neurophysiology.