Epileptic seizure detection using multiwavelet transform based approximate entropy and artificial neural networks

Optimal features for online seizure detection

“…From Fig. 1, the recorded EEG data is high pass filtered with a cut-off frequency of 0.16 Hz to ensure all data is inline with the International Federation of Clinical Neurophysiology recommendation [24].…”

A novel method of EEG data acquisition, feature extraction and feature space creation for early detection of epileptic seizures

“…Having a simple training set makes the classifier's training phase faster. Also, most studies [11,12,15,[18][19][20] use a small data set [21] having a total of only 39 minutes seizure data. We tested our method on an EEG dataset of 22 patients (5 males and 15 females ages 1-22) collected at the Children's Hospital Boston (available as CHB-MIT database) [1,2].…”

“…Patient-related artifacts are common biological signals that can disturb the EEG signals, such as electrical activity from heartbeats, which produce sharp wave artifacts. Technical-related artifacts are related to malfunctioning electrodes or electromagnetic interferences [12,13]. For better EEG interpretation, one can exclude artifacts from an EEG recording; else, their characteristics should be studied in detail, in order to distinguish them from abnormal EEG.…”

“…Since EEG signals are transient, and highly dynamic, feature vectors are formed for each time epoch. The Multi-level Wavelet Decomposition is a popular technique used in previous studies [3,10,11,12,13,[15][16][17][18] for feature vector extraction. This technique decomposes an EEG signal into a number of sub-band signals each depicting a different waveform morphology within a particular frequency range.…”

Automated early detection of drops in commercial egg production using neural networks

“…Therefore, the general scheme includes an input layer, zero or more hidden layers, and an output layer (Kruse et al 2013), as shown in Figure 1. ANNs, and particularly MLPs, have been used in different knowledge areas and shown remarkable results (Guo et al 2010;Samborska et al 2014;Kalhor et al 2016).…”