TY - JOUR
T1 - ENHANCED SALP SWARM ALGORITHM BASED ON CONVOLUTIONAL NEURAL NETWORK OPTIMIZATION FOR AUTOMATIC EPILEPSY DETECTION
AU - Sunaryono, Dwi
AU - Sarno, Riyanarto
AU - Siswantoro, Joko
AU - Raharjo, Agus Budi
AU - Sabilla, Shoffi Izza
AU - Susilo, Rahadian Indarto
AU - Rekha, Kana
N1 - Publisher Copyright:
© 2022 Little Lion Scientific. All rights reserved.
PY - 2022/10/15
Y1 - 2022/10/15
N2 - Epilepsy is a neurological disorder that occurs due to abnormal activity in the brain. Symptoms can vary, such as uncontrolled movements, muscle stiffness, difficulty breathing, loss of consciousness, and even death. Therefore, the multichannel electroencephalogram (EEG) is very important to understand the pattern of seizure occurrence and non-seizure in epilepsy. In this paper, we determine an automatic epilepsy detection method using enhanced Salp Swarm Algorithm (SSA) CNN-based of EEG signals. The signal is transformed into Low Pass Filter (LPF) and High Pass Filter (HPF) with one level, frequencies, and scales using Wavelet Transform. Enhanced SSA was used to determine the number of neurons and the appropriate number of convolution layers in the CNN algorithm for classifying two classes (epilepsy and epilepsy with seizure) using the CHB-MIT dataset from Children's Hospital Boston. The results of the study show that the proposed method produces the highest accuracy of 99.15% and 89.04% of average accuracy. This result is obtained with a computation time on testing data of 0.0001 seconds using a high-end computer. Enhanced SSA was proven to increase the performance of CNN of 81.13%. The proposed method can be used in the automatic detection of epilepsy.
AB - Epilepsy is a neurological disorder that occurs due to abnormal activity in the brain. Symptoms can vary, such as uncontrolled movements, muscle stiffness, difficulty breathing, loss of consciousness, and even death. Therefore, the multichannel electroencephalogram (EEG) is very important to understand the pattern of seizure occurrence and non-seizure in epilepsy. In this paper, we determine an automatic epilepsy detection method using enhanced Salp Swarm Algorithm (SSA) CNN-based of EEG signals. The signal is transformed into Low Pass Filter (LPF) and High Pass Filter (HPF) with one level, frequencies, and scales using Wavelet Transform. Enhanced SSA was used to determine the number of neurons and the appropriate number of convolution layers in the CNN algorithm for classifying two classes (epilepsy and epilepsy with seizure) using the CHB-MIT dataset from Children's Hospital Boston. The results of the study show that the proposed method produces the highest accuracy of 99.15% and 89.04% of average accuracy. This result is obtained with a computation time on testing data of 0.0001 seconds using a high-end computer. Enhanced SSA was proven to increase the performance of CNN of 81.13%. The proposed method can be used in the automatic detection of epilepsy.
KW - CHB-MIT
KW - Convolutional Neural Network
KW - Epilepsy
KW - Salp Swarm Algorithm
KW - Wavelet Transform
UR - http://www.scopus.com/inward/record.url?scp=85140404496&partnerID=8YFLogxK
M3 - Article
AN - SCOPUS:85140404496
SN - 1992-8645
VL - 100
SP - 5615
EP - 5623
JO - Journal of Theoretical and Applied Information Technology
JF - Journal of Theoretical and Applied Information Technology
IS - 19
ER -