Klasifikasi Sinyal Electroencephalogram untuk Mendeteksi Kejang Epilepsi Menggunakan Naive Bayes

Putranto, Dhamai Brillianggara (2019) Klasifikasi Sinyal Electroencephalogram untuk Mendeteksi Kejang Epilepsi Menggunakan Naive Bayes. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 06211440000002-Undergraduate_Theses.pdf] Text
06211440000002-Undergraduate_Theses.pdf - Accepted Version
Restricted to Repository staff only until 1 October 2023.

Download (1MB) | Request a copy

Abstract

Epilepsi merupakan salah satu kelainan kronis pada otak yang paling umum terjadi di dunia. Kejang merupakan ciri khas epilepsi, tetapi tidak semua kejang merupakan manifestasi epilepsi. Electroencephalogram (EEG) adalah alat untuk merekam aktivitas listrik di otak, dengan meletakkan elektroda-elektroda pada daerah kulit kepala. Sinyal EEG memiliki distribusi nongaussian, non-stationer, dan non-linier, sehingga diagnosa secara visual tidak cukup untuk mendeteksi adanya kejang epilepsi, oleh sebab itu diperlukan teknik komputerisasi. Sinyal EEG terdiri atas 5 gelombang frekuensi yaitu delta, theta, alpha, beta dan gamma. Dari kelima gelombang frekuensi tersebut hanya gelombang theta, alpha, dan beta yang mengandung informasi kejang. Berdasarkan uraian tersebut, pada penelitian ini akan dilakukan filtrasi sinyal EEG untuk mendapatkan sinyal dengan gelombang theta, alpha, dan beta menggunakan IIR Butherworth Bandpass Filter. Selanjutnya dari masing-masing gelombang dihitung nilai Power Spectral Density (PSD) dan didapatkan variabel maksimum PSD, Minimum PSD, Mean PSD dan Variance PSD yang akan digunakan untuk proses klasifikasi menggunakan Naïve Bayes Classifier. Dari hasil filtrasi dan ekstraksi variabel didapatkan perbedaan yang signifikan antara sinyal kejang dan non-kejang. Dan dari hasil klasifikasi didapatkan nilai akurasi sebesar 93%,, sensitivitas sebesar 80% dan spesitifitas sebesar 96%.
================================================================================================
Epilepsy is one of the most common chronic disorders of the brain in the world. Seizures are characteristic of epilepsy, but not all seizures are epilepsy manifestations. Electroencephalogram (EEG) is a tool to record electrical activity in the brain, by placing electrodes on the scalp area. EEG signals have non-gaussian, nonstationary, and non-linear distributions, so the diagnosis is not visually enough to detect epileptic seizures, and therefore computerization techniques are required. EEG signal consists of 5 frequency waves, namely delta, theta, alpha, beta and gamma. From the five frequency waves are only theta, alpha, and beta waves that contain seizure information. Based on the description, in this research will be filtrated of EEG signal to get signal with theta, alpha, and beta waves using IIR Butherworth Bandpass Filter. Then, Maximum, Minimum, Mean and Variance variabel are obtained by Power Spectral Density Method, which will be used for classification process using Naïve Bayes Classifier. From result of filtration and variable extraction known that there was significant difference between seizure signal and non-seizure. And from the classification results obtained an accuracy of 93%, 80% sensitivity and specificity of 96%.

Item Type: Thesis (Undergraduate)
Additional Information: RSSt 519.536 Put k-1 2018
Uncontrolled Keywords: Bandpass Filter, Electroencephalogram, Epilepsy, Naïve Bayes, Power Spectral Density.
Subjects: H Social Sciences > HD Industries. Land use. Labor > HD108 Classification (Theory. Method. Relation to other subjects )
Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Q Science > QA Mathematics > QA279.5 Bayesian statistical decision theory.
Divisions: Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: Dhamai Brillianggara Putranto
Date Deposited: 25 Nov 2021 02:26
Last Modified: 25 Nov 2021 02:26
URI: http://repository.its.ac.id/id/eprint/61592

Actions (login required)

View Item View Item