Program Deteksi Penyakit Jantung Fibrilasi Atrium (FA) pada Rekaman Elektrokardiograf (EKG) Melalui Interval RR

Nurhidaya Nurhidaya(1*), Wira Bahari Nurdin(2), Eko Juarlin(3),

(1) Prodi Fisika Universitas Hasanuddin
(2) Prodi Fisika Universitas Hasanuddin
(3) Prodi Fisika Universitas Hasanuddin
(*) Corresponding Author



Abstract


Penelitian tentang  deteksi penyakit jantung Fibrilasi Atrium (FA) pada rekaman elektrokardiograf (EKG) melalui interval RR telah berhasil dilaksanakan. Penelitian ini memanfaatkan karakter fitur statistik interval RR yang digunakan sebagai fitur dasar dalam  mendeteksi Fibrillasi Atrium (FA) dan membedakan Fibrilasi Atrium (FA) dengan pasien normal. Fitur statistik tersebut terdiri dari rata-rata interval RR, standar deviasi interval RR, median interval RR, dan modus interval RR. Berdasarkan hasil eksperimen, terdapat 3 fitur statistik  interval RR yang paling baik dalam mendeteksi Fibrilasi Atrium (FA). Fitur tersebut adalah rata-rata interval RR, median interval RR, dan modus interval RR. Karakter fitur statistik interval RR pada pasien Fibrilasi Atrium (FA) memiliki nilai yang lebih tinggi jika dibanding dengan pasien normal. Dari 33 kode data Arrhythmia yang diuji, hasilnya diperoleh 18 kode data Arrhythmia yang terdeteksi sebagai Fibrilasi Atrium (FA) dan 15 kode data Arrhythmia terdeteksi bukan (FA). Program deteksi ini memiliki kinerja Sensitivitas 66,67%, Spesifitas 54,54%, dan Akurasi 61,53 %.

Kata Kunci: Fibrilasi Atrium (FA), Interval RR.

 

Research on the detection of Atrial Fibrillation (FA) heart disease on electrocardiograph (EKG) recordings through RR intervals has been successfully implemented.This study utilizes the characteristics of RR interval statistics which are used as basic features in detecting Atrial Fibrillation (FA) and differentiating Atrial Fibrillation (FA) from normal patients. The statistical feature consists of the average interval RR, standard deviation interval RR, median RR interval, and RR interval mode. Based on experimental results, there are 3 statistical features of RR that are the best to detecting Atrial Fibrillation (FA). These features are the average interval RR, the median RR interval, and the RR interval mode. The statistical feature of the RR interval in patients with Atrial Fibrillation (FA) has a higher value when compared to normal patients. Of the 33 codes of Arrhythmia data tested, the results obtained 18 codes of Arrhythmia data that were detected as Atrial Fibrillation (FA) and 15 codes of Arrhythmia data were detected not (FA). This detection program has a sensitivity performance of 66.67%, Specificity of 54.54%, and Accuracy Of 61.53%

Keywords: Fibrilasi Atrium (Fa), Interval Rr.


Full Text:

PDF

References


E. Anzihory dan Nuryani, Sistem Deteksi Fibrilasi Atrium Menggunakan Fitur RR Elektokardiogram Dengan Jaringan Syaraf Tiruan, Jurnal Fisika dan Aplikasinya, vol. 12, no. 2, 2016, pp. 57.

K. Farosi, Nuryani, optimasi fuzzy inference system pada sistem deteksi fibrilasi atrium dengan fitur elektrokardiogram, Jurnal Fisika Dan Aplikasinya, vol. 13, no. 1, 2017, pp. 10.

I. Yansen dan yoga yuniadi, Tata laksana fibrilasi atrium kontrol irama atau laju jantung, vol. 40, no.30, 2013, pp. 172.

R. Mabrouki, B. Khaddoumi, and M. Sayadi, Nonlinear statistical methods for atrial fibrillation detection on electrocardiogram, Institute of Electrical and Electronic Engineers, 2014, pp. 1.

A. Surtono dan G. Ahmad Fauzi, Deteksi miokard infark jantung pada rekaman elektrokardiogram menggunakan elevasi segmen ST, Jurnal Teori dan Aplikasi Fisika, vol 04, no.01, 2016, pp. 3.

A. Ghodrati, B. Murray, S. Marinello, RR interval analysis for detection of atrial fibrillation in ecg monitors, 30th Annual International IEEE EMBS ConferenceInternational,

Canada, August 20-24, 2008, pp. 602.

Wen Zhu, Sensitivity, Specificity, Accuracy, Associated Confidence interval and ROC Analysis with Practical SAS Implementations, Health Care and Life Sciences, website dari https://www.lexjansen.com/nesug/nesug10/hl/hl07.pdf,diakses tanggal 21 Mei 2018, pp. 1-2


Article Metrics

Abstract view : 508 times | PDF view : 123 times

Refbacks

  • There are currently no refbacks.