Machine learning untuk klasifikasi kelainan jantung melalui identifikasi sinyal EKG

Alejandro, Ariel Dimitri (2022) Machine learning untuk klasifikasi kelainan jantung melalui identifikasi sinyal EKG. Bachelor thesis, Universitas Pelita Harapan.

Preview	Text (Title) Title.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (198kB) | Preview
Preview	Text (Abstract) Abstract.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (485kB) | Preview
Preview	Text (ToC) ToC.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (644kB) | Preview
	Text (Chapter1) Chapter1.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (1MB)
	Text (Chapter2) Chapter2.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (2MB)
	Text (Chapter3) Chapter3.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (875kB)
	Text (Chapter4) Chapter4.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (5MB)
	Text (Chapter5) Chapter5.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (456kB)
	Text (Bibliography) Bibliography.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (685kB)
	Text (Appendices) Appendices.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (1MB)

Abstract

Sinyal elektrokardiogram (EKG) merupakan sebuah rekaman detak jantung dengan menggunakan elektroda dan dapat mendeteksi adanya perubahan sinyal listrik pada setiap detak jantung. Sinyal EKG ini digunakan untuk menyelidiki beberapa jenis fungsi jantung yang dinilai tidak normal, termasuk aritmia dan gagal jantung (Congestive Heart Failure). Dalam skripsi ini metode yang diusulkan digunakan untuk mengklasifikasikan sinyal EKG adalah dengan menggunakan teknik klasifikasi dari aplikasi Diagnostic Feature Designer yang digunakan untuk mengembangkan fitur dan mengevaluasi segala indikator menggunakan graphical interface, dan Classification Learner yang digunakan untuk melatih model dengan tujuan pengklasifikasian data. Pada akhirnya model yang akan digunakan adalah Discriminant Analysis Classification yang di mana metode ini dapat mengasumsikan bahwa setiap kelas yang berbeda akan menghasilkan data dengan distribusi Gaussian yang berbeda. Hasil percobaan dari 162 data latih menunjukan bahwa akurasi dari Discriminant Analysis Classification adalah 95.8%. Dibandingkan Classifier lain Discriminant Analysis Classification akurasinya paling tinggi, di mana hasil akhir eksperimen menggunakan 10 data uji dari sinyal luar data latih untuk sinyal Aritmia, Congestive Heart Failure, dan Sinus Normal dengan nilai kesuksesan sebesar 100%, 83.3%, dan 100%. / Electrocardiogram (ECG) is a recording of heart rate using electrodes and can detect changes in electrical signal at each heartbeat. This ECG signal is used to investigate several types of heart function that are considered abnormal, including arrhythmias and heart failure (Congestive Heart Failure). In this report, the proposed method used to classify ECG signals is to use the classification technique from the Diagnostic Feature Designer application which is used to develop features and evaluate all indicators using a graphical interface, and the Classification Learner which is used to train the model for the purpose of classifying data. In the end, the method that will be used is Discriminant Analysis Classification where this method can assume that each different class will produce data with a different Gaussian distribution. The experimental results from 162 training data show that the accuracy of the Discriminant Analysis Classification is 95.8%. Compared to other Classifiers, Discriminant Analysis Classification has the highest accuracy, where the final results of the experiment use 10 test data from signals outside the training data for Arrhythmia, Congestive Heart Failure, and Normal Sinus signals with success values of 100%, 83.3%, and 100%.

Actions (login required)

View Item