Deteksi Kelainan Jantung Berdasarkan Suara Detak Jantung Secara Realtime Menggunakan Deep Learning

Hariyadi, Dedi Rachmad (2023) Deteksi Kelainan Jantung Berdasarkan Suara Detak Jantung Secara Realtime Menggunakan Deep Learning. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 07211840000009-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2025. Download (3MB) | Request a copy

Abstract

Jantung merupakan organ terpenting dalam tubuh manusia karena jantung yang memompa darah didalam tubuh manusia. Penyakit jantung merupakan salah satu penyebab kematian terbesar di dunia, Menurut WHO hampir sepertiga dari populasi manusia di dunia yang meninggal diakibatkan oleh penyakit jantung. Untuk itu perlu diketahui keadaan jantung dengan sebuah alat. Auskultasi merupakan sebuah prosedur yang sangat penting dilakukan dalam mendiagnosa sebuah kelainan dalam tubuh melalui bunyi yang dihasilkan dengan cara menempekan stetoskop pada bagian tubuh tertentu. Pemeriksaan bunyi jantung dilakukan pada dada sebelah kiri, jantung normal ditandai dengan irama yang teratur. Saat ini auskultasi dilakukan dengan menggunakan stetoskop manual yang mempunyai kelemahan diantaranya adalah tingkat akurasi yang lemah. Salah satu metode yang dapat digunakan untuk meningkatkan akurasi dalam auskultasi adalah dengan menggunakan bantuan Deep Learning. Sehingga, pada penelitian ini digunakan deep learning yakni CNN (Convolutional Neural Network) untuk mengklasifikasikan kelainan jantung berdasar suara detak jantung. Ekstraksi fitur menggunakan metode MFCC, sedangkan model pelatihan menggunakan Simple CNN dan metode transfer learning dengan 3 arsitektur model yaitu VGG16, VGG19, dan MobileNetV2. Dari pengujian yang dilakukan didapatkan hasil terbaik terdapat pada VGG19 dengan nilai F1-score 0,8812, ROC-AUC score 0,9400, dan akurasi 0.9265
======================================================================================================================================
The heart is the most important organ in the human body because it is the heart that pumps blood in the human body. Heart disease is one of the leading causes of death largest in the world, almost a third of the world's human population who died due to by heart disease. For this reason, it is necessary to know the condition of the heart with a tool. Auscultation is a procedure that is very important in diagnosing a abnormalities in the body through the sound produced by pressing the stethoscope on the certain body parts. Examination of heart sounds is carried out on the left chest, heart Normal is characterized by a regular rhythm. Currently, auscultation is performed using manual stethoscope which has a weakness including the level of accuracy the weak. One method that can be used to improve accuracy in auscultation is to use the help of Deep Learning. So, in this research deep learning, namely CNN (Convolutional Neural Network) is used to Classification of heart defects based on the sound of the heartbeat. Feature extraction uses the MFCC method, while the training model uses Simple CNN and the transfer learning method with 3 model architectures namely VGG16, VGG19, and MobileNetV2. From the tests carried out, the best results were found in VGG19 with an F1-score of 0,8812, ROC-AUC score 0,9400 and an average accuracy of 0.9265

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Jantung, Penyakit Jantung, Deep-Learning, Heart, Heart Disease, Deep-Learning
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. R Medicine > R Medicine (General) > R858 Deep Learning T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5102.9 Signal processing. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7871.674 Detectors. Sensors T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7878 Electronic instruments
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Dedi Rachmad Hariyadi
Date Deposited:	20 Feb 2023 01:03
Last Modified:	20 Feb 2023 01:03
URI:	http://repository.its.ac.id/id/eprint/97580

Actions (login required)

View Item