PERBANDINGAN KINERJA SEGMENTASI JANTUNG JANIN MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORK

FARHAN, ABDULLAH and Nurmaini, Siti (2021) PERBANDINGAN KINERJA SEGMENTASI JANTUNG JANIN MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORK. Undergraduate thesis, Sriwijaya University.

	Text RAMA_56201_09011181722081.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (7MB) | Request a copy
	Text RAMA_56201_09011181722081_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (16MB) | Request a copy
Preview	Text RAMA_56201_09011181722081_0002085908_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (610kB) | Preview
	Text RAMA_56201_09011181722081_0002085908_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (166kB) | Request a copy
	Text RAMA_56201_09011181722081_0002085908_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (124kB) | Request a copy
	Text RAMA_56201_09011181722081_0002085908_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_56201_09011181722081_0002085908_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (30kB) | Request a copy
	Text RAMA_56201_09011181722081_0002085908_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (100kB) | Request a copy
	Text RAMA_56201_09011181722081_0002085908_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (724kB) | Request a copy

Abstract

Congenital heart disease is one of the leading causes of death in the first year of birth. One example of the challenges that exist in medical images, especially the fetal heart, is the poor image quality. In fetal heart echocardiography, the problem that occurs when diagnosing congenital heart disease is that the ultrasound image obtained is susceptible to blurry parts that can damage the image and reduce image quality. Segmentation of the fetal heart using deep learning can help doctors to diagnose congenital heart disease more quickly. The method used in this research is Convolutional Neural Network (CNN) with FractalNet, Resnet and U-Net architectures. In this study, the scenario carried out is to segment 7 classes with the number of models for each class that is opened 12 for the learning rate parameters, and the best loss function. Of the 12 models tested in each class. segmentation of the fetal heart in classes la, lv, ra, rv, hole, aorta, and fetal heart got the results of the dice coefficient 94.23%, 97.44%, 97.83%, 97.37%, 92.17%, 94 0.04%, 90.85%.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Segmentation, Fetal heart, Convolutional Neural Network.
Subjects:	Q Science > Q Science (General) > Q334-342 Computer science. Artificial intelligence. Algorithms. Robotics. Automation.
Divisions:	09-Faculty of Computer Science > 56201-Computer Systems (S1)
Depositing User:	Abdullah Farhan
Date Deposited:	09 Jul 2021 03:54
Last Modified:	09 Jul 2021 03:54
URI:	http://repository.unsri.ac.id/id/eprint/49523

Actions (login required)

View Item