SEGMENTASI SEL TUMOR OTAK DENGAN KOMBINASI ARSITEKTUR 3D U-NET DAN ATTENTION GATE PADA CITRA HASIL MAGNETIC RESONANCE IMAGING (MRI) OTAK

LUBIS, ANDIKA CRISTIAN and Desiani, Anita and Irmeilyana, Irmeilyana (2024) SEGMENTASI SEL TUMOR OTAK DENGAN KOMBINASI ARSITEKTUR 3D U-NET DAN ATTENTION GATE PADA CITRA HASIL MAGNETIC RESONANCE IMAGING (MRI) OTAK. Undergraduate thesis, Sriwijaya University.

	Text RAMA_44201_08011282025025.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (2MB) | Request a copy
	Text RAMA_44201_08011282025025_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (6MB) | Request a copy
	Text RAMA_44201_08011282025025_0011127702_0017057403_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (1MB)
	Text RAMA_44201_08011282025025_0011127702_0017057403_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_44201_08011282025025_0011127702_0017057403_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (230kB) | Request a copy
	Text RAMA_44201_08011282025025_0011127702_0017057403_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_44201_08011282025025_0011127702_0017057403_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (127kB) | Request a copy
	Text RAMA_44201_08011282025025_0011127702_0017057403_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (212kB) | Request a copy

Abstract

Based on the Global Cancer Observatory in 2020, brain tumors occupy the 19^th position among all malignant diseases that most often cause death. To help analyze brain tumor cells, it is necessary to detect the brain using Magnetic Resonance Imaging (MRI). In detecting MRI images, segmentation is applied by applying the Convolutional Neural Network (CNN) method. CNN architecture that is often used for 3-dimensional image segmentation is the 3D U-Net architecture. The purpose of this research is to overcome the shortcomings of the 3D U-Net architecture by combining the 3D U-Net architecture and the attention gate mechanism . Attention gate is placed in the decoder section after the skip connections process so that the features under the skip connections can focus only on important features. With only important features being filtered out, the model can more accurately learn feature patterns. The performance level of the model was measured using accuracy, sensitivity, specificity, Intersection over Union (IoU), and F1-Score, which achieved values of 99.77%, 97.87%, 98.73%, 89.82%, and 94.83%, respectively. The evaluation measures are categorized as very good in segmenting except IoU which is categorized as good. Based on the results of the model performance evaluation measures obtained, the combination of 3D U-Net architecture and attention gate provides excellent results in segmenting brain tumor cells from MRI results.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	deep learning, CNN, Komputasi Matematika
Subjects:	Q Science > QA Mathematics > QA75-76.95 Calculating machines > QA76.9.B45 Big data. Machine learning. Quantitative research. Metaheuristics.
Divisions:	08-Faculty of Mathematics and Natural Science > 44201-Mathematics (S1)
Depositing User:	Andika Cristian Lubis
Date Deposited:	25 Nov 2024 06:03
Last Modified:	25 Nov 2024 06:03
URI:	http://repository.unsri.ac.id/id/eprint/159777

Actions (login required)

View Item