REKONSTRUKSI CITRA ULTRASONOGRAFI TRANSCEREBELLAR 2 DIMENSI MENJADI TRANSCEREBELLAR 3 DIMENSI MENGGUNAKAN METODE 3D-FHNet

RAMADHAN, MUHAMMAD RIZKY and Erwin, Erwin (2023) REKONSTRUKSI CITRA ULTRASONOGRAFI TRANSCEREBELLAR 2 DIMENSI MENJADI TRANSCEREBELLAR 3 DIMENSI MENGGUNAKAN METODE 3D-FHNet. Undergraduate thesis, Sriwijaya University.

	Text RAMA_56201_09011281924063.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (5MB) | Request a copy
	Text RAMA_56201_09011281924063_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (8MB) | Request a copy
	Text RAMA_56201_09011281924063_0029017101_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (1MB)
	Text RAMA_56201_09011281924063_0029017101_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (156kB) | Request a copy
	Text RAMA_56201_09011281924063_0029017101_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (363kB) | Request a copy
	Text RAMA_56201_09011281924063_0029017101_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (723kB) | Request a copy
	Text RAMA_56201_09011281924063_0029017101_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (852kB) | Request a copy
	Text RAMA_56201_09011281924063_0029017101_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (27kB) | Request a copy
	Text RAMA_56201_09011281924063_0029017101_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (3MB) | Request a copy

Abstract

Reconstruction of two-dimensional transcerebellar images into three�dimensional transcerebellar has an important role in diagnosis and treatment planning in neurology. This study proposes the 3D-FHNet method to produce an accurate three-dimensional representation of transcerebellar images. The study began with segmentation of transcerebellar objects on well-labeled ultrasound datasets. The segmentation architecture used is U-Net and LinkNet as a comparison. After training, both architectures can process object segmentation properly. The best performance was produced by U-Net with a pixel accuracy of 99.83%, Mean IU of 89.71%, FPR of 0.91%, Precision of 85.78%, Recall of 85.31%, and F1 Score of 85.31%. With this accuracy, cross-validation was carried out using a 10-Fold K-Fold. The U-Net architecture was chosen to continue the 3D reconstruction process. The reconstruction process was carried out using 2 models as an improvement on the previous results using the PiFUHD method. The PiFUHD method succeeded in creating a 3-dimensional representation with one input from the previous segmentation results. From the results of the 3-dimensional representation, 4 sides are taken and then implemented into the 3D-FHNet method. The resulting object is the same as a 3-dimensional object. Model performance can be calculated by using the input as ground truth or the IoU (Intersection over Union) method with an average of 76.76% for 19 test data. By using the 3D-FHNet method, three-dimensional image reconstruction of the ultrasound image of the fetal head can be performed accurately.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Segmentasi, Rekonstruksi, Transcerebellar
Subjects:	T Technology > T Technology (General) > T57.6-57.97 Operations research. Systems analysis > T57.6.A2-Z General works Simulation Cf. QA76.9.C65 Computer science Cf. TA343 Engineering mathematics
Divisions:	09-Faculty of Computer Science > 56201-Computer Systems (S1)
Depositing User:	Muhammad Rizky Ramadhan
Date Deposited:	01 Aug 2023 08:04
Last Modified:	01 Aug 2023 08:04
URI:	http://repository.unsri.ac.id/id/eprint/124965

Actions (login required)

View Item