Segmentasi dan Pemodelan Saluran Udara Paru Menggunakan Arsitektur VNet. (Segmentation and Modeling of Pulmonary Airway Using VNet Architecture).

Wicaksana, Andika Setianata (2024) Segmentasi dan Pemodelan Saluran Udara Paru Menggunakan Arsitektur VNet. (Segmentation and Modeling of Pulmonary Airway Using VNet Architecture). Undergraduate thesis, Universitas 17 Agustus 1945 Surabaya.

	Text ABSTRAK.pdf Download (1MB)
	Text BAB I.pdf Restricted to Repository staff only Download (26kB) | Request a copy
	Text BAB II.pdf Restricted to Repository staff only Download (660kB) | Request a copy
	Text BAB III.pdf Restricted to Repository staff only Download (223kB) | Request a copy
	Text BAB IV.pdf Restricted to Repository staff only Download (811kB) | Request a copy
	Text BAB V.pdf Restricted to Repository staff only Download (136kB) | Request a copy
	Text Pengalihan Publikasi.pdf Download (262kB)
	Text jurnal_turnitin.pdf Restricted to Repository staff only Download (1MB) | Request a copy
	Other (Thumbnails conversion from text to thumbnail_lightbox) lightbox.jpg Download (0B)
	Other (Thumbnails conversion from text to thumbnail_preview) preview.jpg Download (0B)
	Other (Thumbnails conversion from text to thumbnail_medium) medium.jpg Download (0B)
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Restricted to Repository staff only Download (0B) | Request a copy
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Restricted to Repository staff only Download (0B) | Request a copy
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Restricted to Repository staff only Download (0B) | Request a copy
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Restricted to Repository staff only Download (0B) | Request a copy
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Download (0B)
	Other (Thumbnails conversion from text to thumbnail_medium) medium.jpg Download (0B)
	Other (Thumbnails conversion from text to thumbnail_small) small.jpg Download (0B)
	Other (Thumbnails conversion from text to thumbnail_lightbox) lightbox.jpg Restricted to Repository staff only Download (0B) | Request a copy
	Other (Thumbnails conversion from text to thumbnail_small) small.jpg Restricted to Repository staff only Download (0B) | Request a copy
	Other (Thumbnails conversion from text to thumbnail_small) small.jpg Restricted to Repository staff only Download (0B) | Request a copy
	Other (Thumbnails conversion from text to thumbnail_preview) preview.jpg Restricted to Repository staff only Download (0B) | Request a copy
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Download (0B)
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Restricted to Repository staff only Download (0B) | Request a copy
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Restricted to Repository staff only Download (0B) | Request a copy
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Restricted to Repository staff only Download (0B) | Request a copy
	Other (Thumbnails conversion from text to thumbnail_lightbox) lightbox.jpg Restricted to Repository staff only Download (0B) | Request a copy
	Other (Thumbnails conversion from text to thumbnail_preview) preview.jpg Restricted to Repository staff only Download (0B) | Request a copy
	Other (Thumbnails conversion from text to thumbnail_lightbox) lightbox.jpg Restricted to Repository staff only Download (0B) | Request a copy
	Other (Thumbnails conversion from text to thumbnail_lightbox) lightbox.jpg Download (0B)
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Download (0B)
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Restricted to Repository staff only Download (0B) | Request a copy
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Restricted to Repository staff only Download (0B) | Request a copy
	Other (Generate index codes conversion from text to indexcodes) indexcodes.txt Download (0B)

Abstract

The airway of the lungs is a complex structure with various branching and fine levels, making the diagnosis of lung airway diseases require high precision and time�consuming effort. Therefore, a technique or method is needed to expedite the diagnosis process. In recent years, advancements in image processing technology and artificial neural networks have significantly progressed in the field of lung airway segmentation and modeling. By utilizing segmentation and modeling of lung airways, medical professionals can easily diagnose diseases affecting these structures. The aim of this research is to generate lung airway shapes using Deep Learning image segmentation methods and visualize them in 3D. Another objective is to address existing segmentation challenges.This study focuses on using the VNet architecture method for segmenting parts of the lung airway. The research process begins by inputting 2D chest CT scan images, followed by cleaning and smoothing the object structures from the chest CT scan images, and then proceeding with the segmentation process. The resulting segmented images are further processed to refine the segmented images, which will then undergo 3D modeling.The findings of this study indicate that the segmentation model trained without preprocessing enhancement yielded better evaluation results compared to the model trained using preprocessing enhancement. This is evidenced by the evaluation scores without preprocessing enhancement achieving accuracy of 99.5%, precision of 96.7%, and DSC (Dice Similarity Coefficient) of 95.8%. In contrast, the evaluation scores with preprocessing enhancement achieved accuracy of 97.1%, precision of 60.4%, and DSC of 65.2%. However, the lung airway object cannot be fully visible due to the use of 96 x 96 pixel size for patches and input images, resulting in segmentations that only display the central part of the lung airway with small portions of bronchi on the right and left sides. Existing postprocessing methods can rectify the breakages in segmentation, demonstrating the ability to address unconnected objects using postprocessing fill holes.

Actions (login required)

View Item