Harun, Jacinthya Stephany Ezralia and Iskandar, Aulia Arif and Setiawan, Lyana and Susanto, Hery (2022) ENHANCING PHILADEPHIA MUTATION DETECTION USING OPTIMIZED IMAGE PRE-PROCESSING FOR CONVOLUTION NEURAL NETWORK. Bachelor thesis, Swiss German University.

	Text BE_Jacinthya Harun_1.pdf Restricted to Repository staff only Download (517kB)
	Text BE_Jacinthya Harun_2.pdf Restricted to Repository staff only Download (2MB)
	Text BE_Jacinthya Harun_3.pdf Restricted to Repository staff only Download (1MB)
	Text BE_Jacinthya Harun_4.pdf Restricted to Repository staff only Download (2MB)
	Text BE_Jacinthya Harun_5.pdf Restricted to Repository staff only Download (243kB)
	Text BE_Jacinthya Harun_REF.pdf Download (2MB)
	Text BE_Jacinthya Harun_TOC.pdf Download (1MB)

Abstract

Leukemia is a cancer that have a high mortality rate. Identifying the marker of Leukemia could help determine the type of Leukemia and it's treatment. Philadelphia Mutation (Ph)t(9;22)(q34;q11.2) is one of the markers that commonly occurred in acute lymphocytic Leukemia-ALL (20-30%) and chronic myelocytic Leukemia-CML (90-95%) patients. This marker can be detected by several diagnosis tests, including Karyotyping. However, Karyotyping is known to have high errors due to human factors and poor chromosomal shape and band. This research aims to aid medical professionals in detecting Ph mutation by developing an image processing algorithm using Gaussian Blur, Otsu Thresholding, CLAHE, Median Blur, and Convolution Neural Network. 400 Chromosome images were collected from literature studies, Dharmais Cancer Hospital, RSUP Dr. Cipto Mangunkusumo (RSCM), and Mitra Keluarga Laboratory as the dataset. The first step was the preprocessing method followed by chromosomal classification by using fine-tuned Inception-ResNet V2 CNN. The step achieved the average accuracy, precision, recall, and F1-score of 92.88% 84.31%, 82.55% and 82.77% respectively. The final step was detecting Ph mutation and this step obtained 84.6% accuracy. It is concluded that the image processing algorithm potentially detects Ph mutation and aids the detection of Ph mutation in Karyotyping.

Item Type:	Thesis (Bachelor)
Subjects:	Q Science > QA Mathematics > QA76 Computer software > QA76.87 Neural networks (Computer science)
Divisions:	Faculty of Life Sciences and Technology > Department of Biomedical Engineering
Depositing User:	Mr Arinton Sinaga
Date Deposited:	10 Dec 2024 04:18
Last Modified:	10 Dec 2024 04:18
URI:	http://repository.sgu.ac.id/id/eprint/2670

Actions (login required)

View Item