IMPLEMENTASI ALGORITMA CONVOLUTIONAL NEURAL NETWORK DENGAN ARSITEKTUR RESNET-152 UNTUK KLASIFIKASI PENYAKIT RETINOPATHY OF PREMATURITY BERDASARKAN CITRA FUNDUS

Hana Mumtaz, . (2025) IMPLEMENTASI ALGORITMA CONVOLUTIONAL NEURAL NETWORK DENGAN ARSITEKTUR RESNET-152 UNTUK KLASIFIKASI PENYAKIT RETINOPATHY OF PREMATURITY BERDASARKAN CITRA FUNDUS. Skripsi thesis, Universitas Pembangunan Nasional Veteran Jakarta.

[img] Text
ABSTRAK.pdf
Download (47kB)
[img] Text
AWAL.pdf
Download (578kB)
[img] Text
BAB 1.pdf
Restricted to Repository UPNVJ Only
Download (132kB)
[img] Text
BAB 2.pdf
Restricted to Repository UPNVJ Only
Download (695kB)
[img] Text
BAB 3.pdf
Restricted to Repository UPNVJ Only
Download (717kB)
[img] Text
BAB 4.pdf
Restricted to Repository UPNVJ Only
Download (2MB)
[img] Text
BAB 5.pdf
Download (47kB)
[img] Text
DAFTAR PUSTAKA.pdf
Download (135kB)
[img] Text
RIWAYAT HIDUP.pdf
Restricted to Repository UPNVJ Only
Download (55kB)
[img] Text
LAMPIRAN.pdf
Restricted to Repository UPNVJ Only
Download (909kB)
[img] Text
HASIL PLAGIARISME.pdf
Restricted to Repository staff only
Download (21MB)
[img] Text
ARTIKEL KI.pdf
Restricted to Repository staff only
Download (588kB)

Abstract

Retinopathy of Prematurity (ROP) is one of the leading causes of permanent blindness in premature infants due to excessive oxygen exposure that damages the immature retina. In Indonesia, the prevalence of ROP remains high, while access to accurate and efficient early detection services remains limited, particularly in areas with limited healthcare resources. Therefore, an artificial intelligence-based technological solution is needed to support the detection and classification of ROP severity. This study developed an automatic classification system using a Convolutional Neural Network (CNN) algorithm with a ResNet-152 architecture. The dataset consists of 1591 fundus images obtained from Shenzhen Eye Hospital and Ostrava University Hospital, classified into four classes: Normal, ROP Stage 1, Stage 2, and Stage 3. This study tested 24 experimental schemes by varying the dataset splitting ratio, the amount of training data with and without augmentation, the use of CLAHE, and the use of dropout, both with and without fine-tuning, to understand the effect of each combination on classification performance. The model with the optimal configuration uses the CLAHE method, augmentation reaching 4000 training data, the addition of dropout of 0,3, and fine-tuning of the last 20 layers, achieving an accuracy of 87,20%. Additionally, integrating the model into a GUI interface facilitates healthcare professionals in efficiently classifying ROP. This study demonstrates the significant potential of deep learning technology in supporting early ROP detection, particularly in areas with limited access to specialized expertise.

Item Type: Thesis (Skripsi)
Additional Information: [No.Panggil: 2110511041] [Pembimbing 1: Widya Cholil] [Pembimbing 2: Neny Rosmawarni] [Penguji 1: Ridwan Raafi’udin] [Penguji 2: Nurul Afifah Arifuddin]
Uncontrolled Keywords: Retinopathy of Prematurity, Deep learning, Convolutional Neural Network, ResNet-152
Subjects: T Technology > T Technology (General)
Divisions: Fakultas Ilmu Komputer > Program Studi Informatika (S1)
Depositing User: HANA MUMTAZ
Date Deposited: 13 Jul 2025 21:06
Last Modified: 13 Jul 2025 21:06
URI: http://repository.upnvj.ac.id/id/eprint/37112

Actions (login required)

View Item View Item