Amrullah, Muhammad Ahnaf (2022) Implementasi Jaringan Saraf Konvolusional dengan Inception-V3 untuk Deteksi Katarak Menggunakan Gambar Digital Funduskopi. Other thesis, Institut Teknologi Sepuluh Nopember.
|
Text
06111840000051-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (4MB) | Request a copy |
Abstract
Katarak merupakan salah satu penyakit mata yang paling serius yang dapat menyebabkan kebutaan. Deteksi dan pengobatan dini dapat mengurangi kebutaan pada pasien katarak. Seiring berkembangnya teknologi pelayanan kesehatan saat ini mengintegrasikan alat kesehatan dan teknologi informasi untuk meningkatkan kualitas dan produktivitas dalam pelayanan kesehatan. Hasil gambar funduskopi atau gambar bagian belakang dan dalam mata (fundus) dapat digunakan untuk memprediksi katarak. Dalam Tugas Akhir ini diimplementasikan Convolutional Neural Network (CNN) dengan arsitektur Inception-V3 dalam deteksi katarak berdasarkan gambar digital funduskopi. Terdapat 3 jenis citra fundus yang digunakan yaitu citra fundus normal, citra fundus katarak, dan citra fundus degenerasi makula. Data gambar fundus dipraproses menggunakan histogram equalization dan Contrast Limited Adaptive Histogram Equalization (CLAHE) terhadap channel hijau. Hasil terbaik pada Tugas Akhir ini adalah model dengan praproses CLAHE dengan Fine Tuning yang memiliki akurasi sebesar 98,33%.
==============================================================================================================================
Cataract is one of the most serious eye diseases that can lead to blindness. Early detection and treatment can reduce blindness in cataract patients. As technology develops, current healthcare services integrate medical devices and information technology to improve quality and productivity in healthcare services. Funduscopy image results or images of the back and inside of the eye (fundus) can be used to predict cataracts. In this Final Project, a Convolutional Neural Network (CNN) with the Inception-V3 architecture is implemented in cataract detection based on digital funduscopy images. There are 3 types of fundus images used, namely normal fundus images, cataract fundus images, and macular degeneration fundus images. Fundus image data is preprocessed using histogram equalization and Contrast Limited Adaptive Histogram Equalization (CLAHE) on the green channel. The best result in this Final Project is the model with CLAHE preprocessing with Fine Tuning which has an accuracy of 98.33%.
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSMa 006.42 Amr i-1 2022 |
| Uncontrolled Keywords: | Cataract, Convolutional Neural Network (CNN), Inception-V3. Convolutional Neural Network (CNN), Inception-V3, Katarak. |
| Subjects: | Q Science > QA Mathematics |
| Divisions: | Faculty of Mathematics and Science > Mathematics > 44201-(S1) Undergraduate Thesis |
| Depositing User: | Mr. Marsudiyana - |
| Date Deposited: | 05 Jun 2026 04:56 |
| Last Modified: | 05 Jun 2026 04:56 |
| URI: | http://repository.its.ac.id/id/eprint/133607 |
Actions (login required)
![]() |
View Item |
