Fahria, Tasya Aulia (2024) Klasifikasi Retinopati Diabetik Pada Citra Retina Menggunakan Local Binary Pattern (LBP) dan Support Vector Machine (SVM). Other thesis, Institut Teknologi Sepuluh Nopember.
Text
07311940000065-Undergaduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 July 2026. Download (2MB) | Request a copy |
Abstract
Diabetes Melitus (DM) atau penyakit gula didefinisikan sebagai suatu penyakit atau gangguan metabolisme kronis yang ditandai dengan tingginya kadar gula darah. Salah satu komplikasi diabetes melitus adalah retinopati diabetik, dimana kadar gula yang tinggi pada akhirnya menyebabkan kerusakan pada pembuluh darah di retina mata, terutama pada jaringan yang sensitif terhadap cahaya. Retinopati diabetik seringkali hanya menunjukkan gejala ringan atau tanpa gejala pada awalnya. Retinopati diabetik memiliki dua tahap. Non-Proliferative Diabetic Retinopathy (NPDR) merupakan gejala awal adanya retinopati diabetik yang terdiri dari kondisi ringan, sedang, dan berat. Selanjutnya, Proliverative Diabetic Retinopathy (PDR) ialah stadium lanjut dari NPDR yang mengarah ke kebutaan. Penelitian ini memanfaatkan teknik augmentasi data, reshape lalu resize menjadi 3 varian piksel, yaitu 64x64, 256x256, dan 448x448, dilanjut dengan penggunaan Contrast Limited Adaptive Histogram Equalization (CLAHE) dan Median Filter, serta ekstraksi fitur dari LBP dan Multiclass SVM berbasis one-vs-one untuk mengklasifikasikan retinopati diabetik menjadi 5 kelas, yaitu normal, mild, moderate, severe, dan PDR. Studi ini juga menguji kernel dan hyperparameter. Hasil terbaik dari model ini menunjukkan nilai akurasi sebesar 81%, micro average 81%, macro average 90%, dan AUC 0.81 dengan ukuran piksel 448x448. Harapan dari Tugas Akhir ini adalah dapat memberikan manfaat pada dokter atau tenaga medis dalam mengidentifikasi atau mendiagnosa dengan tingkat akurasi tinggi adanya penyakit retinopati diabetik. Oleh karena itu, prosedur perawatan pasien dapat dilakukan lebih awal sehingga dapat membantu mengurangi kerusakan permanen dan mengantisipasi penyakit retinopati diabetik lebih lanjut berdasarkan hasil klasifikasi citra retina pasien.
====================================================================================================================================
Diabetes Mellitus (DM) or diabetes is defined as a disease or chronic metabolic disorder characterized by high blood sugar levels. One of the complications of diabetes mellitus is diabetic retinopathy, where diabetes levels ultimately cause damage to the blood vessels in the retina of the eye, especially in light-sensitive tissues. Diabetic retinopathy often shows only mild symptoms or no symptoms at first. Diabetic retinopathy has two stages. Non-Proliferative Diabetic Retinopathy (NPDR) is an early symptom of diabetic retinopathy which consists of mild, moderate and severe conditions. Furthermore, Proliverative Diabetic Retinopathy (PDR) is an advanced stage of NPDR which leads to blindness. This research utilizes data augmentation techniques, reshape then resize into 3 pixel variants, namely 64x64, 256x256, and 448x448, followed by the use of Contrast Limited Adaptive Histogram Equalization (CLAHE) and Median Filter, as well as feature extraction from LBP and Multiclass SVM based on one-vs-one to classify diabetic retinopathy into 5 classes, namely normal, mild, moderate, severe, and PDR. This study also tested the kernel and hyperparameters. The best results from this model show an accuracy value of 81%, micro average 81%, macro average 90%, and AUC 0.81 with a pixel size of 448x448. The aim of this final assignment is that it can provide benefits to doctors or medical personnel in identifying or diagnosing with a high level of accuracy the presence of diabetic retinopathy. Therefore, patient treatment procedures can be carried out earlier so that they can help reduce permanent damage and anticipate further diabetic retinopathy disease based on the results of the patient's retinal image classification.
Item Type: | Thesis (Other) |
---|---|
Uncontrolled Keywords: | Diabetes Melitus, Retinopati Diabetik, Klasifikasi; Classification, Diabetic Retinopathy, Diabetes Mellitus |
Subjects: | Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. |
Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Biomedical Engineering > 11410-(S1) Undergraduate Thesis |
Depositing User: | Tasya Aulia Fahria |
Date Deposited: | 15 Feb 2024 02:10 |
Last Modified: | 15 Feb 2024 02:10 |
URI: | http://repository.its.ac.id/id/eprint/107027 |
Actions (login required)
View Item |