Implementasi Perangkat Lunak Deteksi Penyakit Retinopati Hipertensi Di Polimata Rumah sakit Umum Provinsi Nusa Tenggara Barat

Abstract

Hypertensive retinopathy is a type of eye disease where microvascular changes occur in the retina experienced by high blood pressure sufferers. The arterial and venous ratio (AVR) in the retina of the eye is an indicator used to determine the presence of high blood pressure, which is measured by the ratio of the width of the retinal arteries and veins.  Traditionally, ophthalmologists use fundus images or retinal images of the eye to diagnose hypertensive retinopathy's physical symptoms and determine the phase of evolution. Still, traditional methods have limitations because, in the case of borderline stages, the early symptoms of hypertensive retinopathy will be difficult to identify manually, so they are often ignored. Referring to these problems, early diagnosis is needed for accurate prevention and treatment of hypertensive retinopathy. Based on the abovementioned issues, this service activity aims to implement a hypertensive retinopathy disease detection model using a local dataset from a regional general hospital in West Nusa Tenggara (NTB). It will compare the model detection results with those of three eye disease experts. Classification model testing results using the Messidor training and NTB Regional Hospital datasets. In models using the Messidor training dataset, the highest accuracy is a comparison with the results of the most senior expert's observations. The results of the classification model are only a tool to assist ophthalmologists in diagnosing hypertensive retinopathy, while the final decision remains with the expert or ophthalmologist.