Identifikasi Anemia melalui Citra Konjungtiva Mata Menggunakan Metode K-Means Clustering dan Convolutional Neural Network

Abstract

Anemia is a serious health problem that, if not promptly and accurately addressed, can lead to death. Anemia examinations can generally be performed invasively and non-invasively. Invasive examinations require time and expensive costs. The technical process of blood sampling can also cause discomfort, leading some people to hesitate in undergoing examinations. Meanwhile, non-invasive examinations only require observation of the paleness of the color on the conjunctival eye, which is considered far more effective than invasive examinations. However, human observation can be subjective, necessitating a system for identifying anemia using the K-Means Clustering and Convolutional Neural Network (CNN) methods. The data used consisted 400 images for the training process and 40 images for the testing process. The research results show the system’s ability to recognize conjunctival eye images of individuals with anemia or non-anemia with an accuracy rate of 95%.