Euclidean Distance Digital Image Processing For Jaundice Detect
Etika Putri Rahayu1, Melyana Nurul Widyawati2, Suryono Suryono3
1,2Postgraduate Program in Applied Midwifery Science, 3Departement of Physics Faculty of Science and Mathematics, 1,2Poltekkes Kemenkes, 3Diponegoro University
Abstract - Jaundice is a serious health issue. Late treatment of jaundice cases in babies may result in neurodevelopmental disorder and irreversible brain damage. Diagnosis inaccuracy is usually caused by the fact that health professionals and health service providers often rely on visual observation instead of laboratory examination. Lack of expertise in detecting jaundice is a serious matter. This research proposes a method of web-based digital image processing as an alternative for early detection of jaundice based on babies- complexion. Images of babies- complexion and color calibration cards are taken to obtain images for online analysis. Determination of bilirubin levels is carried out using the method of Euclidean approximate distance of RGB values from babies- complexion and those of color calibration cards. Results show correlation of Euclidean distance to bilirubin level of babies of 0.93596 and web-based digital image processing accuracy of 90%. These mean that the information system developed here is capable of detecting jaundice cases. This research was performed observationally in high-risk prenatal ward involving 30 infants as samples.
Keywords: Keyword - Euclidean distance, Image processing, Jaundice, Web based
Topic: Electrical Engineering and Computer Science