Jaundice (Hyperbilirubinemia) detection and prediction system using color card technique

Jaundice is describes as yellow discoloration of the skin and other tissues of a newborn infant and it is conjunctiva due to the bilirubin rises up to 5 mg/dL or 85 μmol/L. In Malaysia itself, about 75% of newborns had jaundiced in the first week of life, this proves by the survey made by Ministry of Health Malaysia. Traditional assessment (Kramer's Rule) is not convenient since there are limits to the accuracy of visual estimation, thus blood test taken from baby's heel remains to be necessary but lead to traumatize to baby and their parents. Moreover the available jaundice meters in the market are expensive. This research work proposed the jaundice (hyperbilirubinemia) detection and prediction system using color card technique. It detects jaundice non-invasively by optical method (spectral reflectance method). This research doesn't involve any subject or infant. It collects data by capturing different type of color shades that represent random bilirubin color in patient's body, human skin color and bilirubin concentration color card using silicon photodiode sensor. This research concern about the human skin factor as it is the major disadvantages in most of the jaundice assessment. Thus, human skin colors represented by Fitzpatrick shades are considered. The final outputs were analyzed using linear regression analysis and characterized based on particular bilirubin value and treatment needed. The final outputs are shown in graphical user interface (Microsoft Visual Basic 2010) on medical staff's PC and can only be viewed by authorized person.