Search

Background: Inborn metabolic disorders (IMDs) form a large group of rare, but often serious, metabolic disorders. Aims: Our objective was to construct a decision tree, based on classification algorithm for the data on three metabolic disorders, enabling us to take decisions on the screening and clinical diagnosis of a patient. Settings and Design: A non-incremental concept learning classification algorithm was applied to a set of patient data and the procedure followed to obtain a decision on a patient’s disorder. Materials and Methods: Initially a training set containing 13 cases was investigated for three inborn errors of metabolism. Results: A total of thirty test cases were investigated for the three inborn errors of metabolism. The program identified 10 cases with galactosemia, another 10 cases with fructosemia and the remaining 10 with propionic acidemia. The program successfully identified all the 30 cases. Conclusions: This kind of decision support systems can help the healthcare delivery personnel immensely for early screening of IMDs.