In pediatric cardiology as well as many other medical specialties, the accurate portrayal of a large volume of patient information is crucial to providing good patient care. Our research aims at utilizing clinical and spatial ontologies representing the human heart, to automatically generate a Mullins-like diagram based on a patient's information in the cardiology databases. Our ontology allows an intuitive way of modeling congenital defects with the structure of the human heart. A prototype system has been implemented using Protege and the Java Advanced Imaging package , and is currently under testing at the Children's Mercy Hospitals and Clinics. A pilot study showed the diagrammatical representation and characterization of congenital heart defects can be used to facilitate a better understanding of a complex congenital heart defects.