| Title | Experience with Baye's Theorem for Computer Analysis of Congenital Heart Disease |
| Author(s) | Homer R. Warner; Alan F. Toronto; and George Veasy |
| Source | Annals of the NY Academy of Sciences, Vol. 115, Pages 558-567 |
| Publication Date | 1964 |
| Abstract | An equation based on Baye's Theorem has been described which has been programmed for a general-purpose digital computer for medical diagnosis. A consideration of the assumptions inherent in this approach was presented along with the methods for handling exceptions, such as mutually exclusive symptoms. Experience with this approach in the field of congenital heart disease indicates that these diseases can be diagnosed with an accuracy equal to that of an experienced specialist in this field. Furthermore, the accuracy of the computer diagnosis is still improving with refinements in the data matrix. It is suggested that perhaps a part of the improvement in the diagnostic accuracy of the physicians in this study may be attributed to their experience in preparing data for the computer and receiving feedback from the computer in the form of a differential diagnosis. |