Production of diagnostic rules from a neurotologic database with decision trees

A decision tree is an artificial intelligence program that is adaptive and is closely related to a neural network, but ran handle missing or nondecisi ve data in decision-making. Data on patients with Meniere's disease, vestib ular schwannoma, traumatic vertigo, sudden deafness, benign paroxysmal posi tional vertigo, and vestibular neuritis were retrieved from the database of the otoneurologic expert system ONE for the development and testing of the accuracy of decision trees in the diagnostic workup. Decision trees were c onstructed separately for each disease. The accuracies of the best decision trees were 94%, 95%, 90%, 99%, 100%, and 100% for the respective diseases. The most important questions concerned the presence of vertigo, hearing lo ss, and tinnitus; duration of vertigo; frequency of vertigo attacks severit y of rotational vertigo; onset and type of hearing loss; and occurrence of head injury in relation to the timing of onset of vertigo. Meniere's diseas e was the most difficult to classify correctly. The validity and structure of the decision trees are easily comprehended and can be used outside the e xpert system.