Childhood Stuttering: A Second Opinion

Childhood Stuttering: A Second Opinion is a decision support system designed to help speech clinicians diagnose and select an appropriate management plan for preschool children at risk for a stuttering disability. Speech-language clinicians can vary widely in their diagnoses of stuttering in young children, but discussing such cases with experienced clinicians reduces this variability. When using Second Opinion®, clinicians with different training and experience arrived at diagnostic opinions that were indistinguishable from those of a panel of five experienced clinicians. In effect, using Second Opinion allowed inexperienced clinicians to "review" cases of incipient stuttering with a panel of experts, a process that should increase the reliability of their diagnoses in the real world. Second Opinion talks clinicians through the evaluation process asking specific questions. Then it analyzes their responses to these questions, assesses of the child's risk of stuttering, and presents clinical management alternatives for consideration. It saves this information in a file that can be printed for the client's folder.

The knowledge in Second Opinion reflects the consensus diagnostic opinions of a panel of expert clinicians:

The knowledge of these five experts for diagnosing incipient stuttering and selecting an appropriate clinical management approach was written into if-then production rules that are used in an Artificial Intelligence program. An important aspect of this project was obtaining a consensus from these experts about what is and is not important in diagnosing incipient stuttering. This knowledge was incorporated into a personal computer-based decision support system. We then systematically compared the output of the system with the diagnostic consensus of the panel as well as the diagnostic conclusions of each panel member with that of every other member on 30 cases of suspected incipient stuttering. Finally, we evaluated the system's ease of access and use by samples of intended users.

This research was supported by Bahill Intelligent Computer Systems and the National Institute of Child Health and Human Development, NIH, grant number 2 R44 HD26209-02A1, although the NIH offer no explicit or implied endorsement.

For further information see Bahill, A.T., Bharathan, K. and Curlee, R.F., "How the testing techniques for a decision support system changed over nine years," IEEE Transactions on Systems, Man and Cybernetics, 25, 1533-1542, 1995.

Would you like to go to Bahill's Systems Engineering Page?