This Phase I project proposes to assess the feasibility of an advanced diagnostic logic system (Diagnostica) to support the clinical assessment process for diagnosis in psychiatry. A working prototype of the diagnostic rules in the American Psychiatric Association Diagnostic and Statistical Manual (DSM-IV) uses and artificial intelligence engine (""""""""XSB"""""""") to implement the logic of DSM-IV along with and an interactive graphical user interface to allow a user to add information and understand conclusions reached by the system. The Phase I programming objectives are to make Diagnostica ready for commercial use by improving its graphical user interface, and finalizing implementation of its logical rules. The resulting system will be a practical tool in clinical settings, and relies on computer science innovations that have preciously neither been explored nor applied in the domain of medical reasoning. With the emergence of decision support systems, the need for better quality diagnostic information is becoming increasingly apparent. This has been due, in part, to the complexity of diagnostic processes and the emphasis on support of financial processes. Within mental health, the DSM-IV provides both a model and a standard for making diagnoses. A software component that provides flexible, complete, and efficient application of this standard is of great value. The innovation of Diagnostica relies on the sophistication of its modeling of DSM-IV rules, and it's flexibility in applying those rules. Diagnostica will automatically track the status of the information entered and allow users to tie up 'loose ends' in documenting the proof of diagnoses formally. AS example, the user may indicate that a set of diagnoses in 'believed true' without specifying the symptoms needed to make the diagnoses formally ( a procedure used routinely in clinical practice). the application will track whatever 'residual' data this is necessary in order to complete formal diagnoses, while leaving the option of when, or if, to complete the process up to the user. Phase II objectives include: (1) extending Diagnostica to provide other software applications needing diagnostic decision support services, and specifically to link Diagnostica to the World Health Organization Schedules for Clinical Assessment in Neuropsychiatry (SCAN); (2) addressing logical modeling of time and creating an effective user interface for repeated assessment; (3) incorporating probabilistic information about sets of symptoms based on empirical information initially obtained in Phase I; and (4) developing and testing """"""""belief revision"""""""" functions to changes in knowledge stemming from repeated clinical assessment.
Computerization of diagnostic logic for clinical use can improve the quality of mental health services by efficient standardization of assessment and through motivating and making more practical the creation of data bases which can be used for clinical quality improvement and knowledge discovery.
