Version 2.1.0.0 CRISP Logo CRISP Homepage Help for CRISP Email Us

Abstract

Grant Number: 1R43MH066434-01
PI Name: OTT, GEOFFREY E.
PI Email: allen@mdlogix.com
PI Title:
Project Title: Diagnostic Logic and Adaptive Assessment for Psychiatry

Abstract: DESCRIPTION (provided by applicant): This SBIR Phase I application proposes to extend an existing multi-tier Internet client-server system ("CliniMetricar") by integrating active diagnostic logic. CliniMetrica currently provides systematic psychiatric assessment using the World Health Organization (WHO) Schedules for Clinical Assessment in Neuropsychiatry (SCAN, version 2.1). CliniMetrica automatically indexes digital audio recording of assessment interviews, and was developed with SBIR support from NIMH Digital video functions are currently being implemented and tested with additional NIMH support, as are functions to support remote psychometrics for the training and monitoring of interviewers. The CliniMetrica system is increasing in sophistication and functionality, and has applications for clinical research (e.g., clinical trials) as well as for more routine clinical practice. The current lack of fully developed and tested diagnostic functionality is a major gap. A number of potential customers have requested integrated DSM-IV and/or lCD-10 diagnostic results to be automatically linked to assessments. In order to meet this market need, we propose to implement the DSM-IV and ICD-10 nosologic systems as two classification knowledge bases that a logic engine will process to generate diagnostic results. These logical functions are difficult to implement and manipulate with procedural languages (e. g. C + +). so the use of a logic engine provides significant technical benefits. This diagnostic logic version of CliniMetrica is referred to as "CliniMetrica-Dx." To assist raters in thorough examination and to support adaptive assessment, a user interface coupled to logic processing modules will allow tracking the diagnostic status of a subject (the sets of true, false, partially true, or partially false DSM-IV and lCD-10 diagnoses during assessment. The logic engine will dynamically generate assessment item subsets (currently from the SCAN) needed to rule-in or rule-out DSM-IV and lCD-10 diagnoses. During the assessment, presentation of these symptom sets to the assessor can further guide the interview. By narrowing the "search space," the efficiency of assessments will be increased. By formalizing the search, reliability and validity can be enhanced In addition, to provide support for nosologic research and development, in Phase II the system will include mappings between DSM-IV and lCD-10, and between current and future versions of the DSM and lCD. The CliniMetrica-Dx system as an application framework will generalize to other clinical psychiatric assessment instruments such as the SCID, to adaptive self-report systems, and to other medical specialties. It also will have applications for education and training.

Thesaurus Terms:
artificial intelligence, computer assisted diagnosis, diagnosis design /evaluation, mental disorder diagnosis, psychometrics
Internet, computer graphics /printing, computer program /software, computer system design /evaluation, disease /disorder classification, indexing
clinical research, human subject

Institution: MEDICAL DECISION LOGIC, INC.
724 DULANEY VALLEY RD, STE 2
TOWSON, MD 21204
Fiscal Year: 2002
Department:
Project Start: 01-AUG-2002
Project End: 31-JAN-2003
ICD: NATIONAL INSTITUTE OF MENTAL HEALTH
IRG: ZRG1


CRISP Homepage Help for CRISP Email Us