Abstract—This paper introduces a top-down algorithm for
diagnosing psychiatric illnesses. It is based on the
conceptualisation of diagnostic categories, diagnosis, and
symptoms as a hierarchical model. The algorithm assumes that
there exist a few close-ended clinical questions that can be used
during clinical interview to rule in and rule out diagnostic
categories, diagnoses and their symptoms. Compared to a more
exhaustive bottom-up and recursive algorithm, which the
authors have previously introduced, this algorithm has the
advantage of being easy to implement requiring a less extensive
knowledgebase. It is expected the algorithm will be used as a
useful screening tool that increases the detection of psychiatric
disorders, which are common but unfortunately currently
under-diagnosed.
Index Terms—Diagnosis, diagnostic algorithm, psychiatry,
screening tool.
The authors are with the School of Electrical Engineering & Computer
Science, University of Newcastle, NSW 2308, Australia (phone: +61 423
281 664; e-mail: irosh.fernando@uon.edu.au,
frans.henskens@newcastle.edu.au).
Cite:Irosh Fernando and Frans A. Henskens, "Drill-Locate-Drill Algorithm for Diagnostic Reasoning in Psychiatry," International Journal of Machine Learning and Computing vol.3, no. 5, pp. 449-452, 2013.