Mining Rare Cases in Post-Operative Pain by Means of Outlier Detection

• Rare cases are often interesting for healthprofessionals, physicians, researchers and clinicians in order toreuse and disseminate experiences in healthcare. However,mining, i.e. identification of rare cases in electronic patientrecords, is non-trivial for information technology. This paperinvestigates a number of well-known clustering algorithms andfinally applies a 2nd order clustering approach by combining theFuzzy C-means algorithm with the Hierarchical one. Theapproach is used in order to identify rare cases from 1572patient cases in the domain of post-operative pain management.The results show that the approach enables identification of rarecases in the domain of post-operative pain management and 18%of cases are identified as rare case.

Subject headings

NATURVETENSKAP  -- Data- och informationsvetenskap (hsv//swe)

NATURAL SCIENCES  -- Computer and Information Sciences (hsv//eng)

Keyword

rare cases

clustering

case mining

medical

Computer Science

datavetenskap

Publication and Content Type

vet (subject category)

ovr (subject category)