Multicriteria Decision Analysis Methodology for Medical Diagnosis Aid

Download
  1. (PDF, 219 KB)
AuthorSearch for:
TypeArticle
ConferenceDixièmes Journées francophones d'information médicale, JFIM 2003, Collection informatique et santé (Springer Verlag), September 4-5, 2003., Tunis, Tunisia
Subjectcomputer-aided diagnosis; classification; multicriteria decision analysis; acute leukemia
AbstractThe aim of this paper is to present the original methodology of acute leukemia diagnosis using a new classification procedure, called PROCTN. This procedure belongs to multicriteria decision analysis area and it is based on the scoring function to determine a subset of prototypes, which represent the closest resemblance with an object to be assigned. Then it applies the majority-voting rule to assign an object to a class. The implementation of PROCTN was carried out on cytological data of 108 cases of acute leukemia, using the classification rules of French, American and British hematologists, and was then applied on an independent test set of 83 cases of acute leukemia. Each case was defined by forty-seven parameters obtained by examining patient's bone marrow smears with optical microscope. In order to determine the percentage of correct classification of each subtype of acute leukemia, we have compared the results obtained by the procedure with the results given previously by the hematologist. 90 % of the cases were correctly classified by the proposed procedure. These primary results are satisfactory and show the efficiency of the new procedure. Although still an investigative method, the preliminary results are very encouraging and demonstrate the potential performances of this procedure for solving medical classification problems.
Publication date
LanguageEnglish
AffiliationNRC Institute for Information Technology; National Research Council Canada
Peer reviewedNo
NRC number46503
NPARC number8914299
Export citationExport as RIS
Report a correctionReport a correction
Record identifier2fbcad3d-150f-4547-94e3-e00361e60e0e
Record created2009-04-22
Record modified2016-05-09
Bookmark and share
  • Share this page with Facebook (Opens in a new window)
  • Share this page with Twitter (Opens in a new window)
  • Share this page with Google+ (Opens in a new window)
  • Share this page with Delicious (Opens in a new window)