Health Evaluation and Promotion
Online ISSN : 1884-4103
Print ISSN : 1347-0086
ISSN-L : 1347-0086
Original Articles
Verification of Distinction Ability by Self Organizing Maps (SOM) about Metabolic Syndrome Data in Health Evaluation and Promotion
Heizo TOKUTAKANobuhiko KASEZAWA
Author information
JOURNAL FREE ACCESS

2010 Volume 37 Issue 3 Pages 389-397

Details
Abstract

  Regarding diagnostic criteria for metabolic syndrome (MS), minor cases where the lipid, blood pressure, and blood glucose levels slightly exceed their normal ranges are considered to require careful observation, because they are border line between “normal” and“abnormal” conditions.
  We devised a method to solve this problem by introducing the concept of an “undeveloped” condition.
  Based on observations of the MS-score distribution for each examinee, checkup data were numerically processed into three areas representing the “normal”, “undeveloped”, and “critical” conditions, and applied to Self-Organizing Maps (SOM) to develop a tool visualizing the MS indicators.
  The distinguishing capacity of our method was validated by comparing checkup data between MS judgment results employing our method and those using conventional diagnostic methods.
  Furthermore, our method can visualize positional relationships between the MS indicators and the three areas on the two-dimensional SOM, promising its application to health guidance.

Content from these authors
© 2010 Japan Society of Health Evaluation and Promotion
Next article
feedback
Top