Related Experiment Video
Updated: Mar 22, 2026
14:27
Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
16.5K
The Study of Misclassification Probability in Discriminant Model of Pattern Identification for Stroke
1KM Fundamental Research Division, Korea Institute of Oriental Medicine, Daejeon 305-811, Republic of Korea.
Evidence-Based Complementary and Alternative Medicine : Ecam
|April 19, 2016
Summary
This study improved pattern identification (PI) accuracy for stroke diagnosis in traditional Korean medicine (TKM). New measures identified misclassified cases, aiding diagnostic standard development.
Area of Science:
- Integrative Medicine
- Traditional Korean Medicine (TKM)
- Diagnostic Accuracy
Background:
- Pattern identification (PI) is fundamental to TKM diagnosis.
- Stroke diagnosis in TKM requires accurate PI.
- Improving PI classification accuracy is crucial for patient outcomes.
Purpose of the Study:
- Identify misclassification objects within the PI discriminant model for stroke.
- Enhance the overall classification accuracy of PI in TKM for stroke patients.
- Develop novel measures to detect and correct diagnostic errors.
Main Methods:
- Included 3306 stroke patients from 15 TKM hospitals (June 2006-December 2012).
- Derived four measures (D, R, S, C scores) from profile graph patterns.
- Applied measures to evaluate misclassification detection effectiveness.
Main Results:
- C score showed the highest misclassification rate (42.60%) in 10-20% filtered data.
- R score had the highest misclassification rate (40.32%) in 30% filtered data.
- D score exhibited the highest misclassification rate across 40-90% filtered data.
Conclusions:
- The developed measures effectively identify misclassification objects in TKM PI for stroke.
- Findings support the refinement of diagnostic standards in TKM.
- Further research can validate these measures for broader clinical application.
