By Topic

Predicting hyperkalemia by the use of a 12-lead temporal-spatial electrocardiograph: clinical evaluations and model simulations

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

3 Author(s)
W. C. Tzeng ; Chung Hua Univ. ; Y. Z. Chan ; J. C. Hsieh

Predicting hyperkalemia by the use of 12-lead electrocardiograph (ECG) is crucial in clinical emergency practice. However, the diagnostic sensitivity of hyperkalemia (serum potassium level >5.0) is from 38% to 54% based on the interpretation of 12-lead ECG reports. In this study, a serial two-staged classifier implanted with K-means algorithm was created to screen hyperkalemia based on feature parameters extracted from 12-lead ECG. The SCP-ECG reports of 56 normal individuals and 41 hyperkalemia patients (serum potassium level >5.3) were collected in the Emergency Department of Wei-Gong Memorial Hospital-Taiwan from January, 2003 through April, 2005. Each of the 12-lead ECG reports was processed to generate two temporal-spatial graphs representing the signals on the limb leads and on chest leads, respectively. The two integrated T-wave volumes shown on the two graphs were selected as the feature parameters of the first staged classifier in screening hyperkalemia. The second staged classifier, which was composed of feature parameters such as PR interval, QRS duration, and QT interval, was used to correct the misclassification from the first stage. A computer program ECGSIM was also adopted to simulate the ECG waveforms from mild to severe hyperkalemia. Based on the results, the classifier showed the sensitivity of 85% (95% CI=77% to 97%) and the specificity of 79% (95 % CI=55% to 97%) in classifying the samples. In conclusion, the two-staged classifier developed in this study shows the ability to extensively and efficiently diagnose patients with hyperkalemia from very mild to severe degrees. In addition, the estimated values of T-wave volumes from clinical data were in accordance with the observation of computer simulations, where T volumes were increased with the increased serum potassium level

Published in:

Computers in Cardiology, 2005

Date of Conference:

25-28 Sept. 2005