Computer Identification of P Waves in Electrocardiograms.

Abstract

Algorithms to locate and identify the P wave in single lead electrocardiograms (ECG's) are developed and evaluated on eighteen 12 second electrocardiogram recordings. The program does not assume that the P wave has any specific time relationship to the occurrence of the QRS complex. The program searches the entire ECG to locate the P wave in any part of the ECG. The program divides the ECG into sections and classifies the sections as P waves or some other wave. The first test selects the sections that have a frequency domain pattern that is similar to a training set P wave. The sections that pass this test are checked for a periodic relationship with past selected sections. If a section was selected at a time assumed to be one P wave period before the present one, the present one will be classified as a P wave. Of the 209 P waves in the ECG records, 123 or 58.85% were located correctly. There were 83 false classifications of sections which were not P waves. (Author)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1977
Accession Number
ADA056532

Entities

People

  • Charles A. Flick

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Biomedical
  • C4I
  • Energy and Power Technologies
  • Materials and Manufacturing Processes
  • Space

DTIC Thesaurus Topics

  • Aerospace Medicine
  • Air Force
  • Algorithms
  • Computer Programs
  • Computers
  • Digital Information
  • Electrical Engineering
  • Electrocardiography
  • Feature Selection
  • Frequency
  • Frequency Domain
  • Heart
  • Indicators
  • Monitoring
  • Pattern Recognition
  • Power Spectra
  • Spectra

Readers

  • Business Analytics
  • Cardiovascular Physiology
  • Combustion Dynamics and Shock Wave Physics.