Title :
Adaptive time-frequency ECG analysis
Author :
Chouvarda, I. ; Maglaveras, N. ; Pappas, C.
Author_Institution :
Lab of Med. Informdtics, Aristotle Univ., Thessaloniki, Greece
Abstract :
Non-stationary analysis of ECG using the Choi-Williams (CW) distribution (H. Choi & W.J. Williams, 1989) is presented. An analysis was carried out on subjects with acute myocardial infarction (AMI) who had undergone thrombolysis, in V1 Holder recordings. Two groups of patients were looked at: successfully and unsuccessfully thrombolysed patients. The time-frequency map generated by the CW transform (a modification of the Wigner-Ville transform) was divided into nine areas relevant to ECG features. Characteristic parameters were extracted from each area, and linear discriminant analysis was applied to these parameters, leading to a prediction index. Applying the method to ECGs six hours after lysis, the successfully thrombolysed patients were shown to be distinct from the unsuccessfully thrombolysed ones
Keywords :
Wigner distribution; adaptive signal processing; electrocardiography; medical signal processing; time-frequency analysis; transforms; Choi-Williams distribution; Choi-Williams transform; ECG features; V1 Holder recordings; Wigner-Ville transform; acute myocardial infarction; adaptive time-frequency ECG analysis; characteristic parameter extraction; linear discriminant analysis; nonstationary analysis; prediction index; thrombolysis; time-frequency map; Ambient intelligence; Angioplasty; Biomedical informatics; Cardiology; Electrocardiography; Linear discriminant analysis; Linear regression; Robustness; Testing; Time frequency analysis;
Conference_Titel :
Computers in Cardiology 2001
Conference_Location :
Rotterdam
Print_ISBN :
0-7803-7266-2
DOI :
10.1109/CIC.2001.977643