Title :
An approach to QRS complex detection using mathematical morphology
Author_Institution :
Dept. of Electr. & Comput. Eng., Toronto Univ., Ont., Canada
Abstract :
An approach to QRS complex detection based on mathematical morphology is presented. QRS complexes are detected by the application of a simple morphological operator. This operator works as a peak-valley extractor and it is controlled by the shape of the structuring element. A set (horizontal line segment) is used as a structuring element, resulting in very fast execution times. The accuracy of this approach has been tested using a standard ECG library; a sensitivity of 99.38% and a positive predictivity of 99.48% have been achieved.
Keywords :
electrocardiography; mathematical morphology; medical signal processing; QRS complex detection; horizontal line segment; mathematical morphology; peak-valley extractor; standard ECG library; structuring element shape; Communication system control; Detectors; Electrocardiography; Hidden Markov models; Libraries; Maximum likelihood detection; Morphological operations; Morphology; Shape control; Testing; Decision Support Techniques; Electrocardiography; Evaluation Studies as Topic; Heart Diseases; Humans; Image Processing, Computer-Assisted; Mathematics; Sensitivity and Specificity; Signal Processing, Computer-Assisted;
Journal_Title :
Biomedical Engineering, IEEE Transactions on
