P Wave Detection and Delineation Based on Distances Transform

Automatic detection of P-wave in an electrocardiogram(ECG) is a crucial task in diagnosis of the presence of arrhythmias. In this paper, we proposed an improved method based on distance transform. Distance transform is an efficient method for locating P-wave critical points, corresponding to subjective judgment of cardiologists. Auxiliary algorithms have also been introduced to improve the precision of distance transform, such as recognition of horizontal segments and rising or declining segment, then, feedback comes to narrow or widen the P wave searching area for the need of distance transform. The method is validated through a QT database. The experiments show that our method provides good results. The average error of P-wave onsets, ends and P-wave location is 8.5ms,10.4ms, and 7.6ms, and P wave boundary point detection accuracy is 98.9%, 97.8% respectively.