A Novel Algorithm for the Real-time Analysis of Uterine Contraction Status

Abstract

Against the problems of existing uterine contraction identification algorithms, mainly in the\r inapplicability to real-time identification of the characteristic points of uterine contraction wave, a novel\r method based on a series of criteria for real-time analysis is proposed in this paper. After confirming the\r beginning, end and peak of the analyzed segment, a series of criterion are used to identify the status of\r current uterine contraction point. Meanwhile, the status of current uterine contraction point is corrected\r by a buffer mechanism to ensure accuracy. The experiments show that the conformity between the realtime\r analysis results and two experts' analysis results are high, and the kappa coefficients are as high\r as 0.874 and 0.872. Moreover, the time required for the analysis of a uterine contraction point is only\r 0:0032\u00060:0002 (mean \u00b1 SD) seconds, which is far less than the acquisition and output time of a uterine\r contraction point by pressure probe. Therefore, the proposed method can meet the requirements of\r real-time analysis in both accuracy and efficiency.