Identification of term and preterm labor in rats using artificial neural networks on uterine electromyography signals
Identification of term and preterm labor in rats using artificial neural networks on uterine electromyography signals
复制标题
DOI：
10.1016/j.ajog.2007.08.039
复制
发表时间：
2008-02-01
影响因子：
9.8
通讯作者：
Garfield, Robert E.
中科院分区：
文献类型：
--
作者：
Shi, Shao-Qing;Maner, William L.;Garfield, Robert E.
目的：本研究采用人工神经网络对子宫肌电图数据，以确定长期和早产rats.Study设计：对照组（组1：n = 4）和早产模型（组2：n = 4，治疗与奥那司酮）。子宫肌电图和子宫内压（IUP）变量通过植入遥测装置测量。对于评估的每个时间点，首先通过视觉和其他方法指定“分娩事件”或“非分娩事件”。共观察到112例分娩和非分娩事件。人工神经网络，然后使用肌电图和子宫内压参数，试图算法，客观识别劳动时间在每个group.Results：组1，所有8（100%）劳动事件和所有44（100%）非劳动事件被正确识别的人工神经网络。对于第2组，22的24（92%）劳动事件和31的36（86%）nonlabour事件被正确地确定了人工神经网络。结论：人工神经网络可以有效地预测足月和早产在怀孕期间使用子宫肌电图和宫内压力变量。
OBJECTIVE: This study was undertaken to use artificial neural networks on uterine electromyography data to identify term and preterm labor in rats.STUDY DESIGN: Controls ( group 1: n = 4) and preterm labor models (group 2: n = 4, treated with onapristone) were used. Uterine electromyography and intrauterine pressure (IUP) variables were measured by implanted telemetric devices. For each timepoint assessed, either a "labor event" or "nonlabor event" was first assigned by using visual and other means. 112 total labor and nonlabor events were observed. Artificial neural networks were then used with electromyography and intrauterine pressure parameters to attempt algorithmic, objective identification for time of labor in each group.RESULTS: For group 1, all 8 (100%) labor events and all 44 (100%) nonlabor events were correctly identified by the artificial neural networks. For group 2, 22 of 24 (92%) labor events and 31 of 36 (86%) nonlabor events were correctly determined by the artificial neural networks.CONCLUSION: Artificial neural networks can effectively predict term and preterm labor during pregnancy with the use of uterine electromyography and intrauterine pressure variables.