摘要
目的:利用超声结合人工神经网络技术预测血清促黄体激素(luteinizing hormone,LH)基础值。方法:选用71例性早熟女童进行常规超声检查子宫、卵巢,以其中61个病例的子宫体积、卵巢体积以及双侧卵巢最大卵泡内径为输入变量,以LH基础值为输出变量,建立BP神经网络,并用另10个病例作为测试检验,将预测结果与实际值作统计分析。结果:BP网络预测的结果与实际值作线性回归,得到回归直线的斜率为1.0088,回归直线的相关系数为0.977。结论:性早熟女童超声检查结合BP神经网络技术可以预测LH基础值。
Objective:Prediction of the basic luteinizing hormone(LH) level with ultrasonic union artificial neural network.Methods:In 71 girls with precocious puberty,the uterine and ovarian were examined with ultrasound.In which 61 cases of uterine volume,ovarian volume and bilateral ovarian follicles biggest diameter as inputs,the basic LH for output variable.The back-propagation(BP) neural network was established using such data.The basic LH of the other ten cases were predicted by this BP neural network.The prediction results and practical value statistical analyzed.Results:Draw regression line with predicted value of basic LH with BP network and actual value of basic LH of ten test cases and find that regression slope coefficient and correlation coefficient were 1.0088 and 0.977,respectively.Conclusion:The ultrasound union BP neural network can be used to predict the basic LH for the girls with precocious puberty.
出处
《医学影像学杂志》
2011年第5期637-639,共3页
Journal of Medical Imaging
