| 中 华 妇 产 科 杂 志 CHINESE JOURNAL OF OBSTETRICS AND GYNECOLOGY 1998年 第11期 |
科技期刊 |
|
胎心率与宫缩描记图的计算机数码转换
邵浩达 阮邦武 王益夫 钟国衡 张明仁 【摘要】 目的 建立胎心率与宫缩描记图的计算机数码转换方法。方法 采用平台扫描器,摄取40例胎儿的胎心率与宫缩描记图,将扫描成像先进行倾斜校正,然后利用对比增强法,去除记录纸上原有的格子背景和手写的注解,仅留下胎心率和宫缩压力曲线,进行数码转换。通过比较实时的记录和以本方法转换的胎心率数值的差异进行方法学鉴定。结果 实时记录与推算值的平均差异为-0.26~-1.26次,其95%可信限值为-7~5次。结论 使用本方法可以有效地将许多贮存的胎心率与宫缩描记图进行数码转换,以便进一步进行回顾性计算机分析研究,为研制高精度的计算机化胎儿预测系统,提供极有价值的资料。 Computerized Transformation of the Cardiotocographic Paper Record to Its Digital Equivalent for Computerised Analysis Shao Haoda, Ruan Bangwu, Wang Yifu, et al. The Chinese University of Hong Kong 【Abstract 】 Objective To analyze cardiotocograph by computer, the tracings recorded in paper form must first be converted into their digital equivalent. We developed a method by which this process may be performed. Methods Paper recordings were first scanned using a conventional flat bed scanner to obtain a digital image. Each image was firstly corrected for rotational misalignment error during scanning and, sceondly the grid was removed by performing logistic contrast enhancement to leave the discrete fetal heart rate and tocographic tracings. The method was validated by comparing differences between the fetal heart rate obtained from the paper record with that directly obtained from the fetal monitor. Results Forty recordings were analyzed. The mean difference per recording between the actual and derived values ranged from -0.26~-1.26 beats per minute. The 95% confidence interval for the pooled differences between the derived and actual fetal heart rate values was -7~5 beats per minute. Conclusion By using the techniques described in this paper, it is now possible to convert the large number of paper records available so that they can be analyzed by computerized cardiotocograph interpreters. 依靠目测对胎心率与宫缩描记图(cardiotocograph, CTG)的判别往往缺乏一致性和重复性,这与判读者本身的经验[1,2]和判读者之间的差异有关。利用计算机进行CTG分析,理论上可以摒除判读者的主观性[3,4]。近年来,这方面的研究主要集中在如何提高CTG分析的预测作用。目前,多数CTG计算机分析系统的研制,系基于大量的电子信号数据的收集,广泛地测试不同临床情况下的运算结果,藉以评定产前的胎心率和胎儿活动的关系。但有关产程内的电子信号数据资料十分有限,且胎儿监护的结果均以CTG图纸留存,由于这些描记在记录纸上的信息不是以数码的形式存在,故不能直接借助计算机进行分析。本研究旨在探索一种在计算机的辅助下,将原描记在图纸上的CTG资料以数码形式转换,以便用于CTG的计算机分析。 资料与方法 1.CTG图像的摄取:首先设定纵横解像度为每厘米60点,并且固定亮度和对比度,然后扫描摄取图像,其数码解像度在纵轴和横轴上均为每厘米60个画素点(pixel)。 结果 40例产程中胎心率实际值和CTG描记图的计算机数码转换值均为负数,此差别的范围为-0.26~-1.26次,其95%可信限为-7.5~5次。见附表。 讨论 新一代的胎儿监护仪已可以直接录取胎心率与宫缩压力的数码数据,输入计算机胎心率分析系统进行分析和作为监测胎儿缺氧的指标[7~9]。可是,目前大多数产科临床医师仍使用热敏记录纸描记胎心率和宫缩压力。已在记录纸上积累的资料,特别是产程内的CTG资料,对于进行胎儿预测因素的回顾性研究是十分有用的,但要利用这些资料,必须先将其转换为数码形式,才能用于计算机分析系统中。 附表 40例胎心率实测值与计算值之间的差异(次/分) |
| 病例 序号 |
差异的范围 | ||
| 最低值 | 最高值 | ||
|
1 |
-7 |
4 |
-0.83±1.13 |
| 2 | -21 | 10 | -0.65±2.60 |
| 3 | -23 | 13 | -1.06±3.06 |
| 4 | -38 | 13 | -1.04±2.50 |
| 5 | -19 | 29 | -1.14±4.04 |
| 6 | -12 | 10 | -0.46±1.64 |
| 7 | -8 | 9 | -0.90±1.46 |
| 8 | -21 | 11 | -0.47±2.39 |
| 9 | -8 | 11 | -0.69±1.67 |
| 10 | -23 | 22 | -0.53±3.24 |
| 11 | -25 | 22 | -0.27±2.49 |
| 12 | -16 | 21 | -0.76±2.48 |
| 13 | -69 | 97 | -0.67±5.94 |
| 14 | -14 | 15 | -0.67±1.93 |
| 15 | -19 | 9 | -0.97±1.79 |
| 16 | -15 | 12 | -0.79±2.26 |
| 17 | -16 | 18 | -0.83±2.51 |
| 18 | -23 | 14 | -1.13±2.72 |
| 19 | -15 | 6 | -0.92±1.69 |
| 20 | -7 | 8 | -1.11±1.01 |
| 21 | -8 | 8 | -1.17±1.35 |
| 22 | -14 | 12 | -1.01±2.33 |
| 23 | -14 | 15 | -1.01±2.79 |
| 24 | -14 | 15 | -0.82±2.35 |
| 25 | -16 | 13 | -1.06±2.73 |
| 26 | -18 | 15 | -1.26±2.44 |
| 27 | -21 | 28 | -0.63±3.55 |
| 28 | -13 | 9 | -0.82±2.18 |
| 29 | -20 | 24 | -0.52±3.39 |
| 30 | -22 | 23 | -0.67±5.99 |
| 31 | -15 | 21 | -0.26±3.33 |
| 32 | -28 | 36 | -0.58±6.73 |
| 33 | -21 | 18 | -0.53±3.04 |
| 34 | -11 | 24 | -0.73±2.01 |
| 35 | -69 | 21 | -0.79±4.96 |
| 36 | -22 | 17 | -0.39±3.98 |
| 37 | -29 | 14 | -0.74±2.92 |
| 38 | -16 | 12 | -0.44±2.29 |
| 39 | -22 | 20 | -0.37±3.16 |
| 40 | -10 | 13 | -0.76±1.76 |
|
作者单位:香港中文大学妇产科学系 参考文献 1 Nielsen PV, Stigsby B, Nickelson C, et al. Intra- and inter-observer variability in the assessment of intrapartum cardiotocograms. Acta Obstet Gynecol Scand, 1987, 66:470-479. (收稿:1998-06-05 修回:1998-08-27) (本文编辑:潘伟) |