中文摘要：

目的：建立在Gates法基础上的双探头几何均数法使用测量体厚替代估算肾脏深度，理论上应能消除深度估算误差的影响。实验检验双探头几何均数用于Gates法测定GFR的实际可行性。方法：以76名健康者为研究对象，同时采用双血浆法（GFRdt）、Gates单探头法（GFRPOST）及双探头几何均数法（GFRGM）测定肾小球滤过率（GFR）。以GFRdt为参照标准，分析GFRPOST和GFRGM1与GFRdt的一致性。结果：①GFRPOST和GFRGM1与GFRdt的配对t检验均无显著性差异；②GFRPOST和GFRGM1与GFRdt的相关系数分别为0.802和0.722，其相关性GFRPOST高于GFRGM1；③GFRPOST和GFRGM1与GFRdt差值的均值±标准差为（-0.5±12.0）和（3.0±13.9）。结论：双探头几何均数法的准确性不及Gates单探头法，未能实现理论上消除肾脏深度估算误差的目的。

英文摘要：

Objective：Glomerular filtration rate（GFR） is determined by the geometric mean method based on Gates method using dual detector with body thickness instead of kidney depth estimation for using single detector, which can eliminate the error of kidney depth estimation theoretically. The feasibility and accuracy of the geometric mean method were investigated by experiments in this study. Methods： Seventy six cases Community Health residents were selected as the research object. The GFRs（GFRdt, GFRPOST and GFRGM） were determined with two-sample method, Gates method using the post single detector and geometric mean method using double detectors respectively. The accuracy of GFRPOST and GFRGM was analysed and compared using GFRdt as the reference. Results：①Paired t test showed no significant differences between GFRPOST, GFRGM and GFRdt. ②Both GFRdt and GFRPOST changed with GFRdt in linear trend. The correlation coefficients were 0.802 and 0.722 for GFRdt and GFRPOST.③The differences between GFRPOST, GFRGM and GFRdt were-0.5±12.0, 3.0±13.9. Conclusion：The accuracy of the geometric mean method using double detectors is poorer than that of Gates method using the post single detector, which fails to eliminate the error of kidney depth estimation.