图像分析表征定性血清hCG侧向流免疫分析剂量反应曲线
本研究通过图像分析表征定性血清hCG侧向流免疫分析的剂量反应曲线(测试线可见度),以理解随hCG浓度增加从阴性到阳性的转变。测试线颜色积分S与hCG浓度关系符合双曲线S=Smax[hCG]/([hCG]+K),K=202 mIU/ml(r=0.997),K处 replicates 的CV为7.3%。用户目视判读的不确定性仅出现在检测声明的灵敏度10 mIU/ml以下。
要点
- 测试线信号与hCG浓度呈双曲线关系,K=202 mIU/ml。
- K处重复检测的CV为7.3%。
- 目视判读不确定性仅出现在低于声明灵敏度10 mIU/ml时。
核心结论
- 该定性血清hCG侧向流免疫分析的剂量反应曲线(测试线积分)为简单双曲线。
- 剂量反应曲线表征有助于验证该检测的分析性能特征。
研究设计与样本
- 使用已知hCG浓度的血清样本,在规定的读取时间5分钟用扫描仪获取装置图像。
- 采用Python和R进行图像分析,获得测试线颜色积分S作为hCG浓度的函数。
主要发现
- S与hCG浓度关系符合双曲线S=Smax[hCG]/([hCG]+K),其中K=202 mIU/ml(r=0.997)。
- K处S重复检测的CV为7.3%。
- 用户目视判读的不确定性仅出现在检测声明灵敏度10 mIU/ml以下,对应S<Smax的3%且信噪比<3的区域。
临床怎么用
- 图像分析可用于验证定性hCG侧向流免疫分析的剂量反应特征。
- 目视判读在高于声明灵敏度10 mIU/ml时结果不确定性低。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND: As an adjunct to verification of performance characteristics of a qualitative serum hCG lateral flow immunoassay (LFI), we performed image analysis to characterize the dose vs response curve (visibility of the test line), as a means of understanding the transition from negative to positive as a function of increasing [hCG]. METHODS: Using serum samples of known [hCG], device images were obtained using a scanner at the prescribed reading time (5 min). Image analysis (using Python and R) was used to obtain the integral (S) of the test-line color as a function of [hCG]. RESULTS: Data for S as a function of [hCG] were well characterized by a simple hyperbola: S = Smax [hCG]/([hCG] + K), where K = 202 mIU/ml (r = 0.997). Replicates of S at K had CV of 7.3 %. By eye, uncertainty of test results among users occurred only below the assay's stated sensitivity of 10 mIU/ml, in region of S < 3 % of Smax, and signal:noise ratio < 3. CONCLUSIONS: By image analysis, the dose vs response (Test line integral) for this qualitative serum hCG LFI was a simple hyperbola. Characterization of the dose vs response curve was useful in verification of the assay's performance characteristics.
引用
- Mattiello CJ, Stickle DF. Characterization by image analysis of the dose vs response curve for a qualitative serum hCG lateral flow immunoassay. Clin Chim Acta. 2022. PMID: 36423702. doi:10.1016/j.cca.2022.11.020
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。