Adaptive Monte Carlo method for top-down assessment of measurement uncertainty

doi:10.3969/j.issn.1673-8640.2022.04.016

Abstract

Abstract:

Objective To investigate the feasibility of adaptive Monte Carlo method（aMCM）for top-down uncertainty assessment. Methods In case of uncertainty assessment based on top-down assessment in clinical laboratories,aMCM was used for verification. If the absolute difference between aMCM and top-down assessment results was less than the specified numerical tolerance,the verification was passed. The aMCM verification program was developed by MatLab software. Results The absolute difference between the results evaluated by aMCM and the top-down assessment was less than the specified numerical tolerance,and the top-down assessment was verified by aMCM. Conclusions The established aMCM can be used to verify the uncertainty evaluation results of the top-down assessment.

Key words: Monte Carlo method, Adaptive Monte Carlo method, Top-down, Measurement uncertainty

CLC Number:

R446.1

WANG Zhifang, LI Yabo, LIANG Man, AN Shuqi, HAN Yanlin. Adaptive Monte Carlo method for top-down assessment of measurement uncertainty[J]. Laboratory Medicine, 2022, 37(4): 382-386.

Figures/Tables 5

References 10

[1]	Joint Committee for Guides in Metrology. International vocabulary of metrology-basic and general concepts and associated terms(VIM)[S]. JCGM 200:2012,JCGM, 2012.
[2]	International Organization for Standardization. Medical laboratories-requirements for quality and competence[S]. ISO 15189:2012,ISO, 2012.
[3]	Joint Committee for Guides in Metrology. Evaluation of measurement data-guide to the expression of uncertainty in measurement[S]. JCGM 100:2008,JCGM, 2008.
[4]	BS PD ISO/TS 20914-2019 Medical laboratories. Practical guidance for the estimation of measurement uncertainty[S]. 2019.
[5]	Joint Committee for Guides in Metrology. Evaluation of measurement data-supplement 1 to the “guide to the expression of uncertainty in measurement”-propagation of distributions using a Monte Carlo method[S]. JCGM 101:2008,JCGM, 2008.
[6]	周卫平, 陶赟, 戴绘芬, 等. GUF法和蒙特卡洛法在TBil参考测量程序测量结果不确定度评定中的应用[J]. 南通大学学报(医学版), 2019, 39(6):455-459.
[7]	汪静, 张传宝, 曾洁, 等. GUM法与蒙特卡洛法在IFCC 37 ℃肌酸激酶催化活性水平参考测量程序不确定度评定中的应用[J]. 临床检验杂志, 2014, 32(5):382-389.
[8]	张卫威, 景蓉蓉, 季伙燕, 等. 同位素稀释质谱法测定血清肌酐浓度的不确定度评定[J]. 检验医学, 2021, 36(3):263-269.
[9]	王惠民, 王清涛. 临床实验室管理学[M]. 2版. 北京: 高等教育出版社, 2016.
[10]	GOST R ISO 21748-2012 统计方法. 在测量不确定性的评估中重复性,再现性和真实性的评估使用指南[S]. 2012.

测定次数	第1天	第2天	第3天	第4天	第5天
1	140	138	143	143	142
2	140	139	144	143	143
3	140	138	144	142	141
4	141	137	145	143	142
5	140	139	143	142	143

测定次数	第1天	第2天	第3天	第4天	第5天
1	140	138	143	143	142
2	140	139	144	143	143
3	140	138	144	142	141
4	141	137	145	143	142
5	140	139	143	142	143

测定次数	第1天	第2天	第3天	第4天
1	496.5	500.5	499.3	497.4
2	499.7	498.2	495.7	499.9
3	500.1	497.1	496.1	501.1
4	497.2	495.2	495.2	498.6

测定次数	第1天	第2天	第3天	第4天
1	496.5	500.5	499.3	497.4
2	499.7	498.2	495.7	499.9
3	500.1	497.1	496.1	501.1
4	497.2	495.2	495.2	498.6

示例	输入量	不确定度	单位	分布类型	概率分布函数
1	u_RM	0.85	mmol/L	正态	N（0,0.85²）
	u_cal	0.71	mmol/L	正态	N（0,0.71²）
2	u_rel（RW）	4.28	%	正态	N（0,0.094 0²）
	u_cal	1.15	%	正态	N（0,1.15²）
3	u_cal	0.63	mmol/L	正态	N（0,0.65²）
	u_RW	0.85	mmol/L	正态	N（0,0.85²）
	u_bias	0.49	mmol/L	正态	N（0,0.45²）
4	u_rel（RW）	1.11	%	正态	N（0,1.11²）
	u_bias	1.68	%	正态	N（0,1.46²）
5	u_rel（RW）	1.58	%	正态	N（0,1.58²）
	u_bias	1.27	%	正态	N（0,0.70²）