Validation and comparison of five data mining algorithms using big data from clinical laboratories to establish reference intervals of thyroid hormones for older adults

Ma, Chaochao; Zou, Yutong; Hou, Li’an; Yin, Yanzhen; Zhao, Fang; Hu, Yingying; Wang, Danchen; Li, Lei; Cheng, Xinqi; Qiu, Ling

doi:10.1016/j.clinbiochem.2022.05.008

Cited by 15 publications

(6 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is because those previous studies have found different levels of thyroid-related hormones between elderly and non-elderly people [ 28 – 30 ] and one study suggest that indirect comparisons need to be made separately in subgroups [ 31 ]. Therefore, we just included non-elderly people in this study and analysis of data from older people has been reported in the other research [ 32 ].…”

Section: Discussionmentioning

confidence: 99%

Utilization of five data mining algorithms combined with simplified preprocessing to establish reference intervals of thyroid-related hormones for non-elderly adults

Zhong

Hou

et al. 2023

BMC Med Res Methodol

Self Cite

View full text Add to dashboard Cite

Background Despite the extensive research on data mining algorithms, there is still a lack of a standard protocol to evaluate the performance of the existing algorithms. Therefore, the study aims to provide a novel procedure that combines data mining algorithms and simplified preprocessing to establish reference intervals (RIs), with the performance of five algorithms assessed objectively as well. Methods Two data sets were derived from the population undergoing a physical examination. Hoffmann, Bhattacharya, Expectation Maximum (EM), kosmic, and refineR algorithms combined with two-step data preprocessing respectively were implemented in the Test data set to establish RIs for thyroid-related hormones. Algorithm-calculated RIs were compared with the standard RIs calculated from the Reference data set in which reference individuals were selected following strict inclusion and exclusion criteria. Objective assessment of the methods is implemented by the bias ratio (BR) matrix. Results RIs of thyroid-related hormones are established. There is a high consistency between TSH RIs established by the EM algorithm and the standard TSH RIs (BR = 0.063), although EM algorithms seems to perform poor on other hormones. RIs calculated by Hoffmann, Bhattacharya, and refineR methods for free and total triiodo-thyronine, free and total thyroxine respectively are close and match the standard RIs. Conclusion An effective approach for objectively evaluating the performance of the algorithm based on the BR matrix is established. EM algorithm combined with simplified preprocessing can handle data with significant skewness, but its performance is limited in other scenarios. The other four algorithms perform well for data with Gaussian or near-Gaussian distribution. Using the appropriate algorithm based on the data distribution characteristics is recommended.

show abstract

Section: Discussionmentioning

confidence: 99%

Utilization of five data mining algorithms combined with simplified preprocessing to establish reference intervals of thyroid-related hormones for non-elderly adults

Zhong

Hou

et al. 2023

BMC Med Res Methodol

Self Cite

View full text Add to dashboard Cite

show abstract

“…The KOSMIC algorithm was used by Sibtain et al to derive the RIs of TSH in the neonatal age period and found it very useful to derive RIs in paediatric populations in developing countries [ 17 ]. A similar study was conducted by Ma et al using five indirect algorithms to verify RIs of thyroid hormones in older people and recommended the use of modified Hoffman, KOSMIC and refineR algorithms [ 18 ]. However, the performance of indirect methods is dependent on many characteristics of the input dataset, such as larger sample sizes (i.e.…”

Section: Discussionmentioning

confidence: 99%

Verification of Reference Interval of Thyroid Hormones With Manual and Automated Indirect Approaches: Comparison of Hoffman, KOSMIC and refineR Methods

Agaravatt,

Kansara,

Khubchandani

et al. 2023

Cureus

View full text Add to dashboard Cite

Introduction: The interpretation of quantitative test results requires the availability of appropriate reference intervals (RIs). Every laboratory has been advised by scientific literature and reagent manufacturers to establish RIs for all analytes. Measuring RIs using direct methods is very costly, and it poses ethical and practical challenges. To overcome these challenges, indirect methods, such as Hoffman, and newer automated approaches, such as KOSMIC and refineR, are used to verify RIs for thyroid hormones. Objective: To verify RIs for thyroid hormones in adult patients using Hoffman, KOSMIC and refineR methods and to compare these with reference ranges given in kit literature or standard textbooks. Materials and methods: The observed values (results) of thyroid hormone were collected from the LIS (Laboratory Information System) of the Biochemistry Department at the B. J. Medical College and Civil Hospital in Ahmedabad between 1 January 2021 and 31 May 2022. Hoffman, KOSMIC and refineR methods were used to verify the RIs. The computerised Hoffman approach, which Katayev et al. describe, is a simple method for determining RI from hospital data. Zierk et al. pre-validated and suggested the KOSMIC method based on Python programming, whereas refineR was proposed by Tatjana et al. based on R programming language. Results: Hoffman, KOSMIC and refineR’s indirect RI techniques revealed comparable results with kit literature in free T3 and T4, whereas higher upper reference limits of thyroid-stimulating hormone (TSH) compared to kit literature were observed with KOSMIC and refineR methods. However, the computerised Hoffman method revealed comparable results with TSH also. Conclusion: Indirect approaches, such as Hoffman, KOSMIC and refineR, provide reliable RI verification for free T3 and T4 utilising patient samples obtained from LIS. However, the manual Hoffman method provides reliable RI verification for TSH data derived from the hospital population as compared to automated approaches, such as KOSMIC and refineR.

show abstract

“…RIs were calculated as the 2.5–97.5th percentile with a 90 % confidence interval (CI) for upper limits (ULs) and lower limits (LLs) [ 18 ]. Further, the Hoffmann method [ [20] , [21] , [22] ] [ [20] , [21] , [22] ], an indirect method capable of identifying healthy populations from mixed populations, was utilized to further establish discrete RI to demonstrate the enrolled healthy pregnant populations and the applicability of RIs established by non-parametric in this study to pregnant women ( Supplement Table 2 ). Second, gestational-age based continuous RIs were established by the GAMLSS model.…”

Section: Methodsmentioning

confidence: 99%

Establishment of discrete reference interval and next-generation reference interval for copper and zinc during pregnancy using real-world data

Guan,

Ma,

Lin

et al. 2024

Heliyon

Self Cite

View full text Add to dashboard Cite

Validation and comparison of five data mining algorithms using big data from clinical laboratories to establish reference intervals of thyroid hormones for older adults

Cited by 15 publications

References 20 publications

Utilization of five data mining algorithms combined with simplified preprocessing to establish reference intervals of thyroid-related hormones for non-elderly adults

Utilization of five data mining algorithms combined with simplified preprocessing to establish reference intervals of thyroid-related hormones for non-elderly adults

Verification of Reference Interval of Thyroid Hormones With Manual and Automated Indirect Approaches: Comparison of Hoffman, KOSMIC and refineR Methods

Establishment of discrete reference interval and next-generation reference interval for copper and zinc during pregnancy using real-world data

Contact Info

Product

Resources

About