A sample with a blood clot may produce an inaccurate outcome in coagulation testing, which may mislead clinicians into making improper clinical decisions. Currently, there is no efficient method to automatically detect clots. This study demonstrates the feasibility of utilizing machine learning (ML) to identify clotted specimens.
The results of coagulation testing with 192 clotted samples and 2,889 no-clot-detected (NCD) samples were retrospectively retrieved from a laboratory information system to form the training dataset and testing dataset. Standard and momentum backpropagation neural networks (BPNNs) were trained and validated using the training dataset with a five-fold cross-validation method. The predictive performances of the models were then assessed based on the testing dataset.
Our results demonstrated that there were intrinsic distinctions between the clotted and NCD specimens regarding differences in the testing results and the separation of the groups (clotted and NCD) in the t-SNE analysis. The standard and momentum BPNNs could identify the sample status (clotted and NCD) with areas under the ROC curves of 0.966 (95% CI, 0.958–0.974) and 0.971 (95% CI, 0.9641–0.9784), respectively.
Here, we have described the application of ML algorithms in identifying the sample status based on the results of coagulation testing. This approach provides a proof-of-concept application of ML algorithms to evaluate the sample quality, and it has the potential to facilitate clinical laboratory automation.
Funding source: Department of Health of Zhejiang Province
Award Identifier / Grant number: 2021KY845
Research funding: This work was supported by the Science Fund of the Health Department of Zhejiang Province. Project ID: 2021KY845.
Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.
Competing interests: Authors state no conflict of interest.
Ethical approval: The study was approved by the institutional research ethics committee of The Third Affiliated Hospital of Zhejiang Chinese Medical University. Approval ID: ZSLL-KY-2021-001-01.
1. Adcock Funk, D, Lippi, G, Favaloro, E. Quality standards for sample processing, transportation, and storage in hemostasis testing. Semin Thromb Hemost 2012;38:576–85. https://doi.org/10.1055/s-0032-1319768.Search in Google Scholar
2. Das, N, Topalovic, M, Janssens, W. Artificial intelligence in diagnosis of obstructive lung disease: current status and future potential. Curr Opin Pulm Med 2018;24:117–23. https://doi.org/10.1097/mcp.0000000000000459.Search in Google Scholar
4. Cleophas, TJ, Cleophas, TF. Artificial intelligence for diagnostic purposes: principles, procedures and limitations. Clin Chem Lab Med 2010;48:159–65. https://doi.org/10.1515/cclm.2010.045.Search in Google Scholar
5. Givens, TB, Braun, P, Fischer, TJ. Predicting the presence of plasma heparin using neural networks to analyze coagulation screening assay optical profiles. Comput Biol Med 1996;26:463–76. https://doi.org/10.1016/s0010-4825(96)00023-6.Search in Google Scholar
6. Han, Q, Zheng, W, Guo, X-D, Zhang, D, Liu, H-F, Yu, L, et al.. A new predicting model of preeclampsia based on peripheral blood test value. Eur Rev Med Pharmacol Sci 2020;24:7222–9. https://doi.org/10.26355/eurrev_202007_21874.Search in Google Scholar PubMed
8. Mishra, A, Ashraf, MZ. Using artificial intelligence to manage thrombosis research, diagnosis, and clinical management. Semin Thromb Hemost 2020.10.1055/s-0039-1697949Search in Google Scholar PubMed
9. Gunnur Dikmen, Z, Pinar, A, Akbiyik, F. Specimen rejection in laboratory medicine: necessary for patient safety? Biochem Med 2015;25:377–85. https://doi.org/10.11613/bm.2015.037.Search in Google Scholar
10. Van Der Maaten, L. Accelerating t-SNE using tree-based algorithms. J Mach Learn Res 2015;15:3221–45.Search in Google Scholar
13. Smith, SA, Travers, RJ, Morrissey, JH. How it all starts: initiation of the clotting cascade. Crit Rev Biochem Mol Biol 2015;50:326–36. https://doi.org/10.3109/10409238.2015.1050550.Search in Google Scholar PubMed PubMed Central
15. Magnette, A, Chatelain, M, Chatelain, B, Ten Cate, H, Mullier, F. Pre-analytical issues in the haemostasis laboratory: guidance for the clinical laboratories. Thromb J 2016;14:49. https://doi.org/10.1186/s12959-016-0123-z.Search in Google Scholar PubMed PubMed Central
16. Dahlbäck, B. Blood coagulation and its regulation by anticoagulant pathways: genetic pathogenesis of bleeding and thrombotic diseases. J Intern Med 2005;257:209–23. https://doi.org/10.1111/j.1365-2796.2004.01444.x.Search in Google Scholar PubMed
18. Thachil, J, Lippi, G, Favaloro, EJ. D-dimer testing: laboratory aspects and current issues. In: Favaloro, EJ, Lippi, G, editors. New York, NY: Springer; 2017;91–104, vol 1646.10.1007/978-1-4939-7196-1_7Search in Google Scholar PubMed
20. Buchtele, N, Schober, A, Schoergenhofer, C, Spiel, AO, Mauracher, L, Weiser, C, et al.. Added value of the DIC score and of D-dimer to predict outcome after successfully resuscitated out-of-hospital cardiac arrest. Eur J Intern Med 2018;57:44–8. https://doi.org/10.1016/j.ejim.2018.06.016.Search in Google Scholar PubMed
22. Naugler, C, Church, DL. Automation and artificial intelligence in the clinical laboratory. Crit Rev Clin Lab Sci 2019;56:98–110. https://doi.org/10.1080/10408363.2018.1561640.Search in Google Scholar PubMed
24. Gruson, D, Helleputte, T, Rousseau, P, Gruson, D. Data science, artificial intelligence, and machine learning: opportunities for laboratory medicine and the value of positive regulation. Clin Biochem 2019;69:1–7. https://doi.org/10.1016/j.clinbiochem.2019.04.013.Search in Google Scholar PubMed
25. Lyu, J, Zhang, J. BP neural network prediction model for suicide attempt among Chinese rural residents. J Affect Disord 2019;246:465–73. https://doi.org/10.1016/j.jad.2018.12.111.Search in Google Scholar PubMed PubMed Central
29. Yang, W, Liu, X, Wang, K, Hu, J, Geng, G, Feng, J. Sex determination of three-dimensional skull based on improved backpropagation neural network. Comput Math Methods Med 2019;2019:9163547. https://doi.org/10.1155/2019/9163547.Search in Google Scholar PubMed PubMed Central
30. Du, KL, Swamy, MNS. Neural networks in a softcomputing framework. Neural Netw Softcomput Framew 2006:1–566.Search in Google Scholar
31. Wang, J, Yang, J, Wu, W. Convergence of cyclic and almost-cyclic learning with momentum for feedforward neural networks. IEEE Trans Neural Netw 2011;22:1297–306. https://doi.org/10.1109/tnn.2011.2159992.Search in Google Scholar
32. Toulon, P, Berruyer, M, Lasne, D, Telion, C, Arcizet, J, Giacomello, R, et al.. Results of a multicentre study aimed at defining the age-specific reference ranges. Thromb Haemost 2016.Search in Google Scholar
33. Ho, P, Ng, C, Rigano, J, Tacey, M, Smith, C, Donnan, G, et al.. Significant age, race and gender differences in global coagulation assays parameters in the normal population. Thromb Res 2017;154:80–3. https://doi.org/10.1016/j.thromres.2017.04.009.Search in Google Scholar PubMed
The online version of this article offers supplementary material (https://doi.org/10.1515/cclm-2021-0081).
© 2021 Walter de Gruyter GmbH, Berlin/Boston