Startseite Medizin Expert-level detection of M-proteins in serum protein electrophoresis using machine learning
Artikel
Lizenziert
Nicht lizenziert Erfordert eine Authentifizierung

Expert-level detection of M-proteins in serum protein electrophoresis using machine learning

  • Eike Elfert , Wolfgang E. Kaminski , Christian Matek , Gregor Hoermann ORCID logo , Eyvind W. Axelsen , Carsten Marr und Armin P. Piehler ORCID logo EMAIL logo
Veröffentlicht/Copyright: 17. Juni 2024
Veröffentlichen auch Sie bei De Gruyter Brill
Manuskript einreichen Informationen für Autor*innen
Clinical Chemistry and Laboratory Medicine (CCLM)
Aus der Zeitschrift Clinical Chemistry and Laboratory Medicine (CCLM) Band 62 Heft 12

Abstract

Objectives

Serum protein electrophoresis (SPE) in combination with immunotyping (IMT) is the diagnostic standard for detecting monoclonal proteins (M-proteins). However, interpretation of SPE and IMT is weakly standardized, time consuming and investigator dependent. Here, we present five machine learning (ML) approaches for automated detection of M-proteins on SPE on an unprecedented large and well-curated data set and compare the performance with that of laboratory experts.

Methods

SPE and IMT were performed in serum samples from 69,722 individuals from Norway. IMT results were used to label the samples as M-protein present (positive, n=4,273) or absent (negative n=65,449). Four feature-based ML algorithms and one convolutional neural network (CNN) were trained on 68,722 randomly selected SPE patterns to detect M-proteins. Algorithm performance was compared to that of an expert group of clinical pathologists and laboratory technicians (n=10) on a test set of 1,000 samples.

Results

The random forest classifier showed the best performance (F1-Score 93.2 %, accuracy 99.1 %, sensitivity 89.9 %, specificity 99.8 %, positive predictive value 96.9 %, negative predictive value 99.3 %) and outperformed the experts (F1-Score 61.2 ± 16.0 %, accuracy 89.2 ± 10.2 %, sensitivity 94.3 ± 2.8 %, specificity 88.9 ± 10.9 %, positive predictive value 47.3 ± 16.2 %, negative predictive value 99.5 ± 0.2 %) on the test set. Interestingly the performance of the RFC saturated, the CNN performance increased steadily within our training set (n=68,722).

Conclusions

Feature-based ML systems are capable of automated detection of M-proteins on SPE beyond expert-level and show potential for use in the clinical laboratory.

Keywords: artificial intelligence; monoclonal gammopathy; myeloma; electrophoresis; machine learning
Corresponding author: Armin P. Piehler, MD, PhD, MLL Munich Leukemia Laboratory GmbH, Max-Lebsche-Platz 31, 81377 Munich, Germany; and Fürst Medical Laboratory, 1051 Oslo, Norway, E-mail:
Eike Elfert and Wolfgang E. Kaminski contributed equally to this work as first authors. Carsten Marr and Armin P. Piehler contributed equally to this work as senior authors.

Funding source: H2020 European Research Council

Award Identifier / Grant number: 866411

Award Identifier / Grant number: 101113551

Acknowledgments

We would like to thank the laboratory technicians at Fürst Medical Laboratory for their participation in the study and their interpretation of protein electrophoresis curves in order to compare performance of the machine learning model with human experts.

  1. Research ethics: The study was approved by the Ethics Committee of the University of Heidelberg (2018-548N-MA), Germany, and the Regional Ethics Committee REC South-East, Norway (231395).

  2. Informed consent: The data used was completely anonymized. Tracing back to single individuals was not possible.

  3. Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission. Conceptualization: Wolfgang E. Kaminski (lead), Armin P. Piehler, Carsten Marr, Christian Matek and Eike Elfert. Data curation: Eyvind W. Axelsen, Eike Elfert, Wolfgang E. Kaminski. Formal Analysis: Eike Elfert (lead), Armin P. Piehler, Carsten Marr, Christian Matek. Funding acquisition: Carsten Marr. Investigation: Eike Elfert (lead), Armin P. Piehler, Carsten Marr, Christian Matek. Methodology: Armin P. Piehler, Carsten Marr, Christian Matek, Wolfgang E. Kaminski, Eike Elfert. Project administration: Armin P. Piehler, Carsten Marr, Christian Matek, Wolfgang E. Kaminski, Eike Elfert. Resources: Eyvind W. Axelsen. Software: Eike Elfert, Eyvind W. Axelsen. Supervision: Wolfgang E. Kaminski, Armin P. Piehler, Carsten Marr, Christian Matek. Validation: Eike Elfert, Armin P. Piehler, Carsten Marr, Christian Matek. Visualization: Eike Elfert. Writing – original draft: Eike Elfert, Wolfgang E. Kaminski, Christian Matek, Gregor Hoermann, Eyvind W. Axelsen, Carsten Marr, Armin P. Piehler. Writing – review & editing: Eike Elfert, Christian Matek, Carsten Marr, Armin P. Piehler.

  4. Competing interests: The authors state no conflict of interest.

  5. Research funding: Funding from the European Research Council (ERC), grant agreement no. 866411 and 101113551.

  6. Data availability: The raw data can be obtained on request from the corresponding author.

References

1. Kyle, RA, Larson, DR, Therneau, TM, Dispenzieri, A, Kumar, S, Cerhan, JR, et al.. Long-term follow-up of monoclonal gammopathy of undetermined significance. N Engl J Med 2018;378:241–9. https://doi.org/10.1056/nejmoa1709974.Suche in Google Scholar PubMed PubMed Central

2. Bray, F, Ferlay, J, Soerjomataram, I, Siegel, RL, Torre, LA, Jemal, A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA A Cancer J Clin 2018;68:394–424. https://doi.org/10.3322/caac.21492.Suche in Google Scholar PubMed

3. Rajkumar, SV, Dimopoulos, MA, Palumbo, A, Blade, J, Merlini, G, Mateos, MV, et al.. International Myeloma Working Group updated criteria for the diagnosis of multiple myeloma. Lancet Oncol 2014;15:538–48. https://doi.org/10.1016/s1470-2045(14)70442-5.Suche in Google Scholar

4. Tate, JR. The paraprotein – an enduring biomarker. Clin Biochem Rev 2019;40:5–22.Suche in Google Scholar

5. Harris, NS, Winter, WE. Multiple myeloma and related serum protein disorders: an electrophoretic guide. New York: Demos Medical Publishing; 2012.Suche in Google Scholar

6. Thoren, KL, McCash, SI, Murata, K. Immunotyping provides equivalent results to immunofixation in a population with a high prevalence of monoclonal gammopathies. J Appl Lab Med 2021;6:1551–60. https://doi.org/10.1093/jalm/jfab067.Suche in Google Scholar PubMed PubMed Central

7. Cárdenas, MC, García-Sanz, R, Puig, N, Pérez-Surribas, D, Flores-Montero, J, Ortiz-Espejo, M, et al.. Recommendations for the study of monoclonal gammopathies in the clinical laboratory. A consensus of the Spanish Society of Laboratory Medicine and the Spanish Society of Hematology and Hemotherapy. Part I: update on laboratory tests for the study of monoclonal gammopathies. Clin Chem Lab Med 2023;61:2115–30. https://doi.org/10.1515/cclm-2023-0326.Suche in Google Scholar PubMed

8. Cárdenas, MC, García-Sanz, R, Puig, N, Pérez-Surribas, D, Flores-Montero, J, Ortiz-Espejo, M, et al.. Recommendations for the study of monoclonal gammopathies in the clinical laboratory. A consensus of the Spanish Society of Laboratory Medicine and the Spanish Society of Hematology and Hemotherapy. Part II: methodological and clinical recommendations for the diagnosis and follow-up of monoclonal gammopathies. Clin Chem Lab Med 2023;61:2131–42. https://doi.org/10.1515/cclm-2023-0325.Suche in Google Scholar

9. O’Connell, TX, Horita, TJ, Kasravi, B. Understanding and interpreting serum protein electrophoresis. Am Fam Physician 2005;71:105–12.Suche in Google Scholar

10. Font, P, Loscertales, J, Soto, C, Ricard, P, Novas, CM, Martín-Clavero, E, et al.. Interobserver variance in myelodysplastic syndromes with less than 5 % bone marrow blasts: unilineage vs. multilineage dysplasia and reproducibility of the threshold of 2 % blasts. Ann Hematol 2015;94:565–73. https://doi.org/10.1007/s00277-014-2252-4.Suche in Google Scholar PubMed

11. Fuentes-Arderiu, X, Dot-Bach, D. Measurement uncertainty in manual differential leukocyte counting. Clin Chem Lab Med 2009;47:112–5. https://doi.org/10.1515/cclm.2009.014.Suche in Google Scholar

12. Matek, C, Schwarz, S, Spiekermann, K, Marr, C. Human-level recognition of blast cells in acute myeloid leukaemia with convolutional neural networks. Nat Mach Intell 2019;1:538–44. https://doi.org/10.1038/s42256-019-0101-9.Suche in Google Scholar

13. Esteva, A, Kuprel, B, Novoa, RA, Ko, J, Swetter, SM, Blau, HM, et al.. Dermatologist-level classification of skin cancer with deep neural networks. Nature 2017;542:115–8. https://doi.org/10.1038/nature21056.Suche in Google Scholar PubMed PubMed Central

14. Shalev-Shwartz, S, Ben-David, S. Understanding machine learning: from theory to algorithms. New York: Cambridge University Press; 2014.10.1017/CBO9781107298019Suche in Google Scholar

15. Bizopoulos, P, Koutsouris, D. Deep learning in cardiology. IEEE Rev Biomed Eng 2019;12:168–93. https://doi.org/10.1109/rbme.2018.2885714.Suche in Google Scholar PubMed

16. McBee, MP, Awan, OA, Colucci, AT, Ghobadi, CW, Kadom, N, Kansagra, AP, et al.. Deep learning in radiology. Acad Radiol 2018;25:1472–80. https://doi.org/10.1016/j.acra.2018.02.018.Suche in Google Scholar PubMed

17. Yang, H-C, Islam, MM, Jack Li, Y-C. Potentiality of deep learning application in healthcare. Comput Methods Progr Biomed 2018;161:A1. https://doi.org/10.1016/j.cmpb.2018.05.014.Suche in Google Scholar PubMed

18. Altinier, S, Sarti, L, Varagnolo, M, Zaninotto, M, Maggini, M, Plebani, M. An expert system for the classification of serum protein electrophoresis patterns. Clin Chem Lab Med 2008;46:1458–63. https://doi.org/10.1515/cclm.2008.284.Suche in Google Scholar

19. Kratzer, MA, Ivandic, B, Fateh-Moghadam, A. Neuronal network analysis of serum electrophoresis. J Clin Pathol 1992;45:612–5. https://doi.org/10.1136/jcp.45.7.612.Suche in Google Scholar PubMed PubMed Central

20. Ognibene, A, Graziani, MS, Caldini, A, Terreni, A, Righetti, G, Varagnolo, MC, et al.. Computer-assisted detection of monoclonal components: results from the multicenter study for the evaluation of CASPER (Computer Assisted Serum Protein Electrophoresis Recognizer) algorithm. Clin Chem Lab Med 2008;46:1183–8. https://doi.org/10.1515/cclm.2008.221.Suche in Google Scholar PubMed

21. Chen, R, Jaye, DL, Roback, JD, Sherman, MA, Smith, GH. Automated serum protein electrophoresis interpretation using machine learning-based algorithm for paraprotein detection. Am J Clin Pathol 2020;154:S7–8. https://doi.org/10.1093/ajcp/aqaa137.013.Suche in Google Scholar

22. Chabrun, F, Dieu, X, Ferre, M, Gaillard, O, Mery, A, Chao de la Barca, JM, et al.. Achieving expert-level interpretation of serum protein electrophoresis through deep learning driven by human reasoning. Clin Chem 2021;67:1406–14. https://doi.org/10.1093/clinchem/hvab133.Suche in Google Scholar PubMed

23. Pedregosa, F, Varoquaux, G, Gramfort, A, Michel, V, Thirion, B, Grisel, O, et al.. Scikit-learn: machine learning in Python. J Mach Learn Res 2011;12:2825–30.Suche in Google Scholar

24. Cournapeau, D. scikit-learn: machine learning in Python — scikit-learn 1.3.1 documentation; 2007. Available from: https://scikit-learn.org/stable [Accessed 25 Jan 2024].Suche in Google Scholar

25. Ciregan, D, Meier, U, Schmidhuber, J. Multi-column deep neural networks for image classification. In: 2012 IEEE conference on computer vision and pattern recognition. Providence: Institute of Electrical and Electronics Engineers (IEEE); 2012:3642–9 pp.10.1109/CVPR.2012.6248110Suche in Google Scholar

26. Russakovsky, O, Deng, J, Su, H, Krause, J, Satheesh, S, Ma, S, et al.. ImageNet large scale visual recognition challenge. Int J Comput Vis 2015;115:211–52. https://doi.org/10.1007/s11263-015-0816-y.Suche in Google Scholar

27. Breiman, L. Random forests. Mach Learn 2001;45:5–32.10.1023/A:1010933404324Suche in Google Scholar

28. Geurts, P, Ernst, D, Wehenkel, L. Extremely randomized trees. Mach Learn 2006;63:3–42. https://doi.org/10.1007/s10994-006-6226-1.Suche in Google Scholar

29. Freund, Y. Boosting a weak learning algorithm by majority. Inf Comput 1995;121:256–85. https://doi.org/10.1006/inco.1995.1136.Suche in Google Scholar

30. Friedman, JH. Stochastic gradient boosting. Comput Stat Data Anal 2002;38:367–78. https://doi.org/10.1016/s0167-9473(01)00065-2.Suche in Google Scholar

31. LeCun, Y, Bengio, Y, Hinton, G. Deep learning. Nature 2015;521:436–44. https://doi.org/10.1038/nature14539.Suche in Google Scholar PubMed

32. Liu, J, Osadchy, M, Ashton, L, Foster, M, Solomon, CJ, Gibson, SJ. Deep convolutional neural networks for Raman spectrum recognition: a unified solution. Analyst 2017;142:4067–74. https://doi.org/10.1039/c7an01371j.Suche in Google Scholar PubMed

33. Chollet, F. Deep learning mit Python und Keras: Das Praxis-Handbuch vom Entwickler der Keras-Bibliothek. Frechen: MITP-Verlags GmbH & Co. KG; 2018.Suche in Google Scholar

34. Chollet, F. Keras documentation; 2015. Available from: https://keras.io [Accessed 25 Jan 2024].Suche in Google Scholar

35. Hanley, JA, McNeil, BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29–36. https://doi.org/10.1148/radiology.143.1.7063747.Suche in Google Scholar PubMed

36. Guyon, I, Gunn, S, Nikravesh, M, Zadeh, LA. Feature extraction: foundations and applications. Berlin: Springer; 2008.Suche in Google Scholar

37. Atkinson, KE. An introduction to numerical analysis. New Jersey: John Wiley & Sons; 2008.Suche in Google Scholar

38. Graf, O. Arbeitsphysiologie. Berlin: Springer; 2013.Suche in Google Scholar

39. Grinsztajn, L, Oyallon, E, Varoquaux, G. Why do tree-based models still outperform deep learning on typical tabular data? Adv Neural Inf Process Syst 2022;35:507–20.Suche in Google Scholar

40. Hu, H, Xu, W, Jiang, T, Cheng, Y, Tao, X, Liu, W, et al.. Expert-level immunofixation electrophoresis image recognition based on explainable and generalizable deep learning. Clin Chem 2023;69:130–9. https://doi.org/10.1093/clinchem/hvac190.Suche in Google Scholar PubMed

41. Watson, DS, Krutzinna, J, Bruce, IN, Griffiths, CE, McInnes, IB, Barnes, MR, et al.. Clinical applications of machine learning algorithms: beyond the black box. BMJ 2019;364:l886. https://doi.org/10.1136/bmj.l886.Suche in Google Scholar PubMed

42. Poon, AIF, Sung, JJY. Opening the black box of AI-Medicine. J Gastroenterol Hepatol 2021;36:581–4. https://doi.org/10.1111/jgh.15384.Suche in Google Scholar PubMed

Supplementary Material

This article contains supplementary material (https://doi.org/10.1515/cclm-2024-0222).

Received: 2024-02-16
Accepted: 2024-06-02
Published Online: 2024-06-17
Published in Print: 2024-11-26

© 2024 Walter de Gruyter GmbH, Berlin/Boston

Artikel in diesem Heft

  1. Frontmatter
  2. Editorial
  3. External quality assurance (EQA): navigating between quality and sustainability
  4. Reviews
  5. Molecular allergology: a clinical laboratory tool for precision diagnosis, stratification and follow-up of allergic patients
  6. Nitrous oxide abuse direct measurement for diagnosis and follow-up: update on kinetics and impact on metabolic pathways
  7. Opinion Papers
  8. A vision to the future: value-based laboratory medicine
  9. Point-of-care testing, near-patient testing and patient self-testing: warning points
  10. Navigating the path of reproducibility in microRNA-based biomarker research with ring trials
  11. Point/Counterpoint
  12. Six Sigma – is it time to re-evaluate its value in laboratory medicine?
  13. The value of Sigma-metrics in laboratory medicine
  14. Genetics and Molecular Diagnostics
  15. Analytical validation of the amplification refractory mutation system polymerase chain reaction-capillary electrophoresis assay to diagnose spinal muscular atrophy
  16. Can we identify patients carrying targeted deleterious DPYD variants with plasma uracil and dihydrouracil? A GPCO-RNPGx retrospective analysis
  17. General Clinical Chemistry and Laboratory Medicine
  18. Comparison of ChatGPT, Gemini, and Le Chat with physician interpretations of medical laboratory questions from an online health forum
  19. External quality assessment performance in ten countries: an IFCC global laboratory quality project
  20. Multivariate anomaly detection models enhance identification of errors in routine clinical chemistry testing
  21. Enhanced patient-based real-time quality control using the graph-based anomaly detection
  22. Performance evaluation and user experience of BT-50 transportation unit with automated and scheduled quality control measurements
  23. Stability of steroid hormones in dried blood spots (DBS)
  24. Quantification of C1 inhibitor activity using a chromogenic automated assay: analytical and clinical performances
  25. Reference Values and Biological Variations
  26. Time-dependent characteristics of analytical measurands
  27. Cancer Diagnostics
  28. Expert-level detection of M-proteins in serum protein electrophoresis using machine learning
  29. An automated workflow based on data independent acquisition for practical and high-throughput personalized assay development and minimal residual disease monitoring in multiple myeloma patients
  30. Cardiovascular Diseases
  31. Analytical validation of the Mindray CL1200i analyzer high sensitivity cardiac troponin I assay: MERITnI study
  32. Diabetes
  33. Limitations of glycated albumin standardization when applied to the assessment of diabetes patients
  34. Patient result monitoring of HbA1c shows small seasonal variations and steady decrease over more than 10 years
  35. Letters to the Editor
  36. Inaccurate definition of Bence Jones proteinuria in the EFLM Urinalysis Guideline 2023
  37. Use of the term “Bence-Jones proteinuria” in the EFLM European Urinalysis Guideline 2023
  38. Is uracil enough for effective pre-emptive DPD testing?
  39. Reply to: “Is uracil enough for effective pre-emptive DPD testing?”
  40. Accurate predictory role of monocyte distribution width on short-term outcome in sepsis patients
  41. Reply to: “Accurate predictory role of monocyte distribution width on short-term outcome in sepsis patients”
  42. Spurious parathyroid hormone (PTH) elevation caused by macro-PTH
  43. Setting analytical performance specifications for copeptin-based testing
  44. Serum vitamin B12 levels during chemotherapy against diffuse large B-cell lymphoma: a case report and review of the literature
  45. Evolution of acquired haemoglobin H disease monitored by capillary electrophoresis: a case of a myelofibrotic patient with a novel ATRX mutation
https://doi.org/10.1515/cclm-2024-0222
Schlagwörter für diesen Artikel
artificial intelligence; monoclonal gammopathy; myeloma; electrophoresis; machine learning

Artikel in diesem Heft

  1. Frontmatter
  2. Editorial
  3. External quality assurance (EQA): navigating between quality and sustainability
  4. Reviews
  5. Molecular allergology: a clinical laboratory tool for precision diagnosis, stratification and follow-up of allergic patients
  6. Nitrous oxide abuse direct measurement for diagnosis and follow-up: update on kinetics and impact on metabolic pathways
  7. Opinion Papers
  8. A vision to the future: value-based laboratory medicine
  9. Point-of-care testing, near-patient testing and patient self-testing: warning points
  10. Navigating the path of reproducibility in microRNA-based biomarker research with ring trials
  11. Point/Counterpoint
  12. Six Sigma – is it time to re-evaluate its value in laboratory medicine?
  13. The value of Sigma-metrics in laboratory medicine
  14. Genetics and Molecular Diagnostics
  15. Analytical validation of the amplification refractory mutation system polymerase chain reaction-capillary electrophoresis assay to diagnose spinal muscular atrophy
  16. Can we identify patients carrying targeted deleterious DPYD variants with plasma uracil and dihydrouracil? A GPCO-RNPGx retrospective analysis
  17. General Clinical Chemistry and Laboratory Medicine
  18. Comparison of ChatGPT, Gemini, and Le Chat with physician interpretations of medical laboratory questions from an online health forum
  19. External quality assessment performance in ten countries: an IFCC global laboratory quality project
  20. Multivariate anomaly detection models enhance identification of errors in routine clinical chemistry testing
  21. Enhanced patient-based real-time quality control using the graph-based anomaly detection
  22. Performance evaluation and user experience of BT-50 transportation unit with automated and scheduled quality control measurements
  23. Stability of steroid hormones in dried blood spots (DBS)
  24. Quantification of C1 inhibitor activity using a chromogenic automated assay: analytical and clinical performances
  25. Reference Values and Biological Variations
  26. Time-dependent characteristics of analytical measurands
  27. Cancer Diagnostics
  28. Expert-level detection of M-proteins in serum protein electrophoresis using machine learning
  29. An automated workflow based on data independent acquisition for practical and high-throughput personalized assay development and minimal residual disease monitoring in multiple myeloma patients
  30. Cardiovascular Diseases
  31. Analytical validation of the Mindray CL1200i analyzer high sensitivity cardiac troponin I assay: MERITnI study
  32. Diabetes
  33. Limitations of glycated albumin standardization when applied to the assessment of diabetes patients
  34. Patient result monitoring of HbA1c shows small seasonal variations and steady decrease over more than 10 years
  35. Letters to the Editor
  36. Inaccurate definition of Bence Jones proteinuria in the EFLM Urinalysis Guideline 2023
  37. Use of the term “Bence-Jones proteinuria” in the EFLM European Urinalysis Guideline 2023
  38. Is uracil enough for effective pre-emptive DPD testing?
  39. Reply to: “Is uracil enough for effective pre-emptive DPD testing?”
  40. Accurate predictory role of monocyte distribution width on short-term outcome in sepsis patients
  41. Reply to: “Accurate predictory role of monocyte distribution width on short-term outcome in sepsis patients”
  42. Spurious parathyroid hormone (PTH) elevation caused by macro-PTH
  43. Setting analytical performance specifications for copeptin-based testing
  44. Serum vitamin B12 levels during chemotherapy against diffuse large B-cell lymphoma: a case report and review of the literature
  45. Evolution of acquired haemoglobin H disease monitored by capillary electrophoresis: a case of a myelofibrotic patient with a novel ATRX mutation
Jetzt anmelden und 20% sparen
Institutioneller Zugang
Wie funktioniert Authentifizierung?
Haben Sie eine Idee, wie wir unsere Website verbessern können?
Bitte schreiben Sie uns.
© 2025 De Gruyter Brill
Heruntergeladen am 8.12.2025 von https://www.degruyterbrill.com/document/doi/10.1515/cclm-2024-0222/html?srsltid=AfmBOopfwq9P37mcbHUoMZvFP1NKZrdVzJfe4QIUjoTwbhy5kBigeNVP
Button zum nach oben scrollen