Advertisement

Regression, Classification and Ensemble Machine Learning Approaches to Forecasting Clinical Outcomes in Ischemic Stroke

  • Ahmedul Kabir
  • Carolina RuizEmail author
  • Sergio A. Alvarez
  • Majaz Moonis
Conference paper
Part of the Communications in Computer and Information Science book series (CCIS, volume 881)

Abstract

We applied different machine learning approaches to predict (forecast) the clinical outcome, measured by the modified Rankin Scale (mRS) score, of ischemic stroke patients 90 days after stroke. Regression, multinomial classification, and ordinal regression tasks were considered. M5 model trees followed by bootstrap aggregating as a meta-learning technique produced the best regression results. The same regression technique when used for classification after discretization of the target attribute also performed better than regular multinomial classification. For the ordinal regression task, the logit link function (ordinal logistic regression) outperformed the alternatives. We discuss the methodology used, and compare the results with other standard predictive techniques. We also analyze the results to provide insights into the factors that affect stroke outcomes.

Keywords

Ischemic stroke mRS score M5 model tree Bootstrap aggregating Ordinal regression 

Notes

Acknowledgements

The authors thank Prof. Dr. Klaus Brinker for suggesting using ordinal regression as an additional technique in this research.

References

  1. 1.
    Raffeld, M., Debette, S., Woo, D.: International stroke genetics consortium update. Stroke 47(4), 1144–1145 (2016).  https://doi.org/10.1161/STROKEAHA.116.012682CrossRefGoogle Scholar
  2. 2.
    Mozaffarian, D., Benjamin, E.J., Go, A.S., et al.: Heart disease and stroke statistics—2016 update. Circulation 133(4), e38–e360 (2016).  https://doi.org/10.1161/CIR.0000000000000350CrossRefGoogle Scholar
  3. 3.
    Rankin, J.: Cerebral vascular accidents in patients over the age of 60: II. Prognosis. Scott. Med. J. 2, 200–215 (1957)CrossRefGoogle Scholar
  4. 4.
    Kabir, A., Ruiz, C., Alvarez, S.A., Moonis, M.: Predicting outcome of ischemic stroke patients using bootstrap aggregating with M5 Model trees. In: van den Broek, E.L., Fred, A., Gamboa, H., Vaz, M. (eds.) 10th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2017, Proceedings Volume 5: HealthInf, 21–23 February, Porto, Protugal, pp. 178–187 (2017)Google Scholar
  5. 5.
    van Swieten, J.C., Koudstaal, P.J., Visser, M.C., Schouten, H.J., van Gijn, J.: Interobserver agreement for the assessment of handicap in stroke patients. Stroke 19(5), 604–607 (1988)CrossRefGoogle Scholar
  6. 6.
    Banks, J.L., Marotta, C.A.: Outcomes validity and reliability of the modified rankin scale: implications for stroke clinical trials: a literature review and synthesis. Stroke 38(3), 1091–1096 (2007)CrossRefGoogle Scholar
  7. 7.
    Bruno, A., Shah, N., Lin, C., Close, B., Hess, D.C., Davis, K., Baute, V., Switzer, J.A., Waller, J., Nichols, F.T.: Improving modified rankin scale assessment with a simplified questionnaire. Stroke 41(5), 1048–1050 (2010).  https://doi.org/10.1161/STROKEAHA.109.571562CrossRefGoogle Scholar
  8. 8.
    Tan, P.-N., Steinbach, M., Kumar, V.: Introduction to Data Mining. Addison-Wesley, Boston (2005)Google Scholar
  9. 9.
    Breiman, L., Friedman, J.H., Olshen, R.A., Stone, C.J.: Classification and Regression Trees. Wadsworth, Belmont (1984)zbMATHGoogle Scholar
  10. 10.
    Quinlan, J.R.: Learning with continuous classes. In: Adams, A., Sterling, L. (eds.) 5th Australian Joint Conference on Artificial Intelligence, Hobart, Australia, pp. 343–348 (1992)Google Scholar
  11. 11.
    Wang, Y., Witten, I.H.: Induction of model trees for predicting continuous classes. In: van Someren, M., Widmer, G. (eds.) 9th European Conference on Machine Learning. University of Economics, Prague (1996)Google Scholar
  12. 12.
    Etemad-Shahidi, A., Mahjoobi, J.: Comparison between M5′ model tree and neural networks for prediction of significant wave height in Lake Superior. Ocean Eng. 36(15–16), 1175–1181 (2009).  https://doi.org/10.1016/j.oceaneng.2009.08.008CrossRefGoogle Scholar
  13. 13.
    Breiman, L.: Bagging predictors. Mach. Learn. 24(2), 123–140 (1996)MathSciNetzbMATHGoogle Scholar
  14. 14.
    Aslam, J., Popa, R., Rivest, R.: On estimating the size and confidence of a statistical audit. USENIX/ACCURATE Electron. Voting Technol. Workshop 7, 8 (2007)Google Scholar
  15. 15.
    McCullagh, P.: Regression models for ordinal data. J. R. Stat. Soc.: Ser. B (Methodol.) 42(2), 109–142 (1980)MathSciNetzbMATHGoogle Scholar
  16. 16.
    Herbrich, R., Graepel, T., Obermayer, K.: Regression models for ordinal data: a machine learning approach. Technische Universität Berlin, Fachbereich 13, Informatik, Berlin (1999)Google Scholar
  17. 17.
    Anderson, J.: Regression and ordered categorical variables. J. R. Stat. Soc.: Ser. B (Methodol.) 46(1), 1–30 (1984)MathSciNetzbMATHGoogle Scholar
  18. 18.
    Agresti, A.: An Introduction to Categorical Data Analysis. Wiley-Interscience, Hoboken (2007)CrossRefGoogle Scholar
  19. 19.
    Bender, R., Grouven, U.: Ordinal logistic regression in medical research. J. R. Coll. Phys. Lond. 31(5), 546–551 (1997)Google Scholar
  20. 20.
    Moonis, M., Kane, K., Schwiderski, U., Sandage, B.W., Fisher, M.: HMG-CoA reductase inhibitors improve acute ischemic stroke outcome. Stroke 36(6), 1298–1300 (2005)CrossRefGoogle Scholar
  21. 21.
    Marini, C., De Santis, F., Sacco, S., Russo, T., Olivieri, L., Totaro, R., Carolei, A.: Contribution of atrial fibrillation to incidence and outcome of ischemic stroke. Stroke 36(6), 1115–1119 (2005)CrossRefGoogle Scholar
  22. 22.
    Yong, M., Kaste, M.: Dynamic of hyperglycemia as a predictor of stroke outcome in the ECASS-II trial. Stroke 39(10), 2749–2755 (2008).  https://doi.org/10.1161/STROKEAHA.108.514307CrossRefGoogle Scholar
  23. 23.
    Nogueira, R.G., Liebeskind, D.S., Sung, G., Duckwiler, G., Smith, W.S.: Multi MERCI writing committee: predictors of good clinical outcomes, mortality, and successful revascularization in patients with acute ischemic stroke undergoing thrombectomy: pooled analysis of the mechanical embolus removal in cerebral ischemia (MERCI) and multi MERCI trials. Stroke 40(12), 3777–3783 (2009).  https://doi.org/10.1161/STROKEAHA.109.561431CrossRefGoogle Scholar
  24. 24.
    Henninger, N., Lin, E., Baker, S.P., Wakhloo, A.K., Takhtani, D., Moonis, M.: Leukoaraiosis predicts poor 90-Day outcome after acute large cerebral artery occlusion. Cerebrovascular Diseases. 33(6), 525–531 (2012).  https://doi.org/10.1159/000337335CrossRefGoogle Scholar
  25. 25.
    Kissela, B., Lindsell, C.J., Kleindorfer, D., Alwell, K., Moomaw, C.J., Woo, D., Flaherty, M.L., Air, E., Broderick, J., Tsevat, J.: Clinical prediction of functional outcome after ischemic stroke: the surprising importance of periventricular white matter disease and race. Stroke 40(2), 530–536 (2008).  https://doi.org/10.1161/STROKEAHA.108.521906CrossRefGoogle Scholar
  26. 26.
    Optimising Analysis of Stroke Trials Collaboration: Calculation of sample size for stroke trials assessing functional outcome: comparison of binary and ordinal approaches. Int. J. Stroke 3(2), 78–84 (2008).  https://doi.org/10.1111/j.1747-4949.2008.00184.x
  27. 27.
    Sandset, E.C., Bath, P.M., Boysen, G., Jatuzis, D., Kõrv, J., Lüders, S., Murray, G.D., Richter, P.S., Roine, R.O., Terént, A., Thijs, V., Berge, E.: SCAST study group: the angiotensin-receptor blocker candesartan for treatment of acute stroke (SCAST): a randomised, placebo-controlled, double-blind trial. Lancet 377(9767), 741–750 (2011).  https://doi.org/10.1016/S0140-6736(11)60104-9CrossRefGoogle Scholar
  28. 28.
    Gialanella, B., Santoro, R., Ferlucci, C.: Predicting outcome after stroke: the role of basic activities of daily living. Eur. J. Phys. Rehabil. Med. 49(5), 629–637 (2013)Google Scholar
  29. 29.
    Granger, C.V., Hamilton, B.B., Keith, R.A., Zielezny, M., Sherwin, F.S.: Advances in functional assessment for medical rehabilitation. Top. Geriatr. Rehabil. 1(3), 59–74 (1986)CrossRefGoogle Scholar
  30. 30.
    Brown, A.W., Therneau, T.M., Schultz, B.A., Niewczyk, P.M., Granger, C.V.: Measure of functional independence dominates discharge outcome prediction after inpatient rehabilitation for stroke. Stroke 46(4), 1038–1044 (2015).  https://doi.org/10.1161/STROKEAHA.114.007392CrossRefGoogle Scholar
  31. 31.
    Adams Jr., H.P., Bendixen, B.H., Kappelle, L.J., Biller, J., Love, B.B., Gordon, D.L., Marsh 3rd, E.E.: Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 24(1), 35–41 (1993)CrossRefGoogle Scholar
  32. 32.
    Brott, T., Adams Jr., H.P., Olinger, C.P., Marler, J.R., Barsan, W.G., Biller, J., Spilker, J., Holleran, R., Eberle, R., Hertzberg, V., et al.: Measurements of acute cerebral infarction: a clinical examination scale. Stroke 20(7), 864–870 (1989)CrossRefGoogle Scholar
  33. 33.
    Nakayama, H., Jorgensen, H.S., Raaschou, H.O., Olsen, T.S.: The influence of age on stroke outcome. The Copenhagen Stroke Study. Stroke 25(4), 808–813 (1994)CrossRefGoogle Scholar
  34. 34.
    Frank, E., Hall, M.A., Witten, I.H.: The WEKA Workbench. Online Appendix for “Data Mining: Practical Machine Learning Tools and Techniques”. Morgan Kaufmann, Burlington (2016)Google Scholar
  35. 35.
    Kohavi, R.: A study of cross-validation and bootstrap for accuracy estimation and model selection. In: Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI), vol. 2, pp. 1137–1145 (1995)Google Scholar
  36. 36.
    Lee Rodgers, J., Nicewander, W.A.: Thirteen ways to look at the correlation coefficient. Am. Stat. 42(1), 59–66 (1988)CrossRefGoogle Scholar
  37. 37.
    Moore, D.S., Notz, W.I., Fligner, M.A.: The Basic Practice of Statistics. W.H. Freeman and Co., New York (2015)Google Scholar
  38. 38.
    Quinlan, J.R.: C4.5 - Programs for Machine Learning. Morgan Kaufmann, San Francisco (1992)Google Scholar
  39. 39.
    Christensen, R.H.B.: Regression Models for Ordinal Data [R package ordinal version 2015.6-28]. https://cran.r-project.org/web/packages/ordinal/index.html
  40. 40.
    Bozdogan, H.: Model selection and Akaike’s Information Criterion (AIC): the general theory and its analytical extensions. Psychometrika 52(3), 345–370 (1987)MathSciNetCrossRefGoogle Scholar
  41. 41.
    Willmott, C.J., Matsuura, K.: Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance. Clim. Res. 30(1), 79–82 (2005).  https://doi.org/10.3354/cr030079CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  • Ahmedul Kabir
    • 1
  • Carolina Ruiz
    • 1
    Email author
  • Sergio A. Alvarez
    • 2
  • Majaz Moonis
    • 3
  1. 1.Worcester Polytechnic InstituteWorcesterUSA
  2. 2.Boston CollegeChestnut HillUSA
  3. 3.University Massachusetts Medical SchoolWorcesterUSA

Personalised recommendations