Machine learning algorithm to predict in-hospital mortality after aneurysmal subarachnoid hemorrhage

Machine learning (ML) methodology surpasses the traditional tools of statistical analysis in processing big data clinical datasets .

Aim. To develop an ML algorithm of application of recurrent neural network to analyze clinical datasets of patients with aneurysmal subarachnoid hemorrhage (SAH).

Materials and methods. A big data registry included retrospective data from 2,631 patients with an arterial aneurysm. From these, 390 individuals were selected who required treatment for SAH in an intensive care unit (ICU) setting. The raw dataset contained 7290 features, from which 12 features were selected to train the following ML models: logistic regression, support vector machine, random forest, XGBoost, multilayer perceptron and long short-term memory network (LSTM) were tested. Data preprocessing and modeling were provided in Python (version 3.11.4) using scikitlearn, tensorfl ow, keras and hyperopt libraries. The values and 95% confi dence intervals (CI) of AUROC and AURPC, predictive value, specifi city and sensitivity were calculated.

Results. We recruited 246 (63%) females and 144 (37%) males with mean age of 54±12.9 years. Death occurred in 133 (34%) patients including 33 patients deceased during 24 hours after admission. The best model for predicting lethal outcome was LSTM. After comparison with other ML algorithms LSTM showed the highest predictive values (AUROC – 0.83; 95% CI: 0.72–0.92, AURPC – 0.62; 95% CI 0.39–0.81) in term of in-hospital mortality. For the period in ICU from day 3 to day 6, the model’s positive predictive value was 0.83, sensitivity 0.95 and specifi city 0.58.

Conclusions. LSTM may be applied to development of automatic algorithms in management of critically ill patients after SAH.

1. van Gijn J, Kerr R.S., Rinkel G.J. Subarachnoid haemorrhage. Lancet. 2007; 369(9558): 306–318. https://doi.org/10.1016/s0140-6736(07)60153-6. PMID: 17258671

2. Suarez J.I., Tarr R.W., Selman W.R. Aneurysmal subarachnoid hemorrhage. N Engl J Med. 2006; 354(4): 387–396. https://doi.org/10.1056/nejmra052732. PMID: 16436770

3. Lawton M.T., Vates G.E. Subarachnoid Hemorrhage. The New England journal of medicine. 2017; 377(3): 257–266. https://doi.org/10.1056/nejmcp1605827. PMID: 28723321

4. Brisman J.L., Song J.K., Newell D.W. Cerebral aneurysms. N Engl J Med. 2006; 355(9): 928–939. https://doi.org/10.1056/nejmra052760. PMID: 16943405

5. Wang R., Zhang J., Shan B., et al. XGBoost Machine Learning Algorithm for Prediction of Outcome in Aneurysmal Subarachnoid Hemorrhage. Neuropsychiatr Dis Treat. 2022; 18: 659–667. https://doi.org/10.2147/ndt.s349956. PMID: 35378822

6. Jaja B.N.R., Saposnik G., Lingsma H.F., et al. Development and validation of outcome prediction models for aneurysmal subarachnoid haemorrhage: the SAHIT multinational cohort study. Bmj. 2018; 360: j5745. https://doi.org/10.1136/bmj.j5745. PMID: 29348138

7. Dengler N.F., Madai V.I., Unteroberdörster M., et al. Outcome prediction in aneurysmal subarachnoid hemorrhage: a comparison of machine learning methods and established clinico-radiological scores. Neurosurg Rev. 2021; 44(5): 2837–2846. https://doi.org/10.1007/s10143-020-01453-6. PMID: 33474607

8. Yu D., Williams G.W., Aguilar D., et al. Machine learning prediction of the adverse outcome for nontraumatic subarachnoid hemorrhage patients. Ann Clin Transl Neurol. 2020; 7(11): 2178– 2185. https://doi.org/10.1002/acn3.51208. Epub 2020 Sep 29. PMID: 32990362

9. Tabaie A., Nemati S., Allen J.W., et al. Assessing contribution of higher order clinical risk factors to prediction of outcome in aneurysmal subarachnoid hemorrhage patients. AMIA Annu Symp Proc. 2019; 2019: 848–856. PMID: 32308881

10. Štrumbelj E., Kononenko I. Explaining prediction models and individual predictions with feature contributions. Knowledge and Information Systems. 2013; 41: 647–665. https://doi.org/10.1007/s10115-013-0679-x

11. Oh C.H., Kim J.W., Kim G.H., et al. Serum Lactate could predict mortality in patients with spontaneous subarachnoid hemorrhage in the emergency department. Front Neurol. 2020; 11: 975. https://doi.org/10.3389/fneur.2020.00975. PMID: 33013645

12. Kissoon N.R., Mandrekar J.N., Fugate J.E., et al. Positive fluid balance is associated with poor outcomes in subarachnoid hemorrhage. J Stroke Cerebrovasc Dis. 2015; 24(10): 2245–2251. https://doi.org/10.1016/j.jstrokecerebrovasdis.2015.05.027. PMID: 26277290

13. Martini R.P., Deem S., Brown M., et al. The association between fluid balance and outcomes after subarachnoid hemorrhage. Neurocrit Care. 2012; 17(2): 191–198. https://doi.org/10.1007/s12028-011-9573-0. PMID: 21688008

14. Deem S., Diringer M., Livesay S., Treggiari M.M. Hemodynamic management in the prevention and treatment of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage. Neurocrit Care. 2023; 39(1): 81–90. https://doi.org/10.1007/s12028-02301738-w. PMID: 37160848

15. Hosmann A., Schnackenburg P., Rauscher S., et al. Brain tissue oxygen response as indicator for cerebral lactate levels in aneurysmal subarachnoid hemorrhage patients. J Neurosurg Anesthesiol. 2022; 34(2): 193–200. https://doi.org/10.1097/ANA.0000000000000713 PMID: 32701532

16. Treggiari M.M., Deem S. Which H. is the most important in triple-H therapy for cerebral vasospasm? Curr Opin Crit Care. 2009; 15(2): 83–86. https://doi.org/10.1097/mcc.0b013e32832922d1. PMID: 19276798

17. Sen J., Belli A., Albon H., et al. Triple-H therapy in the management of aneurysmal subarachnoid haemorrhage. Lancet Neurol. 2003; 2(10): 614–621. https://doi.org/10.1016/s14744422(03)00531-3. PMID: 14505583

18. Solenski N.J., Haley E.C. Jr., Kassell N.F., et al. Medical complications of aneurysmal subarachnoid hemorrhage: a report of the multicenter, cooperative aneurysm study. Participants of the Multicenter Cooperative Aneurysm Study. Crit Care Med. 1995; 23(6): 1007–1017. https://doi.org/10.1097/00003246-19950600000004. PMID: 7774210

19. Festic E., Rabinstein A.A., Freeman W.D., et al. Blood transfusion is an important predictor of hospital mortality among patients with aneurysmal subarachnoid hemorrhage. Neurocrit Care. 2013; 18(2): 209–215. https://doi.org/10.1007/s12028-0129777-y PMID: 22965325

20. Bakker J., Nijsten M.W., Jansen T.C. Clinical use of lactate monitoring in critically ill patients. Ann Intensive Care. 2013; 3(1): 12. https://doi.org/10.1186/2110-5820-3-12. PMID: 23663301

21. Krishna U., Joshi S.P., Modh M. An evaluation of serial blood lactate measurement as an early predictor of shock and its outcome in patients of trauma or sepsis. Indian J Crit Care Med. 2009; 13(2): 66–73. https://doi.org/10.4103/0972-5229.56051. PMID: 19881186

22. Ahn S.H., Savarraj J.P., Pervez M., et al. The subarachnoid hemorrhage early brain edema score predicts delayed cerebral ischemia and clinical outcomes. Neurosurgery. 2018; 83(1): 137–145. https://doi.org/10.1093/neuros/nyx364. PMID: 28973675

23. Suzuki H. What is early brain injury? Transl Stroke Res. 2015; 6(1): 1–3. https://doi.org/10.1007/s12975-014-0380-8. PMID: 25502277

24. Fujii M., Yan J., Rolland W.B., et al. Early brain injury, an evolving frontier in subarachnoid hemorrhage research. Transl Stroke Res. 2013; 4(4): 432–446. https://doi.org/10.1007/s12975-0130257-2. PMID: 23894255

25. Savarraj J., Parsha K., Hergenroeder G., et al. Early brain injury associated with systemic inflammation after subarachnoid hemorrhage. Neurocrit Care. 2018; 28(2): 203–211. https://doi.org/10.1007/s12028-017-0471-y. PMID: 29043545

26. Eibach M., Won S.Y., Bruder M., et al. Age dependency and modification of the Subarachnoid Hemorrhage Early Brain Edema Score. J Neurosurg. 2020; 134(3): 946–952. https://doi.org/10.3171/2019.12.jns192744. PMID: 32197254