Advertisement

Prediction of outcome after aneurysmal subarachnoid haemorrhage using data from patient admission

  • Christian Rubbert
  • Kaustubh R. Patil
  • Kerim Beseoglu
  • Christian Mathys
  • Rebecca May
  • Marius G. Kaschner
  • Benjamin Sigl
  • Nikolas A. Teichert
  • Johannes Boos
  • Bernd Turowski
  • Julian Caspers
Neuro

Abstract

Objectives

The pathogenesis leading to poor functional outcome after aneurysmal subarachnoid haemorrhage (aSAH) is multifactorial and not fully understood. We evaluated a machine learning approach based on easily determinable clinical and CT perfusion (CTP) features in the course of patient admission to predict the functional outcome 6 months after ictus.

Methods

Out of 630 consecutive subarachnoid haemorrhage patients (2008–2015), 147 (mean age 54.3, 66.7% women) were retrospectively included (Inclusion: aSAH, admission within 24 h of ictus, CTP within 24 h of admission, documented modified Rankin scale (mRS) grades after 6 months. Exclusion: occlusive therapy before first CTP, previous aSAH, CTP not evaluable). A random forests model with conditional inference trees was optimised and trained on sex, age, World Federation of Neurosurgical Societies (WFNS) and modified Fisher grades, aneurysm in anterior vs. posterior circulation, early external ventricular drainage (EVD), as well as MTT and Tmax maximum, mean, standard deviation (SD), range, 75th quartile and interquartile range to predict dichotomised mRS (≤ 2; > 2). Performance was assessed using the balanced accuracy over the training and validation folds using 20 repeats of 10-fold cross-validation.

Results

In the final model, using 200 trees and the synthetic minority oversampling technique, median balanced accuracy was 84.4% (SD 0.7) over the training folds and 70.9% (SD 1.2) over the validation folds. The five most important features were the modified Fisher grade, age, MTT range, WFNS and early EVD.

Conclusions

A random forests model trained on easily determinable features in the course of patient admission can predict the functional outcome 6 months after aSAH with considerable accuracy.

Key Points

• Features determinable in the course of admission of a patient with aneurysmal subarachnoid haemorrhage (aSAH) can predict the functional outcome 6 months after the occurrence of aSAH.

• The top five predictive features were the modified Fisher grade, age, the mean transit time (MTT) range from computed tomography perfusion (CTP), the WFNS grade and the early necessity for an external ventricular drainage (EVD).

• The range between the minimum and the maximum MTT may prove to be a valuable biomarker for detrimental functional outcome.

Keywords

Subarachnoid haemorrhage Aneurysm Multidetector computed tomography Critical care outcomes Machine learning 

Abbreviations and acronyms

aSAH

Aneurysmal subarachnoid haemorrhage

CTP

CT perfusion

DCI

Delayed cerebral ischaemia

EVD

External ventricular drainage

GCS

Glasgow coma scale

mRS

Modified Rankin scale

MTT

Mean transit time

SD

Standard deviation

SMOTE

Synthetic minority oversampling technique

WFNS

World Federation of Neurosurgical Societies

Notes

Acknowledgements

Editing assistance of an earlier version of the manuscript: Bonnie Hami, M.A. (Cleveland, OH, USA).

Funding

The authors state that this work has not received any funding.

Compliance with ethical standards

Guarantor

The scientific guarantor of this publication is Christian Rubbert.

Conflict of interest

The authors of this manuscript declare no relationships with any companies whose products or services may be related to the subject matter of the article.

Statistics and biometry

One of the authors has significant statistical expertise.

Informed consent

Written informed consent was waived by the institutional review board.

Ethical approval

Institutional review board approval was obtained.

Methodology

• retrospective

• cross-sectional study

• performed at one institution

References

  1. 1.
    Feigin VL, Lawes CMM, Bennett DA, Anderson CS (2003) Stroke epidemiology: a review of population-based studies of incidence, prevalence, and case-fatality in the late 20th century. Lancet Neurol 2:43–53CrossRefGoogle Scholar
  2. 2.
    van Gijn J, Rinkel GJE (2001) Subarachnoid haemorrhage: diagnosis, causes and management. Brain 124:249–278CrossRefGoogle Scholar
  3. 3.
    Rinkel GJE, Algra A (2011) Long-term outcomes of patients with aneurysmal subarachnoid haemorrhage. Lancet Neurol 10:349–356CrossRefGoogle Scholar
  4. 4.
    Fujii M, Yan J, Rolland WB et al (2013) Early brain injury, an evolving frontier in subarachnoid haemorrhage research. Transl Stroke Res 4:432–446Google Scholar
  5. 5.
    Dreier JP (2011) The role of spreading depression, spreading depolarisation and spreading ischemia in neurological disease. Nat Med 17:439–447Google Scholar
  6. 6.
    Macdonald RL (2014) Delayed neurological deterioration after subarachnoid haemorrhage. Nat Rev Neurol 10:44–58CrossRefGoogle Scholar
  7. 7.
    Østergaard L, Aamand R, Karabegovic S et al (2013) The role of the microcirculation in delayed cerebral ischemia and chronic degenerative changes after subarachnoid hemorrhage. J Cereb Blood Flow Metab 33:1825–1837CrossRefGoogle Scholar
  8. 8.
    Rosen DS, Macdonald RL (2005) Subarachnoid hemorrhage grading scales: a systematic review. Neurocrit Care 2:110–118CrossRefGoogle Scholar
  9. 9.
    Takagi K, Tamura A, Nakagomi T et al (1999) How should a subarachnoid hemorrhage grading scale be determined? A combinatorial approach based solely on the Glasgow coma scale. J Neurosurg 90:680–687CrossRefGoogle Scholar
  10. 10.
    Fisher CM, Kistler JP, Davis JM (1980) Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery 6:1–9CrossRefGoogle Scholar
  11. 11.
    Frontera JA, Claassen J, Schmidt JM et al (2006) Prediction of symptomatic vasospasm after subarachnoid hemorrhage: the modified Fisher scale. Neurosurgery 59:21–27 discussion 21–7CrossRefGoogle Scholar
  12. 12.
    Etminan N, Beseoglu K, Heiroth H-J et al (2013) Early perfusion computerized tomography imaging as a radiographic surrogate for delayed cerebral ischemia and functional outcome after subarachnoid hemorrhage. Stroke 44:1260–1266CrossRefGoogle Scholar
  13. 13.
    Mathys C, Martens D, Reichelt DC et al (2013) Long-term impact of perfusion CT data after subarachnoid hemorrhage. Neuroradiology 55:1323–1331CrossRefGoogle Scholar
  14. 14.
    Caspers J, Rubbert C, Turowski B et al (2015) Timing of mean transit time maximization is associated with neurological outcome after subarachnoid hemorrhage. Clin Neuroradiol 27:15–22CrossRefGoogle Scholar
  15. 15.
    Diringer MN, Bleck TP, Claude Hemphill J 3rd et al (2011) Critical care management of patients following aneurysmal subarachnoid hemorrhage: recommendations from the Neurocritical Care Society's Multidisciplinary Consensus Conference. Neurocrit Care 15:211–240CrossRefGoogle Scholar
  16. 16.
    Jaja BNR, Cusimano MD, Etminan N et al (2013) Clinical prediction models for aneurysmal subarachnoid hemorrhage: a systematic review. Neurocrit Care 18:143–153CrossRefGoogle Scholar
  17. 17.
    Turowski B, Schramm P (2015) An appeal to standardize CT- and MR-perfusion. Clin Neuroradiol 25:205–210CrossRefGoogle Scholar
  18. 18.
    Wittsack H-J, Wohlschläger AM, Ritzl EK et al (2008) CT-perfusion imaging of the human brain: advanced deconvolution analysis using circulant singular value decomposition. Comput Med Imaging Graph 32:67–77CrossRefGoogle Scholar
  19. 19.
    Breiman L (2001) Random forests. Machine Learning 45:5–32CrossRefGoogle Scholar
  20. 20.
    Hothorn T, Hornik K, Zeileis A (2006) Unbiased recursive partitioning: a conditional inference framework. J Comput Graph Stat 15:651–674CrossRefGoogle Scholar
  21. 21.
    Farrell B, Godwin J, Richards S, Warlow C (1991) The United Kingdom transient ischaemic attack (UK-TIA) aspirin trial: final results. J Neurol Neurosurg Psychiatry 54:1044–1054CrossRefGoogle Scholar
  22. 22.
    Cremers CHP, Vos PC, van der Schaaf IC et al (2015) CT perfusion during delayed cerebral ischemia after subarachnoid hemorrhage: distinction between reversible ischemia and ischemia progressing to infarction. Neuroradiology 57:897–902CrossRefGoogle Scholar
  23. 23.
    Calamante F, Christensen S, Desmond PM et al (2010) The physiological significance of the time-to-maximum (Tmax) parameter in perfusion MRI. Stroke 41:1169–1174CrossRefGoogle Scholar
  24. 24.
    Molinaro AM, Simon R, Pfeiffer RM (2005) Prediction error estimation: a comparison of resampling methods. Bioinformatics 21:3301–3307CrossRefGoogle Scholar
  25. 25.
    Velez DR, White BC, Motsinger AA et al (2007) A balanced accuracy function for epistasis modeling in imbalanced datasets using multifactor dimensionality reduction. Genet Epidemiol 31:306–315CrossRefGoogle Scholar
  26. 26.
    Chawla NV, Bowyer KW, Hall LO, Kegelmeyer WP (2002) SMOTE: synthetic minority over-sampling technique. J Artif Intell Res 16:321–357Google Scholar
  27. 27.
    Strobl C, Boulesteix A-L, Zeileis A, Hothorn T (2007) Bias in random forest variable importance measures: illustrations, sources and a solution. BMC Bioinformatics 8:25CrossRefGoogle Scholar
  28. 28.
    Strobl C, Boulesteix A-L, Kneib T et al (2008) Conditional variable importance for random forests. BMC Bioinformatics 9:307CrossRefGoogle Scholar
  29. 29.
    de Toledo P, Rios PM, Ledezma A et al (2009) Predicting the outcome of patients with subarachnoid hemorrhage using machine learning techniques. IEEE Trans Inf Technol Biomed 13:794–801CrossRefGoogle Scholar
  30. 30.
    Rosengart AJ, Schultheiss KE, Tolentino J, Macdonald RL (2007) Prognostic factors for outcome in patients with aneurysmal subarachnoid hemorrhage. Stroke 38:2315–2321CrossRefGoogle Scholar
  31. 31.
    Ransom ER, Mocco J, Komotar RJ et al (2007) External ventricular drainage response in poor grade aneurysmal subarachnoid hemorrhage: effect on preoperative grading and prognosis. Neurocrit Care 6:174–180CrossRefGoogle Scholar
  32. 32.
    Rodriguez-Régent C, Hafsa M, Turc G et al (2016) Early quantitative CT perfusion parameters variation for prediction of delayed cerebral ischemia following aneurysmal subarachnoid hemorrhage. Eur Radiol 26:2956–2963CrossRefGoogle Scholar
  33. 33.
    Murphy A, de Manoel AL O, Burgers K et al (2015) Early CT perfusion changes and blood-brain barrier permeability after aneurysmal subarachnoid hemorrhage. Neuroradiology 57:767–773CrossRefGoogle Scholar
  34. 34.
    Cremers CHP, van der Schaaf IC, Wensink E et al (2014) CT perfusion and delayed cerebral ischemia in aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis. J Cereb Blood Flow Metab 34:200–207CrossRefGoogle Scholar
  35. 35.
    Malinova V, Dolatowski K, Schramm P et al (2016) Early whole-brain CT perfusion for detection of patients at risk for delayed cerebral ischemia after subarachnoid hemorrhage. J Neurosurg 125:128–136CrossRefGoogle Scholar
  36. 36.
    Othman AE, Brockmann C, Yang Z et al (2015) Effects of radiation dose reduction in volume perfusion CT imaging of acute ischemic stroke. Eur Radiol 25:3415–3422CrossRefGoogle Scholar
  37. 37.
    Othman AE, Afat S, Brockmann C et al (2017) Low-dose volume-perfusion CT of the brain: effects of radiation dose reduction on performance of perfusion CT algorithms. Clin Neuroradiol 27:311–318CrossRefGoogle Scholar
  38. 38.
    Othman AE, Afat S, Brockmann MA et al (2016) Radiation dose reduction in perfusion CT imaging of the brain: a review of the literature. J Neuroradiol 43:1–5CrossRefGoogle Scholar
  39. 39.
    Othman AE, Afat S, Nikoubashman O et al (2016) Volume perfusion CT imaging of cerebral vasospasm: diagnostic performance of different perfusion maps. Neuroradiology 58:787–792CrossRefGoogle Scholar
  40. 40.
    Zijlstra IA, Gathier CS, Boers AM et al (2016) Association of automatically quantified total blood volume after aneurysmal subarachnoid hemorrhage with delayed cerebral ischemia. AJNR Am J Neuroradiol 37:1588–1593CrossRefGoogle Scholar
  41. 41.
    Boers AM, Zijlstra IA, Gathier CS et al (2014) Automatic quantification of subarachnoid hemorrhage on noncontrast CT. AJNR Am J Neuroradiol 35:2279–2286CrossRefGoogle Scholar

Copyright information

© European Society of Radiology 2018

Authors and Affiliations

  • Christian Rubbert
    • 1
  • Kaustubh R. Patil
    • 2
    • 3
  • Kerim Beseoglu
    • 4
  • Christian Mathys
    • 1
    • 5
  • Rebecca May
    • 1
  • Marius G. Kaschner
    • 1
  • Benjamin Sigl
    • 1
  • Nikolas A. Teichert
    • 1
  • Johannes Boos
    • 1
  • Bernd Turowski
    • 1
  • Julian Caspers
    • 1
    • 6
  1. 1.University Düsseldorf, Medical FacultyDepartment of Diagnostic and Interventional RadiologyDüsseldorfGermany
  2. 2.Institute of Neuroscience and Medicine, Brain & Behaviour (INM-7)Research Centre JülichJülichGermany
  3. 3.Institute of Systems Neuroscience, Medical FacultyHeinrich Heine University DüsseldorfDüsseldorfGermany
  4. 4.Department of Neurosurgery, Medical FacultyHeinrich-Heine-UniversityDüsseldorfGermany
  5. 5.Institute of Radiology and Neuroradiology, Evangelisches KrankenhausUniversity of OldenburgOldenburgGermany
  6. 6.Institute of Neuroscience and Medicine (INM-1)Research Centre JülichJülichGermany

Personalised recommendations