Prediction of stroke thrombolysis outcome using CT brain machine learning - PubMed (original) (raw)

Prediction of stroke thrombolysis outcome using CT brain machine learning

Paul Bentley et al. Neuroimage Clin. 2014.

Abstract

A critical decision-step in the emergency treatment of ischemic stroke is whether or not to administer thrombolysis - a treatment that can result in good recovery, or deterioration due to symptomatic intracranial haemorrhage (SICH). Certain imaging features based upon early computerized tomography (CT), in combination with clinical variables, have been found to predict SICH, albeit with modest accuracy. In this proof-of-concept study, we determine whether machine learning of CT images can predict which patients receiving tPA will develop SICH as opposed to showing clinical improvement with no haemorrhage. Clinical records and CT brains of 116 acute ischemic stroke patients treated with intravenous thrombolysis were collected retrospectively (including 16 who developed SICH). The sample was split into training (n = 106) and test sets (n = 10), repeatedly for 1760 different combinations. CT brain images acted as inputs into a support vector machine (SVM), along with clinical severity. Performance of the SVM was compared with established prognostication tools (SEDAN and HAT scores; original, or after adaptation to our cohort). Predictive performance, assessed as area under receiver-operating-characteristic curve (AUC), of the SVM (0.744) compared favourably with that of prognostic scores (original and adapted versions: 0.626-0.720; p < 0.01). The SVM also identified 9 out of 16 SICHs, as opposed to 1-5 using prognostic scores, assuming a 10% SICH frequency (p < 0.001). In summary, machine learning methods applied to acute stroke CT images offer automation, and potentially improved performance, for prediction of SICH following thrombolysis. Larger-scale cohorts, and incorporation of advanced imaging, should be tested with such methods.

Keywords: Imaging; Machine learning; Prediction; Stroke; Thrombolysis.

PubMed Disclaimer

Figures

Fig. 1

Fig. 1

Example of CT normalisation pipeline in one subject. Source images were acquired in top and bottom sections (A) that were normalised respectively to whole-brain and bottom CT templates (B). The two normalised images (C) were joined (D) whereby voxels that were sampled in both images were averaged, and some voxels were sampled in neither (seen as a black ‘join’ anteriorly). The resultant images were inclusively masked by a brain template, but sometimes this included patients' cranium (revealed as anomalous Hounsfield Unit > 200): E shows where this occurred in ≥ 5 subjects, these voxels then being excluded. Most patients also showed a thin non-sampled join, although the location of this differed slightly across subjects. These and other anomalous voxels that occurred in < 5 subjects were replaced by their mean from the remaining set of subjects, and are identified in F.

Similar articles

Cited by

References

    1. Campbell B.C., Christensen S., Parsons M.W., Churilov L., Desmond P.M., Barber P.A., Butcher K.S., Levi C.R., De Silva D.A., Lansberg M.G., Mlynash M., Olivot J.M., Straka M., Bammer R., Albers G.W., Donnan G.A., Davis S.M., Investigators E.a.D. Advanced imaging improves prediction of hemorrhage after stroke thrombolysis. Ann. Neurol. 2012;73(4):510–519. - PMC - PubMed
    1. Charidimou A., Kakar P., Fox Z., Werring D.J. Cerebral microbleeds and the risk of intracerebral haemorrhage after thrombolysis for acute ischaemic stroke: systematic review and meta-analysis. J. Neurol. Neurosurg. Psychiatry. 2013;84:277–280. - PMC - PubMed
    1. Dharmasaroja P., Dharmasaroja P.A. Prediction of intracerebral hemorrhage following thrombolytic therapy for acute ischemic stroke using multiple artificial neural networks. Neurol. Res. 2012;34:120–128. - PubMed
    1. Dubey N., Bakshi R., Wasay M., Dmochowski J. Early computed tomography hypodensity predicts hemorrhage after intravenous tissue plasminogen activator in acute ischemic stroke. J. Neuroimaging. 2001;11:184–188. - PubMed
    1. Erten-Lyons D., Dodge H.H., Woltjer R., Silbert L.C., Howieson D.B., Kramer P., Kaye J.A. Neuropathologic basis of age-associated brain atrophy. JAMA Neurol. 2013;70:616–622. - PMC - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources