A deep learning analysis of stroke onset time prediction and comparison to DWI-FLAIR mismatch

Abstract

Introduction: When time since stroke onset is unknown, DWI-FLAIR mismatch rating is an established technique for patient stratification. A visible DWI lesion without corresponding parenchymal hyperintensity on FLAIR suggests time since onset of under 4.5 h and thus a potential benefit from intravenous thrombolysis. To improve accuracy and availability of the mismatch concept, deep learning might be able to augment human rating and support decision-making in these cases.

Methods: We used unprocessed DWI and coregistered FLAIR imaging data to train a deep learning model to predict dichotomized time since ischemic stroke onset. We analyzed the performance of Group Convolutional Neural Networks compared to other deep learning methods. Unlabeled imaging data was used for pre-training. Prediction performance of the best deep learning model was compared to the performance of four independent junior and senior raters. Additionally, in cases deemed indeterminable by human raters, model ratings were used to augment human performance. Post-hoc gradient-based explanations were analyzed to gain insights into model predictions.

Results: Our best predictive model performed comparably to human raters. Using model ratings in cases deemed indeterminable by human raters improved rating accuracy and interrater agreement for junior and senior ratings. Post-hoc explainability analyses showed that the model localized stroke lesions to derive predictions.

Discussion: Our analysis shows that deep learning based clinical decision support has the potential to improve the accessibility of the DWI-FLAIR mismatch concept by supporting patient stratification.

Keywords: Acute ischemic stroke; Artificial intelligence; Cerebrovascular accident; Computer aided; DWI-FLAIR-mismatch; Decision support; Deep learning; Diffusion-weighted imaging; Fluid attenuated inversion recovery; Machine learning; Magnetic resonance imaging; Precision medicine; Wake up stroke.

Fig. 1

Deep learning architecture, visualized using the PlotNeuralNet software (Iqbal, Dec. 25, 2018.).

Fig. 2

Flowchart of patients used for labeled and unlabeled dataset(DWI = Diffusion-weighted imaging, FLAIR = Fluid-attenuation inversion recovery, TSS = time since stroke).

Fig. 3

Receiver operating characteristics curve for DL model identifying time since stroke onset within 4.5 h.

Fig. 4

Example xAI image showing a patient with a single infarct (A), scattered infarct (B) pattern and severe leukoaraiosis (C). Top row: DWI image (left), DWI with xAI heatmap overlay (right), bottom row: coregistered FLAIR image (left), with xAI heatmap overlay (right).