Imaging Biomarkers for Intra-arterial Stroke Therapy

Cortical and Internal Watershed Infarcts Might Be Key Signs for Predicting Neurological Deterioration in Patients with Internal Carotid Artery Occlusion with Mild Symptoms

Background

Treatment for acute ischemic stroke due to large vessel occlusion (LVO) with mild symptoms is under discussion. Although most patients have good outcomes, some patients deteriorate and have unfavorable results. Imaging findings that predict the prognosis of LVO with mild symptoms are needed to identify patients who require treatment. In this study, we focused on watershed infarctions (WSIs), because this clinical phenomenon quite sensitively reflects changes in cerebral blood flow. The purpose of this study was to assess positive rates of WSI on MRI findings in patients with internal carotid artery (ICA) occlusion, and compare WSI-positive rates between patients divided according to their clinical course.

Methods

We retrospectively collected data of 1,531 patients who presented with acute ischemic stroke between June 2006 and July 2019. Among them, we chose symptomatic ICA occlusion patients with a past history of atrial fibrillation who were treated conservatively. We divided these patients into two groups, those with maintenance or improvement in their NIHSS score after hospitalization, and those whose NIHSS score worsened. We compared WSI-positive rates between these two groups.

Results

Thirty-seven of the 1,531 patients were included in this study. Of them, total NIHSS score was maintained or improved in 8 patients (group A), 3 of whom (37.5%) had internal watershed infarctions (IWIs). In group B, consisting of patients whose NIHSS score worsened by >2 at 7 days from symptom onset, 24 (82.8%) had IWIs. Group A thus had statistically lower IWI positivity rates than group B (p = 0.02). Three patients (37.5%) in group A had cortical watershed infarctions (CWIs), while 27 patients in group B (93.1%) had CWIs. Group A thus had a significantly lower CWI positivity rate than group B (p = 0.002).

Conclusion

In patients with mildly symptomatic ICA occlusion, CWIs and IWIs might be key signs for predicting neurological deterioration after hospitalization.

Automatic segmentation of cerebral infarcts in follow-up computed tomography images with convolutional neural networks

Background and purpose

Infarct volume is a valuable outcome measure in treatment trials of acute ischemic stroke and is strongly associated with functional outcome. Its manual volumetric assessment is, however, too demanding to be implemented in clinical practice.

Objective

To assess the value of convolutional neural networks (CNNs) in the automatic segmentation of infarct volume in follow-up CT images in a large population of patients with acute ischemic stroke.

Materials and methods

We included CT images of 1026 patients from a large pooling of patients with acute ischemic stroke. A reference standard for the infarct segmentation was generated by manual delineation. We introduce three CNN models for the segmentation of subtle, intermediate, and severe hypodense lesions. The fully automated infarct segmentation was defined as the combination of the results of these three CNNs. The results of the three-CNNs approach were compared with the results from a single CNN approach and with the reference standard segmentations.

Results

The median infarct volume was 48 mL (IQR 15–125 mL). Comparison between the volumes of the three-CNNs approach and manually delineated infarct volumes showed excellent agreement, with an intraclass correlation coefficient (ICC) of 0.88. Even better agreement was found for severe and intermediate hypodense infarcts, with ICCs of 0.98 and 0.93, respectively. Although the number of patients used for training in the single CNN approach was much larger, the accuracy of the three-CNNs approach strongly outperformed the single CNN approach, which had an ICC of 0.34.

Conclusion

Convolutional neural networks are valuable and accurate in the quantitative assessment of infarct volumes, for both subtle and severe hypodense infarcts in follow-up CT images. Our proposed three-CNNs approach strongly outperforms a more straightforward single CNN approach.