Performance characteristics of methods for quantifying spontaneous intracerebral haemorrhage: data from the Efficacy of Nitric Oxide in Stroke (ENOS) trial

The probability of cerebral amyloid angiopathy according to the Simplified Edinburgh CT criteria in a large, unselected lobar intracerebral hemorrhage population

PURPOSE

Early identification of the underlying cause of intracerebral hemorrhage (ICH) is important for treatment and prognosis. This study aims to investigate the association of hematoma volume and other clinical parameters on the distribution of cerebral amyloid angiopathy (CAA) probability according to the simplified Edinburgh CT criteria in a large, unselected intracerebral hemorrhage (ICH) population.

METHOD

Patients with spontaneous ICH residing in Skane county registered with clinical data in the Swedish Stroke Register 2016-2020 were included. Radiological parameters were evaluated using baseline non-contrast CT (NCCT) for categorization according to the simplified Edinburgh CT criteria by the presence of subarachnoid hemorrhage (SAH) and fingerlike-projections (FLP). Multivariable logistic regression analysis was used to determine factors associated with an increased (intermediate/high) CAA probability.

RESULTS

Of 666 patients with lobar ICH, 190 (29%) had high, 92 (14%) had intermediate, and 384 (58%) had low CAA probability. Patients with increased CAA probability presented more often with decreased level of consciousness, larger hematoma volumes, and higher 90-day mortality. Baseline hematoma volume [10-30 ml (OR = 4.03;95%CI: 2.26-7.19); 30-80 ml (OR = 12.00;95%CI:7.26-22.53); >80 ml (OR = 30.00;95%CI:15.94-59.09)], female sex (OR = 1.58;95%CI:1.08-2.32) and age (OR = 1.04;95%CI:1.02-1.06) were associated with an increased odds of having an increased CAA probability.

CONCLUSION

We identified a strong association between baseline hematoma volume and an increased probability of CAA in lobar ICH patients on NCCT, indicating that large hematoma volumes alone may contribute to the occurrence of FLP and SAH, and act as a confounder for the simplified Edinburgh CT criteria. Validation against MRI is warranted.

Radiological features of brain hemorrhage through automated segmentation from computed tomography in stroke and traumatic brain injury

Introduction

Radiological assessment is necessary to diagnose spontaneous intracerebral hemorrhage (ICH) and traumatic brain injury intracranial hemorrhage (TBI-bleed). Artificial intelligence (AI) deep learning tools provide a means for decision support. This study evaluates the hemorrhage segmentations produced from three-dimensional deep learning AI model that was developed using non-contrast computed tomography (CT) imaging data external to the current study.

Methods

Non-contrast CT imaging data from 1263 patients were accessed across seven data sources (referred to as sites) in Norway and Sweden. Patients were included based on ICH, TBI-bleed, or mild TBI diagnosis. Initial non-contrast CT images were available for all participants. Hemorrhage location frequency maps were generated. The number of estimated haematoma clusters was correlated with the total haematoma volume. Ground truth expert annotations were available for one ICH site; hence, a comparison was made with the estimated haematoma volumes. Segmentation volume estimates were used in a receiver operator characteristics (ROC) analysis for all samples (i.e., bleed detected) and then specifically for one site with few TBI-bleed cases.

Results

The hemorrhage frequency maps showed spatial patterns of estimated lesions consistent with ICH or TBI-bleed presentations. There was a positive correlation between the estimated number of clusters and total haematoma volume for each site (correlation range: 0.45-0.74; each p-value < 0.01) and evidence of ICH between-site differences. Relative to hand-drawn annotations for one ICH site, the VIOLA-AI segmentation mask achieved a median Dice Similarity Coefficient of 0.82 (interquartile range: 0.78 and 0.83), resulting in a small overestimate in the haematoma volume by a median of 0.47 mL (interquartile range: 0.04 and 1.75 mL). The bleed detection ROC analysis for the whole sample gave a high area-under-the-curve (AUC) of 0.92 (with sensitivity and specificity of 83.28% and 95.41%); however, when considering only the mild head injury site, the TBI-bleed detection gave an AUC of 0.70.

Discussion

An open-source segmentation tool was used to visualize hemorrhage locations across multiple data sources and revealed quantitative hemorrhage site differences. The automated total hemorrhage volume estimate correlated with a per-participant hemorrhage cluster count. ROC results were moderate-to-high. The VIOLA-AI tool had promising results and might be useful for various types of intracranial hemorrhage.

Accuracy of automated intracerebral hemorrhage volume measurement on non-contrast computed tomography: a Swedish Stroke Register cohort study

PURPOSE

Hematoma volume is the strongest predictor of patient outcome after intracerebral hemorrhage (ICH). The aim of this study was to validate novel fully automated software for quantification of ICH volume on non-contrast computed tomography (CT).

METHODS

The population was defined from the Swedish Stroke Register (RS) and included all patients with an ICH diagnosis during 2016-2019 in Region Skåne. Hemorrhage volume on their initial head CT was measured using ABC/2 and manual segmentation (Sectra IDS7 volume measurement tool) and the automated volume quantification tool (qER-NCCT) by Qure.ai. The first 500 were examined by two independent readers.

RESULTS

A total of 1649 ICH patients were included. The qER-NCCT had 97% sensitivity in identifying ICH. In total, there was excellent agreement between volumetric measurements of ICH volumes by qER-NCCT and manual segmentation by interclass correlation (ICC = 0.96), and good agreement (ICC = 0.86) between qER-NCCT and ABC/2 method. The qER-NCCT showed volume underestimation, mainly in large (> 30 ml) heterogenous hemorrhages. Interrater agreement by (ICC) was 0.996 (95% CI: 0.99-1.00) for manual segmentation.

CONCLUSION

Our study showed excellent agreement in volume quantification between the fully automated software qER-NCCT and manual segmentation of ICH on NCCT. The qER-NCCT would be an important additive tool by aiding in early diagnostics and prognostication for patients with ICH and in provide volumetry on a population-wide level. Further refinement of the software should address the underestimation of ICH volume seen in a portion of large, heterogenous, irregularly shaped ICHs.

Computed tomography in acute intracerebral hemorrhage: neuroimaging predictors of hematoma expansion and outcome

Intracerebral hemorrhage (ICH) accounts for 10–20% of all strokes worldwide and is associated with serious outcomes, including a 30-day mortality rate of up to 40%. Neuroimaging is pivotal in diagnosing ICH as early detection and determination of underlying cause, and risk for expansion/rebleeding is essential in providing the correct treatment. Non-contrast computed tomography (NCCT) is the most used modality for detection of ICH, identification of prognostic markers and measurements of hematoma volume, all of which are of major importance to predict outcome. The strongest predictors of 30-day mortality and functional outcome for ICH patients are baseline hematoma volume and hematoma expansion. Even so, exact hematoma measurement is rare in clinical routine practice, primarily due to a lack of tools available for fast, effective, and reliable volumetric tools. In this educational review, we discuss neuroimaging findings for ICH from NCCT images, and their prognostic value, as well as the use of semi-automatic and fully automated hematoma volumetric methods and assessment of hematoma expansion in prognostic studies.

Measures of intracranial compartments in acute intracerebral haemorrhage: data from the Rapid Intervention with Glyceryl Trinitrate in Hypertensive Stroke-2 Trial (RIGHT-2)

BACKGROUND AND PURPOSE

Intracerebral haemorrhage volume (ICHV) is prognostically important but does not account for intracranial volume (ICV) and cerebral parenchymal volume (CPV). We assessed measures of intracranial compartments in acute ICH using computerised tomography scans and whether ICHV/ICV and ICHV/CPV predict functional outcomes. We also assessed if cistern effacement, midline shift, old infarcts, leukoaraiosis and brain atrophy were associated with outcomes.

METHODS

Data from 133 participants from the Rapid Intervention with Glyceryl Trinitrate in Hypertensive Stroke-2 Trial trial were analysed. Measures included ICHV (using ABC/2) and ICV (XYZ/2) (by independent observers); ICHV, ICV and CPV (semiautomated segmentation, SAS); atrophy (intercaudate distance, ICD, Sylvian fissure ratio, SFR); midline shift; leukoaraiosis and cistern effacement (visual assessment). The effects of these measures on death at day 4 and poor functional outcome at day 90 (modified Rankin scale, mRS of >3) was assessed.

RESULTS

ICV was significantly different between XYZ and SAS: mean (SD) of 1357 (219) vs 1420 (196), mean difference (MD) 62 mL (p<0.001). There was no significant difference in ICHV between ABC/2 and SAS. There was very good agreement for ICV measured by SAS, CPV, ICD, SFR, leukoaraiosis and cistern score (all interclass correlations, n=10: interobserver 0.72-0.99, intraobserver 0.73-1.00). ICHV/ICV and ICHV/CPV were significantly associated with mRS at day 90, death at day 4 and acute neurological deterioration (all p<0.05), similar to ICHV. Midline shift and cistern effacement at baseline were associated with poor functional outcome but old infarcts, leukoaraiosis and brain atrophy were not.

CONCLUSIONS

Intracranial compartment measures and visual estimates are reproducible. ICHV adjusted for ICH and CPV could be useful to prognosticate in acute stroke. The presence of midline shift and cistern effacement may predict outcome but the mechanisms need validation in larger studies.

Association Between Perihematomal Perfusion and Intracerebral Hemorrhage Outcome

BACKGROUND

The prognostic impact of perihematomal hypoperfusion in patients with acute intracerebral hemorrhage (ICH) remains unclear. We tested the hypothesis that perihematomal hypoperfusion predicts poor ICH outcome and explored whether hematoma growth (HG) is the pathophysiological mechanism behind this association.

METHODS

A prospectively collected single-center cohort of consecutive ICH patients undergoing computed tomography perfusion on admission was analyzed. Cerebral blood flow (pCBF) was measured in the manually outlined perihematomal low-density area. pCBF was categorized into normal (40-55 mL/100 g/min), low (< 40 mL/100 g/min), and high (> 55 mL/100 g/min). HG was calculated as total volume increase from baseline to follow-up CT. A modified Rankin scale > 2 at three months was the outcome of interest. The association between cerebral perfusion and outcome was investigated with logistic regression, and potential mediators of this relationship were explored with mediation analysis.

RESULTS

A total of 155 subjects were included, of whom 55 (35.5%) had poor outcome. The rates of normal pCBF, low pCBF, and high pCBF were 17.4%, 68.4%, and 14.2%, respectively. After adjustment for confounders and keeping subjects with normal pCBF as reference, the risk of poor outcome was increased in patients with pCBF < 40 mL/100 g/min (odds ratio 6.11, 95% confidence interval 1.09-34.35, p = 0.040). HG was inversely correlated with pCBF (R = -0.292, p < 0.001) and mediated part of the association between pCBF and outcome (proportion mediated: 82%, p = 0.014).

CONCLUSION

Reduced pCBF is associated with poor ICH outcome in patients with mild-moderate severity. HG appears a plausible biological mediator but does not fully account for this association, and other mechanisms might be involved.

Apolipoprotein E genotype predicts subarachnoid extension in spontaneous intracerebral haemorrhage

BACKGROUND AND PURPOSE

Spontaneous intracerebral haemorrhage (ICH) with subarachnoid extension (SAHE) predicts poor outcomes and haematoma expansion in spontaneous ICH and is also a potential predictor of the severity of vascular amyloid deposition. The biological underpinnings of SAHE remain elusive. A study was conducted to identify risk factors associated with SAHE.

METHODS

A retrospective analysis was performed of an ongoing prospective cohort of primary spontaneous supratentorial ICH patients admitted to Tongji Hospital. SAHE was rated on baseline noncontrast computed tomography images by investigators blinded to the clinical data.

RESULTS

A total of 189 patients were enrolled. Apolipoprotein E (APOE) ε2 copies (p = 0.020), but not APOE ε4 copies (p > 0.2), were more common in patients with SAHE in univariate analysis. After controlling for confounding factors in multiple logistic regression, lobar haematoma (odds ratio [OR] 14.21, 95% confidence interval [CI] 5.89-34.33; p < 0.001), large haematoma volume (OR 1.04, 95% CI 1.02-1.06; p < 0.001) and APOE ε2 copies (OR 3.07, 95% CI 1.05-8.97; p = 0.041) were three independent predictors of SAHE. For subgroup analysis stratified by location, APOE ε2 showed a possible association with SAHE in lobar ICH (p = 0.026) but not in deep ICH (p > 0.2). No significant association was found between APOE ε4 copies and either lobar (p > 0.2) or deep ICH (p > 0.2).

CONCLUSIONS

The APOE ε2 allele predicts SAHE in spontaneous supratentorial ICH. The association may predominantly apply to lobar ICH. Given the established relationship between the APOE ε2 allele and pathological cerebrovascular changes, our findings suggest that SAHE involves genetically driven vessel pathology.

Novel and accurate non-linear index for the automated detection of haemorrhagic brain stroke using CT images

Brain stroke is an emergency medical condition which occurs mainly due to insufficient blood flow to the brain. It results in permanent cellular-level damage. There are two main types of brain stroke, ischemic and hemorrhagic. Ischemic brain stroke is caused by a lack of blood flow, and the haemorrhagic form is due to internal bleeding. The affected part of brain will not function properly after this attack. Hence, early detection is important for more efficacious treatment. Computer-aided diagnosis is a type of non-invasive diagnostic tool which can help in detecting life-threatening disease in its early stage by utilizing image processing and soft computing techniques. In this paper, we have developed one such model to assess intracerebral haemorrhage by employing non-linear features combined with a probabilistic neural network classifier and computed tomography (CT) images. Our model achieved a maximum accuracy of 97.37% in discerning normal versus haemorrhagic subjects. An intracerebral haemorrhage index is also developed using only three significant features. The clinical and statistical validation of the model confirms its suitability in providing for improved treatment planning and in making strategic decisions.

Ultra-Early Venous Thromboembolism (VTE) Prophylaxis in Spontaneous Intracerebral Hemorrhage (sICH)

OBJECTIVE

To determine if ultra-early (<24 h) venous thromboembolism (VTE) prophylaxis was associated with hematoma growth in spontaneous intracerebral hemorrhage (ICH).

BACKGROUND

Patients with ICH have a high risk of VTE. Pharmacological prophylaxis such as unfractionated heparin (UFH) have been demonstrated to reduce VTE. However, published datasets exclude patients with recent ICH out of concern for hematoma enlargement. American Heart/Stroke Association guidelines recommend UFH 1-4 days after hematoma stabilization while the European Stroke Organization has no recommendations on when to begin UFH. Our institutional practice is to obtain stability CT scans at 6 to 24 h and to begin UFH following documented clinical and radiologic stability. We examined the impact of this practice on hematoma expansion.

METHODS

We performed a retrospective cohort analysis of consecutive ICH patients treated at a single tertiary academic referral center in the US. Demographic and clinical characteristics were abstracted. ICH volume was measured via 3D volumetrics for a CT head done on admission, follow-up stability, and prior to discharge. The primary outcome was analyzed as ≥3 mL hematoma enlargement. Secondary outcomes include hematoma expansion of ≥6mL and ≥ 33%, length of stay (LOS), discharge disposition and mortality.

RESULTS

A total of 163 ICH patients were analyzed. There were 58 (35.6%) patients in the ultra-early UFH group and UFH was initiated on average at 13.8 h from initial scan. There were 105 (64.6%) patients in the standard group who initiated UFH at an average of 46.6 h. The primary outcome of hematoma enlargement ≥3 mL was observed in 2/58(3.4%) patients with ultra-early initiation of UFH and in 7/105(6.7%) in the standard group (p=0.49). Secondary outcomes were not significant including hematoma expansion in the ultra-early group ≥ 6 mL 3/58 (5.2%) and ≥33% 7/58 (12.1%) (p=0.91, 0.61, respectively) as well as mortality or LOS.

CONCLUSION

Venous thromboembolism prophylaxis started ultra-early (≤24 h) after ICH was not associated with hematoma expansion.

Original and Modified Graeb Score Correlation With Intraventricular Hemorrhage and Clinical Outcome Prediction in Hyperacute Intracranial Hemorrhage

Background and Purpose- The Graeb score is a visual rating scale of intraventricular hemorrhage (IVH) on noncontrast head CT. Little data exist in the hyperacute (<6 hour) period for reliability and predictive value of the modified Graeb Score (mGS) or the original Graeb Score (oGS) for clinical outcomes or their correlation with quantitative IVH volumes. Methods- A retrospective analysis of multicenter prospective intracranial hemorrhage study was performed. oGS and mGS inter-observer agreement and IVH volume correlation on the baseline noncontrast head CT were calculated by intraclass correlation coefficient and Pearson coefficient respectively. Predictors of poor outcome (modified Rankin Scale scores ≥4) at 3 months were identified using a backward stepwise selection multivariable analysis. oGS and mGS performance for modified Rankin Scale scores ≥4 was determined by receiver operating characteristic analysis. Results- One hundred forty-one patients (65±12 years) with median (interquartile range) time to CT of 82.5 (70.3-157.5) minutes were included. IVH was observed in 43 (30%) patients. Inter-observer agreement was excellent for both oGS (intraclass correlation coefficient, 0.90 [95% CI, 0.80-0.95]) and mGS (intraclass correlation coefficient, 0.97 [95% CI, 0.84-0.99]). mGS (R=0.79; P<0.01) correlated better than oGS (R=0.71; P<0.01) with IVH volumes (P=0.02). Models of thresholded oGS and mGS were not different from a model of planimetric baseline intracranial hemorrhage and IVH volume for poor outcome prediction. Area under the curves were 0.70, 0.73, and 0.72, respectively. Conclusions- Excellent correlation for oGS and mGS with IVH volume was seen. Thresholded oGS and mGS are reasonable surrogates for planimetric IVH volume for hyperacute intracranial hemorrhage studies.

Real-world Independent Testing of e-ASPECTS Software (RITeS): statistical analysis plan

Background: Artificial intelligence-based software may automatically detect ischaemic stroke lesions and provide an Alberta Stroke Program Early CT score (ASPECTS) on CT, and identify arterial occlusion and provide a collateral score on CTA. Large-scale independent testing will inform clinical use, but is lacking. We aim to test e-ASPECTS and e-CTA (Brainomix, Oxford UK) using CT scans obtained from a range of clinical studies. Methods: Using prospectively collected baseline CT and CTA scans from 10 national/international clinical stroke trials or registries (total >6600 patients), we will select a large clinically representative sample for testing e-ASPECTS and e-CTA compared to previously acquired independent expert human interpretation (reference standard). Our primary aims are to test agreement between software-derived and masked human expert ASPECTS, and the diagnostic accuracy of e-ASPECTS for identifying all causes of stroke symptoms using follow-up imaging and final clinical opinion as diagnostic ground truth. Our secondary aims are to test when and why e-ASPECTS is more or less accurate, or succeeds/fails to produce results, agreement between e-CTA and human expert CTA interpretation, and repeatability of e-ASPECTS/e-CTA results. All testing will be conducted on an intention-to-analyse basis. We will assess agreement between software and expert-human ratings and test the diagnostic accuracy of software.  Conclusions: RITeS will provide comprehensive, robust and representative testing of e-ASPECTS and e-CTA against the current gold-standard, expert-human interpretation.

Antihyperthermic treatment decreases perihematomal hypodensity

OBJECTIVE

To investigate the effect on perihematomal hypodensity and outcome of a decrease in body temperature in the first 24 hours in patients with intracerebral hemorrhage (ICH).

METHODS

In this retrospective study on a prospectively registered database, among the 1,100 patients, 795 met all the inclusion criteria. Temperature variations in the first 24 hours and perihematomal hypodensity (PHHD) were recorded. Patients ≥37.5°C were treated with antihyperthermic drugs for at least 48 hours. The main objective was to determine the association among temperature variation, PHHD, and outcome at 3 months.

RESULTS

The decrease in temperature in the first 24 hours increased the possibility of good outcome 11-fold. Temperature decrease, lower PHHD volume, and a good outcome were observed in 31.8% of the patients who received antihyperthermic treatment.

CONCLUSION

The administration of early antihyperthermic treatment in patients with spontaneous ICH with a basal axillary temperature ≥37.5°C resulted in good outcome in a third of the treated patients.

Subarachnoid Extension Predicts Lobar Intracerebral Hemorrhage Expansion

Background and Purpose- We investigated whether subarachnoid extension (SAHE) of intracerebral hemorrhage (ICH) is associated with hematoma expansion (HE). Methods- Retrospective analysis of patients with primary spontaneous ICH admitted at 3 academic hospitals in Italy. The study population was divided into a development and a replication cohort. SAHE was rated on baseline noncontrast computed tomography by investigators blinded to clinical data. The main outcome of interest was HE, defined as ICH growth >33% mL and/or >6 mL. Predictors of HE were explored with multivariable logistic regression stratified by ICH location (lobar versus nonlobar). Results- A total of 360 and 192 patients were included in the development and replication cohort, respectively. SAHE was identified with good interrater reliability (K=0.82), and its frequency was 27.8% in the development and 24.5% in the replication cohort. In univariate analysis, HE was more common in patients with SAHE (52.0% versus 27.3%; P<0.001). When controlling for confounders in logistic regression, SAHE was an independent predictor of lobar HE (odds ratio, 6.00 [95% CI, 2.16-16.64]; P=0.001) whereas there was no association with HE in nonlobar ICH (odds ratio, 0.55 [95% CI, 0.17-1.84]; P=0.334). The increased risk of HE in lobar ICH with SAHE was confirmed in the replication cohort (odds ratio, 3.46 [95% CI, 1.07-11.20]; P=0.038). Conclusions- SAHE predicts HE in lobar ICH. This may improve the stratification of HE risk in clinical practice or future trials targeting HE. Further research is needed to confirm our findings and characterize the underlying biological mechanisms.

Comparison of ABC Methods with Computerized Estimates of Intracerebral Hemorrhage Volume: The INTERACT2 Study

BACKGROUND AND PURPOSE

Hematoma volume is a key determinant of outcome in acute intracerebral hemorrhage (ICH). We aimed to compare estimates of ICH volume between simple (ABC/2, length, width, and height) and gold standard planimetric software approaches.

METHODS

Data are from the second Intensive Blood Pressure Reduction in Acute Cerebral Hemorrhage Trial (INTERACT2). Multivariable linear regression was used to compare ICH volumes on baseline CT scans using the ABC/2, modified ABC/2 (mABC/2), and MIStar software. Other aspects of ICH morphology examined included location, irregularity, heterogeneity, intraventricular and subarachnoid hemorrhage extension (SAH) of hematoma, and associated white matter lesions and brain atrophy.

RESULTS

In 2,084 patients with manual and semiautomated measurements, median (IQR) ICH volumes for each approach were: ABC/2 11.1 (5.11-20.88 mL), mABC/2 7.8 (3.88-14.11 mL), and MIStar 10.7 (5.59-18.66 mL). Median differences between ABC/2 and MIStar, and mABC/2 and MIStar were 0.34 (-1.01 to 2.96) and -2.4 (-4.95 to -0.7416), respectively. Hematoma volumes differed significantly with irregular shape (ABC/2 and MIStar, p < 0.001; mABC/2 and MIStar, p = 0.007) and larger volumes (mABC/2 and MIStar, p < 0.001; ABC/2 and MIStar, p = 0.07). ICH with SAH showed a significant discrepancy between ABC/2 and MIStar (p < 0.001).

CONCLUSIONS

Overall, ABC/2 performs better than mABC/2 in estimating ICH volume. The largest discrepancies were evidenced against automated software for irregular-shaped and large ICH with SAH, but the clinical significance of this is uncertain.

Intracerebral Haemorrhage Segmentation in Non-Contrast CT

A 3-dimensional (3D) convolutional neural network is presented for the segmentation and quantification of spontaneous intracerebral haemorrhage (ICH) in non-contrast computed tomography (NCCT). The method utilises a combination of contextual information on multiple scales for fast and fully automatic dense predictions. To handle a large class imbalance present in the data, a weight map is introduced during training. The method was evaluated on two datasets of 25 and 50 patients respectively. The reference standard consisted of manual annotations for each ICH in the dataset. Quantitative analysis showed a median Dice similarity coefficient of 0.91 [0.87-0.94] and 0.90 [0.85-0.92] for the two test datasets in comparison to the reference standards. Evaluation of a separate dataset of 5 patients for the assessment of the observer variability produced a mean Dice similarity coefficient of 0.95 ± 0.02 for the inter-observer variability and 0.97 ± 0.01 for the intra-observer variability. The average prediction time for an entire volume was 104 ± 15 seconds. The results demonstrate that the method is accurate and approaches the performance of expert manual annotation.

Extent of surrounding edema does not correlate with acute complications after radiosurgery for melanoma brain metastases

AIM

To assess whether extent of surrounding edema correlates with acute adverse clinical outcomes within 3 months after stereotactic radiosurgery (SRS) for melanoma brain metastases (BM).

METHODS

Patients with melanoma BM treated with SRS were included in a single center retrospective analysis. A contrast-enhanced magnetic resonance image (MRI) brain was acquired on the day of treatment and used to calculate the volume of the largest lesion (the index BM) and total volume of all BM. Their corresponding volume of surrounding edema was defined based on the fluid attenuated inversion recovery (FLAIR) sequence. After SRS, MRI was performed every 3 months for at least 2 years if the patient remained well enough to do so. Adverse neurologic events after SRS were defined using common terminology criteria for adverse events (CTCAE) version 5.0. Multivariate regression analyses assessed for associations between BM size and edema at baseline with increasing edema and neurologic adverse events within 3 months after SRS.

RESULTS

Mean volume of the index BM reduced from 2.2 to 0.5 cm³ at 3 months after SRS (p = 0.03). Mean volume of edema surrounding the index BM was 6.4 cm³ at baseline, 10.2 cm³ at 3 months and 5.5 cm³ at 6 months. There were 7/43 (16%) patients that experienced an adverse neurological event within 3 months (attributable to any cause) and 4/43 (9%) were associated with an increase in BM edema. On univariate and multivariate analyses, there were no correlations between any baseline factors and volume of edema at 3 months. However, SRS dose delivered and systemic therapy use within 4 weeks of SRS both correlated with a reduction in edema surrounding the index BM.

CONCLUSION

A transient increase in mean volume of edema was apparent at 3 months after SRS. However, this resolved by 6 months and did not correlate with adverse events or dexamethasone requirement. Thus, the clinical significance is uncertain.

Clinical Outcomes and Neuroimaging Profiles in Nondisabled Patients With Anticoagulant-Related Intracerebral Hemorrhage

Background and Purpose- The aim of this study was to prospectively validate our prior findings of smaller hematoma volume and lesser neurological deficit in nonvitamin K oral anticoagulant (NOAC) compared with Vitamin K antagonist (VKA)-related intracerebral hemorrhage (ICH). Methods- Prospective 12-month observational study in 15 tertiary stroke centers in the United States, Europe, and Asia. Consecutive patients with premorbid modified Rankin Scale score of <2 with acute nontraumatic anticoagulant-related ICH divided into 2 groups according to the type of anticoagulant: NOAC versus VKA. We recorded baseline ICH volume, significant hematoma expansion (absolute [12.5 mL] or relative [>33%] increase), neurological severity measured by National Institutes of Health Stroke Scale score, 90-day mortality, and functional status (modified Rankin Scale score). Results- Our cohort comprised 196 patients, 62 NOAC related (mean age, 75.0±11.4 years; 54.8% men) and 134 VKA related (mean age, 72.3±10.5; 73.1% men). There were no differences in vascular comorbidities, antiplatelet, and statin use; NOAC-related ICH patients had lower median baseline hematoma volume (13.8 [2.5-37.6] versus 19.5 [6.6-52.0] mL; P=0.026) and were less likely to have severe neurological deficits (National Institutes of Health Stroke Scale score of >10 points) on admission (37% versus 55.3%, P=0.025). VKA-ICH were more likely to have significant hematoma expansion (37.4% versus 17%, P=0.008). NOAC pretreatment was independently associated with smaller baseline hematoma volume (standardized linear regression coefficient:-0.415 [95% CI, -0.780 to -0.051]) resulting in lower likelihood of severe neurological deficit (odds ratio, 0.44; 95% CI, 0.22-0.85) in multivariable-adjusted models. Conclusions- Patients with NOAC-related ICH have smaller baseline hematoma volumes and lower odds of severe neurological deficit compared with VKA-related ICH. These findings are important for practicing clinicians making anticoagulation choices.

Abilities of a Densitometric Analysis of Computed Tomography Images and Hemorrhagic Parameters to Predict Outcome Favorability in Patients With Intracerebral Hemorrhage

BACKGROUND

Intracerebral hemorrhage (ICH) is one of the most devastating subtypes of stroke. A rapid assessment of ICH severity involves the use of computed tomography (CT) and derivation of the hemorrhage volume, which is often estimated using the ABC/2 method. However, these estimates are highly inaccurate and may not be feasible for anticipating outcome favorability.

OBJECTIVE

To predict patient outcomes via a quantitative, densitometric analysis of CT images, and to compare the predictive power of these densitometric parameters with the conventional ABC/2 volumetric parameter and segmented hemorrhage volumes.

METHODS

Noncontrast CT images of 87 adult patients with ICH (favorable outcomes = 69, unfavorable outcomes = 12, and deceased = 6) were analyzed. In-house software was used to calculate the segmented hemorrhage volumes, ABC/2 and densitometric parameters, including the skewness and kurtosis of the density distribution, interquartile ranges, and proportions of specific pixels in sets of CT images. Nonparametric statistical analyses were conducted.

RESULTS

The densitometric parameter interquartile range exhibited greatest accuracy (82.7%) in predicting favorable outcomes. The combination of skewness and the interquartile range effectively predicted mortality (accuracy = 83.3%). The actual volume of the ICH exhibited good coherence with ABC/2 (R = 0.79). Both parameters predicted mortality with moderate accuracy (<78%) but were less effective in predicting unfavorable outcomes.

CONCLUSION

Hemorrhage volume was rapidly estimated and effectively predicted mortality in patients with ICH; however, this value may not be useful for predicting favorable outcomes. The densitometric analysis exhibited significantly higher power in predicting mortality and favorable outcomes in patients with ICH.

Prehospital transdermal glyceryl trinitrate in patients with ultra-acute presumed stroke (RIGHT-2): an ambulance-based, randomised, sham-controlled, blinded, phase 3 trial

BACKGROUND

High blood pressure is common in acute stroke and is a predictor of poor outcome; however, large trials of lowering blood pressure have given variable results, and the management of high blood pressure in ultra-acute stroke remains unclear. We investigated whether transdermal glyceryl trinitrate (GTN; also known as nitroglycerin), a nitric oxide donor, might improve outcome when administered very early after stroke onset.

METHODS

We did a multicentre, paramedic-delivered, ambulance-based, prospective, randomised, sham-controlled, blinded-endpoint, phase 3 trial in adults with presumed stroke within 4 h of onset, face-arm-speech-time score of 2 or 3, and systolic blood pressure 120 mm Hg or higher. Participants were randomly assigned (1:1) to receive transdermal GTN (5 mg once daily for 4 days; the GTN group) or a similar sham dressing (the sham group) in UK-based ambulances by paramedics, with treatment continued in hospital. Paramedics were unmasked to treatment, whereas participants were masked. The primary outcome was the 7-level modified Rankin Scale (mRS; a measure of functional outcome) at 90 days, assessed by central telephone follow-up with masking to treatment. Analysis was hierarchical, first in participants with a confirmed stroke or transient ischaemic attack (cohort 1), and then in all participants who were randomly assigned (intention to treat, cohort 2) according to the statistical analysis plan. This trial is registered with ISRCTN, number ISRCTN26986053.

FINDINGS

Between Oct 22, 2015, and May 23, 2018, 516 paramedics from eight UK ambulance services recruited 1149 participants (n=568 in the GTN group, n=581 in the sham group). The median time to randomisation was 71 min (IQR 45-116). 597 (52%) patients had ischaemic stroke, 145 (13%) had intracerebral haemorrhage, 109 (9%) had transient ischaemic attack, and 297 (26%) had a non-stroke mimic at the final diagnosis of the index event. In the GTN group, participants' systolic blood pressure was lowered by 5·8 mm Hg compared with the sham group (p<0·0001), and diastolic blood pressure was lowered by 2·6 mm Hg (p=0·0026) at hospital admission. We found no difference in mRS between the groups in participants with a final diagnosis of stroke or transient ischaemic stroke (cohort 1): 3 (IQR 2-5; n=420) in the GTN group versus 3 (2-5; n=408) in the sham group, adjusted common odds ratio for poor outcome 1·25 (95% CI 0·97-1·60; p=0·083); we also found no difference in mRS between all patients (cohort 2: 3 [2-5]; n=544, in the GTN group vs 3 [2-5]; n=558, in the sham group; 1·04 [0·84-1·29]; p=0·69). We found no difference in secondary outcomes, death (treatment-related deaths: 36 in the GTN group vs 23 in the sham group [p=0·091]), or serious adverse events (188 in the GTN group vs 170 in the sham group [p=0·16]) between treatment groups.

INTERPRETATION

Prehospital treatment with transdermal GTN does not seem to improve functional outcome in patients with presumed stroke. It is feasible for UK paramedics to obtain consent and treat patients with stroke in the ultra-acute prehospital setting.

FUNDING

British Heart Foundation.

Automated segmentation of haematoma and perihaematomal oedema in MRI of acute spontaneous intracerebral haemorrhage

BACKGROUND

Spontaneous intracerebral haemorrhage (SICH) is a common condition with high morbidity and mortality. Segmentation of haematoma and perihaematoma oedema on medical images provides quantitative outcome measures for clinical trials and may provide important markers of prognosis in people with SICH.

METHODS

We take advantage of improved contrast seen on magnetic resonance (MR) images of patients with acute and early subacute SICH and introduce an automated algorithm for haematoma and oedema segmentation from these images. To our knowledge, there is no previously proposed segmentation technique for SICH that utilises MR images directly. The method is based on shape and intensity analysis for haematoma segmentation and voxel-wise dynamic thresholding of hyper-intensities for oedema segmentation.

RESULTS

Using Dice scores to measure segmentation overlaps between labellings yielded by the proposed algorithm and five different expert raters on 18 patients, we observe that our technique achieves overlap scores that are very similar to those obtained by pairwise expert rater comparison. A further comparison between the proposed method and a state-of-the-art Deep Learning segmentation on a separate set of 32 manually annotated subjects confirms the proposed method can achieve comparable results with very mild computational burden and in a completely training-free and unsupervised way.

CONCLUSION

Our technique can be a computationally light and effective way to automatically delineate haematoma and oedema extent directly from MR images. Thus, with increasing use of MR images clinically after intracerebral haemorrhage this technique has the potential to inform clinical practice in the future.

Absolute risk and predictors of the growth of acute spontaneous intracerebral haemorrhage: a systematic review and meta-analysis of individual patient data

Background

Intracerebral haemorrhage growth is associated with poor clinical outcome and is a therapeutic target for improving outcome. We aimed to determine the absolute risk and predictors of intracerebral haemorrhage growth, develop and validate prediction models, and evaluate the added value of CT angiography.

Methods

In a systematic review of OVID MEDLINE—with additional hand-searching of relevant studies' bibliographies— from Jan 1, 1970, to Dec 31, 2015, we identified observational cohorts and randomised trials with repeat scanning protocols that included at least ten patients with acute intracerebral haemorrhage. We sought individual patient-level data from corresponding authors for patients aged 18 years or older with data available from brain imaging initially done 0·5–24 h and repeated fewer than 6 days after symptom onset, who had baseline intracerebral haemorrhage volume of less than 150 mL, and did not undergo acute treatment that might reduce intracerebral haemorrhage volume. We estimated the absolute risk and predictors of the primary outcome of intracerebral haemorrhage growth (defined as >6 mL increase in intracerebral haemorrhage volume on repeat imaging) using multivariable logistic regression models in development and validation cohorts in four subgroups of patients, using a hierarchical approach: patients not taking anticoagulant therapy at intracerebral haemorrhage onset (who constituted the largest subgroup), patients taking anticoagulant therapy at intracerebral haemorrhage onset, patients from cohorts that included at least some patients taking anticoagulant therapy at intracerebral haemorrhage onset, and patients for whom both information about anticoagulant therapy at intracerebral haemorrhage onset and spot sign on acute CT angiography were known.

Findings

Of 4191 studies identified, 77 were eligible for inclusion. Overall, 36 (47%) cohorts provided data on 5435 eligible patients. 5076 of these patients were not taking anticoagulant therapy at symptom onset (median age 67 years, IQR 56–76), of whom 1009 (20%) had intracerebral haemorrhage growth. Multivariable models of patients with data on antiplatelet therapy use, data on anticoagulant therapy use, and assessment of CT angiography spot sign at symptom onset showed that time from symptom onset to baseline imaging (odds ratio 0·50, 95% CI 0·36–0·70; p<0·0001), intracerebral haemorrhage volume on baseline imaging (7·18, 4·46–11·60; p<0·0001), antiplatelet use (1·68, 1·06–2·66; p=0·026), and anticoagulant use (3·48, 1·96–6·16; p<0·0001) were independent predictors of intracerebral haemorrhage growth (C-index 0·78, 95% CI 0·75–0·82). Addition of CT angiography spot sign (odds ratio 4·46, 95% CI 2·95–6·75; p<0·0001) to the model increased the C-index by 0·05 (95% CI 0·03–0·07).

Interpretation

In this large patient-level meta-analysis, models using four or five predictors had acceptable to good discrimination. These models could inform the location and frequency of observations on patients in clinical practice, explain treatment effects in prior randomised trials, and guide the design of future trials.

Funding

UK Medical Research Council and British Heart Foundation.

A nested mechanistic sub-study into the effect of tranexamic acid versus placebo on intracranial haemorrhage and cerebral ischaemia in isolated traumatic brain injury: study protocol for a randomised controlled trial (CRASH-3 Trial Intracranial Bleeding Mechanistic Sub-Study [CRASH-3 IBMS])

Background

Tranexamic acid prevents blood clots from breaking down and reduces bleeding. However, it is uncertain whether tranexamic acid is effective in traumatic brain injury. The CRASH-3 trial is a randomised controlled trial that will examine the effect of tranexamic acid (versus placebo) on death and disability in 13,000 patients with traumatic brain injury. The CRASH-3 trial hypothesizes that tranexamic acid will reduce intracranial haemorrhage, which will reduce the risk of death. Although it is possible that tranexamic acid will reduce intracranial bleeding, there is also a potential for harm. In particular, tranexamic acid may increase the risk of cerebral thrombosis and ischaemia. The protocol detailed here is for a mechanistic sub-study nested within the CRASH-3 trial. This mechanistic sub-study aims to examine the effect of tranexamic acid (versus placebo) on intracranial bleeding and cerebral ischaemia.

Methods

The CRASH-3 Intracranial Bleeding Mechanistic Sub-Study (CRASH-3 IBMS) is nested within a prospective, double-blind, multi-centre, parallel-arm randomised trial called the CRASH-3 trial. The CRASH-3 IBMS will be conducted in a cohort of approximately 1000 isolated traumatic brain injury patients enrolled in the CRASH-3 trial. In the CRASH-3 IBMS, brain scans acquired before and after randomisation are examined, using validated methods, for evidence of intracranial bleeding and cerebral ischaemia. The primary outcome is the total volume of intracranial bleeding measured on computed tomography after randomisation, adjusting for baseline bleeding volume. Secondary outcomes include progression of intracranial haemorrhage (from pre- to post-randomisation scans), new intracranial haemorrhage (seen on post- but not pre-randomisation scans), intracranial haemorrhage following neurosurgery, and new focal ischaemic lesions (seen on post-but not pre-randomisation scans). A linear regression model will examine whether receipt of the trial treatment can predict haemorrhage volume. Bleeding volumes and new ischaemic lesions will be compared across treatment groups using relative risks and 95% confidence intervals.

Discussion

The CRASH-3 IBMS will provide an insight into the mechanism of action of tranexamic acid in traumatic brain injury, as well as information about the risks and benefits. Evidence from this trial could inform the management of patients with traumatic brain injury.

Trial registration

The CRASH-3 trial was prospectively registered and the CRASH-3 IBMS is an addition to the original protocol registered at the International Standard Randomised Controlled Trials registry (ISRCTN15088122) 19 July 2011, and ClinicalTrials.gov on 25 July 2011 (NCT01402882).

Electronic supplementary material

The online version of this article (doi:10.1186/s13063-017-2073-6) contains supplementary material, which is available to authorized users.

Development and Validation of an Automatic Segmentation Algorithm for Quantification of Intracerebral Hemorrhage

BACKGROUND AND PURPOSE

ABC/2 is still widely accepted for volume estimations in spontaneous intracerebral hemorrhage (ICH) despite known limitations, which potentially accounts for controversial outcome-study results. The aim of this study was to establish and validate an automatic segmentation algorithm, allowing for quick and accurate quantification of ICH.

METHODS

A segmentation algorithm implementing first- and second-order statistics, texture, and threshold features was trained on manual segmentations with a random-forest methodology. Quantitative data of the algorithm, manual segmentations, and ABC/2 were evaluated for agreement in a study sample (n=28) and validated in an independent sample not used for algorithm training (n=30).

RESULTS

ABC/2 volumes were significantly larger compared with either manual or algorithm values, whereas no significant differences were found between the latter (P<0.0001; Friedman+Dunn's multiple comparison). Algorithm agreement with the manual reference was strong (concordance correlation coefficient 0.95 [lower 95% confidence interval 0.91]) and superior to ABC/2 (concordance correlation coefficient 0.77 [95% confidence interval 0.64]). Validation confirmed agreement in an independent sample (algorithm concordance correlation coefficient 0.99 [95% confidence interval 0.98], ABC/2 concordance correlation coefficient 0.82 [95% confidence interval 0.72]). The algorithm was closer to respective manual segmentations than ABC/2 in 52/58 cases (89.7%).

CONCLUSIONS

An automatic segmentation algorithm for volumetric analysis of spontaneous ICH was developed and validated in this study. Algorithm measurements showed strong agreement with manual segmentations, whereas ABC/2 exhibited its limitations, yielding inaccurate overestimations of ICH volume. The refined, yet time-efficient, quantification of ICH by the algorithm may facilitate evaluation of clot volume as an outcome predictor and trigger for surgical interventions in the clinical setting.

Continuing versus Stopping Prestroke Antihypertensive Therapy in Acute Intracerebral Hemorrhage: A Subgroup Analysis of the Efficacy of Nitric Oxide in Stroke Trial

BACKGROUND AND PURPOSE

More than 50% of patients with acute intracerebral hemorrhage (ICH) are taking antihypertensive drugs before ictus. Although antihypertensive therapy should be given long term for secondary prevention, whether to continue or stop such treatment during the acute phase of ICH remains unclear, a question that was addressed in the Efficacy of Nitric Oxide in Stroke (ENOS) trial.

METHODS

ENOS was an international multicenter, prospective, randomized, blinded endpoint trial. Among 629 patients with ICH and systolic blood pressure between 140 and 220 mmHg, 246 patients who were taking antihypertensive drugs were assigned to continue (n = 119) or to stop (n = 127) taking drugs temporarily for 7 days. The primary outcome was the modified Rankin Score at 90 days. Secondary outcomes included death, length of stay in hospital, discharge destination, activities of daily living, mood, cognition, and quality of life.

RESULTS

Blood pressure level (baseline 171/92 mmHg) fell in both groups but was significantly lower at 7 days in those patients assigned to continue antihypertensive drugs (difference 9.4/3.5 mmHg, P < .01). At 90 days, the primary outcome did not differ between the groups; the adjusted common odds ratio (OR) for worse outcome with continue versus stop drugs was .92 (95% confidence interval, .45-1.89; P = .83). There was no difference between the treatment groups for any secondary outcome measure, or rates of death or serious adverse events.

CONCLUSIONS

Among patients with acute ICH, immediate continuation of antihypertensive drugs during the first week did not reduce death or major disability in comparison to stopping treatment temporarily.

Intraventricular Extension of Supratentorial Intracerebral Hemorrhage: The Modified Graeb Scale Improves Outcome Prediction in Lund Stroke Register

BACKGROUND/AIMS

The modified Graeb Scale (mGS) is a semi-quantitative method to assess the extension of intraventricular hemorrhage (IVH) in patients with intracerebral hemorrhage (ICH). The mGS has been shown to prognosticate outcome after ICH in cohorts derived from convenience samples. We evaluated the external validity of mGS in supratentorial ICH-patients from an unselected cohort.

METHODS

ICH-patients were included prospectively and consecutively in Lund Stroke Register. Follow-up survival status was obtained from the National Census Office; functional outcome was obtained from the Swedish Stroke Register or medical records. Using multivariate analyses, we examined if mGS was related to 30-day survival or poor functional outcome (modified Rankin Scale ≥4) at 90 days.

RESULTS

Of 198 supratentorial ICH-patients, 86 (43%) had IVH (median mGS 12, range 1-28). In multivariate regression analyses, the mGS independently predicted 30-day mortality (per point; OR 1.16; 95% CI 1.06-1.27; p = 0.002) and poor functional outcome (OR 1.11; 95% CI 1.02-1.20; p = 0.011) after ICH. In receiver-operator characteristic analysis, the addition of mGS tended to be associated with a higher prognostic accuracy for survival (area under curve 0.886 vs. not including mGS 0.812; p = 0.053).

CONCLUSIONS

The mGS improves outcome prediction after supratentorial ICH beyond other previously established factors in an unselected population.

Reversal strategies for vitamin K antagonists in acute intracerebral hemorrhage

OBJECTIVE

There is little evidence to guide treatment strategies for intracerebral hemorrhage on vitamin K antagonists (VKA-ICH). Treatments utilized in clinical practice include fresh frozen plasma (FFP) and prothrombin complex concentrate (PCC). Our aim was to compare case fatality with different reversal strategies.

METHODS

We pooled individual ICH patient data from 16 stroke registries in 9 countries (n = 10 282), of whom 1,797 (17%) were on VKA. After excluding 250 patients with international normalized ratio < 1.3 and/or missing data required for analysis, we compared all-cause 30-day case fatality using Cox regression.

RESULTS

We included 1,547 patients treated with FFP (n = 377, 24%), PCC (n = 585, 38%), both (n = 131, 9%), or neither (n = 454, 29%). The crude case fatality and adjusted hazard ratio (HR) were highest with no reversal (61.7%, HR = 2.540, 95% confidence interval [CI] = 1.784-3.616, p < 0.001), followed by FFP alone (45.6%, HR = 1.344, 95% CI = 0.934-1.934, p = 0.112), then PCC alone (37.3%, HR = 1.445, 95% CI = 1.014-2.058, p = 0.041), compared to reversal with both FFP and PCC (27.8%, reference). Outcomes with PCC versus FFP were similar (HR = 1.075, 95% CI = 0.874-1.323, p = 0.492); 4-factor PCC (n = 441) was associated with higher case fatality compared to 3-factor PCC (n = 144, HR = 1.441, 95% CI = 1.041-1.995, p = 0.027).

INTERPRETATION

The combination of FFP and PCC might be associated with the lowest case fatality in reversal of VKA-ICH, and FFP may be equivalent to PCC. Randomized controlled trials with functional outcomes are needed to establish the most effective treatment.