Quantitative CT radiomics-based models for prediction of haematoma expansion and poor functional outcome in primary intracerebral haemorrhage

Objectives To test radiomics-based features extracted from noncontrast CT of patients with spontaneous intracerebral haemorrhage for prediction of haematoma expansion and poor functional outcome and compare them with radiological signs and clinical factors. Materials and methods Seven hundred fifty-four radiomics-based features were extracted from 1732 scans derived from the TICH-2 multicentre clinical trial. Features were harmonised and a correlation-based feature selection was applied. Different elastic-net parameterisations were tested to assess the predictive performance of the selected radiomics-based features using grid optimisation. For comparison, the same procedure was run using radiological signs and clinical factors separately. Models trained with radiomics-based features combined with radiological signs or clinical factors were tested. Predictive performance was evaluated using the area under the receiver operating characteristic curve (AUC) score. Results The optimal radiomics-based model showed an AUC of 0.693 for haematoma expansion and an AUC of 0.783 for poor functional outcome. Models with radiological signs alone yielded substantial reductions in sensitivity. Combining radiomics-based features and radiological signs did not provide any improvement over radiomics-based features alone. Models with clinical factors had similar performance compared to using radiomics-based features, albeit with low sensitivity for haematoma expansion. Performance of radiomics-based features was boosted by incorporating clinical factors, with time from onset to scan and age being the most important contributors for haematoma expansion and poor functional outcome prediction, respectively. Conclusion Radiomics-based features perform better than radiological signs and similarly to clinical factors on the prediction of haematoma expansion and poor functional outcome. Moreover, combining radiomics-based features with clinical factors improves their performance. Key Points • Linear models based on CT radiomics-based features perform better than radiological signs on the prediction of haematoma expansion and poor functional outcome in the context of intracerebral haemorrhage. • Linear models based on CT radiomics-based features perform similarly to clinical factors known to be good predictors. However, combining these clinical factors with radiomics-based features increases their predictive performance.

The initial CT blend sign is not associated with poor patient outcomes after stereotactic minimally invasive surgery

Background The initial CT blend sign is an imaging marker that has been used to predict haematoma expansion and poor outcomes in patients with small-volume intracerebral haemorrhage (ICH). However, the association of the blend sign with the outcomes of patients undergoing surgery remains unclear. The present study aimed to retrospectively evaluate the influence of the initial CT blend sign on short-term outcomes in patients with hypertensive ICH after stereotactic minimally invasive surgery (sMIS). Methods We enrolled 242 patients with spontaneous ICH. The patients were assigned to the blend sign group (91 patients) or non-blend sign (control) group (151 patients) based on the initial CT features. The NIHSS, GCS and mRS were used to assess the effects of sMIS. The rates of severe pulmonary infection and cardiac complications were also compared between the two groups. Results Statistically significant differences in the NIHSS and GCS scores were not observed between the blend sign group and the control group. No significant differences in the proportion of patients with good outcomes during the follow-up period were observed between the two groups. A higher rate of re-haemorrhage was noted in the blend sign group. Significant differences in the rates of severe pulmonary infection and cardiac complications were not observed between the two groups. Conclusions The initial CT blend sign is not associated with poor outcomes in patients with hypertensive ICH after sMIS. ICH patients with the CT blend sign should undergo sMIS if they are suitable candidates for surgery.

Haemostatic therapy in spontaneous intracerebral haemorrhage patients with high-risk of haematoma expansion by CT marker: a systematic review and meta-analysis of randomised trials

Background and purposeCurrent randomised controlled trials (RCTs) showed an uncertain benefit of haemostatic therapy on preventing haematoma expansion and improving the outcome in patients with intracerebral haemorrhage (ICH). This meta-analysis aims to systematically evaluate the effect of haemostatic agents on the prevention of haemorrhage growth in patients with high-risk spontaneous ICH predicted by CT signs in RCTs.MethodsA comprehensive search of PubMed, EMBASE and Cochrane library from 1 January 2005 to 30 June 2021 was conducted. RCTs that compared haemostatic agents with placebo for the treatment of spontaneous patients with ICH with high-risk haemorrhage growth were included. The primary endpoint was haematoma expansion at 24 hours. Other major endpoints of interest included 90-day functional outcome and mortality.ResultsThe meta-analysis included four RCTs that randomised 2666 patients with ICH with high-risk haemorrhage growth. Haemostatic therapy reduced the rate of haematoma expansion at a marginally statistically significant level when compared with placebo (OR 0.84; 95% CI 0.70 to 1.00; p=0.051). Subgroup analysis for patients with black hole sign on CT revealed a significant reduction of haematoma expansion with haemostatic therapy (OR 0.61; 95% CI 0.39 to 0.94; p=0.03). However, both the primary analysis and subgroup analyses showed that haemostatic therapy could not reduce the rate of poor functional outcome (modified Rankin Scale >3) or death.ConclusionsHaemostatic therapy showed a marginally significant benefit in reducing early haematoma expansion in patients with high-risk spontaneous ICH predicted by markers on CT scan. However, no significant improvement in functional outcome or reduction of mortality was observed.

Deep learning for automatically predicting early haematoma expansion in Chinese patients

Background and purposeEarly haematoma expansion is determinative in predicting outcome of intracerebral haemorrhage (ICH) patients. The aims of this study are to develop a novel prediction model for haematoma expansion by applying deep learning model and validate its prediction accuracy.MethodsData of this study were obtained from a prospectively enrolled cohort of patients with primary supratentorial ICH from our centre. We developed a deep learning model to predict haematoma expansion and compared its performance with conventional non-contrast CT (NCCT) markers. To evaluate the predictability of this model, it was also compared with a logistic regression model based on haematoma volume or the BAT score.ResultsA total of 266 patients were finally included for analysis, and 74 (27.8%) of them experienced early haematoma expansion. The deep learning model exhibited highest C statistic as 0.80, compared with 0.64, 0.65, 0.51, 0.58 and 0.55 for hypodensities, black hole sign, blend sign, fluid level and irregular shape, respectively. While the C statistics for swirl sign (0.70; p=0.211) and heterogenous density (0.70; p=0.141) were not significantly higher than that of the deep learning model. Moreover, the predictive value for the deep learning model was significantly superior to that of the logistic model of haematoma volume (0.62; p=0.042) and the BAT score (0.65; p=0.042).ConclusionsCompared with the conventional NCCT markers and BAT predictive model, the deep learning algorithm showed superiority for predicting early haematoma expansion in ICH patients.

Admission leukocytosis and its implications on intra cerebral haemorrhage

Background: Intracerebral haemorrhage is one amongst the most common subtype of stroke. It is a catastrophic disease with significant rate of mortality and may lead to severe disabilities. Immediate and effective treatment is a prime requisite of ICH, as rapid mortality occurs within first 24 hours. Definitive diagnosis of ICH is difficult as its symptoms are similar to ischemic stroke. Aim of current investigation was to establish a relationship between intra-cerebral haemorrhage and leukocytosis and to use it as an early tool for detecting haematoma expansion for prognostication and developing newer drugs using a suitable therapeutic target. Methods: Current investigation was an observational study carried out on 100 patients with intra-cerebral haemorrhage. Differential counts were studied with respect to influence of particular subtypes on hematoma expansion. Follow up NCCT was done after 48 hours of the event. Results: Results of present investigation revealed that mean age of the patients was 56 years, 82% were males and all the patients were hypertensive. It was observed that majority of patients with neutrophilic leukocytosis, did not show hematoma expansion and neutrophilic leukocytosis was preferentially present in patients with higher initial bleed volumes. Significant association was observed between monocytosis and haematoma expansion and association between lymphocytosis and volume expansion was observed to be non-significant. Conclusions: Current study findings can aid in early risk stratification and prognostication of ICH patients and can also provide a tool for identification of new therapeutic targets for controlling haematoma expansion.

Initial CT blend sign is not associated with poor outcome in patients following stereotactic minimally invasive surgery

Background: The initial computed tomography (CT) blend sign has been used as an imaging marker to predict haematoma expansion and poor outcomes in patients with a small volume intracerebral haemorrhage (ICH). However, the relationship between the blend sign and outcomes remains elusive. The present study aimed to retrospectively measure the impact of initial CT blend signs on short-term outcomes in patients with hypertensive ICH who underwent stereotactic minimally invasive surgery (sMIS). Methods: We enrolled 242 patients with spontaneous ICH. Based on the initial CT features, the patients were assigned to a blend sign group (91 patients) or a nonblend sign (control) group (151 patients). The NIHSS, GCS and mRS were used to measure the effects of sMIS. The rates of severe pulmonary infection and cardiac complications were also compared between the two groups. Results: No significant differences in NIHSS and GCS scores were observed between the two groups. The proportion of patients with good outcomes during follow-up was not different between the two groups. The rate of rehaemorrhaging increased in the blend sign group. No significant differences in severe pulmonary infections and cardiac complications were noted between the two groups. Conclusions: The initial CT blend sign was not associated with poor outcomes in patients with hypertensive ICH who underwent sMIS. ICH patients with CT blend signs should undergo sMIS if they are suitable candidates for surgery.