Inter-Rater Reliability for Thrombolysis in Cerebral Infarction with TICI 2c Category

A Deep Learning Approach to Predict Recanalization First-Pass Effect following Mechanical Thrombectomy in Patients with Acute Ischemic Stroke

BACKGROUND AND PURPOSE

Following endovascular thrombectomy in patients with large-vessel occlusion stroke, successful recanalization from 1 attempt, known as the first-pass effect, has correlated favorably with long-term outcomes. Pretreatment imaging may contain information that can be used to predict the first-pass effect. Recently, applications of machine learning models have shown promising results in predicting recanalization outcomes, albeit requiring manual segmentation. In this study, we sought to construct completely automated methods using deep learning to predict the first-pass effect from pretreatment CT and MR imaging.

MATERIALS AND METHODS

Our models were developed and evaluated using a cohort of 326 patients who underwent endovascular thrombectomy at UCLA Ronald Reagan Medical Center from 2014 to 2021. We designed a hybrid transformer model with nonlocal and cross-attention modules to predict the first-pass effect on MR imaging and CT series.

RESULTS

The proposed method achieved a mean 0.8506 (SD, 0.0712) for cross-validation receiver operating characteristic area under the curve (ROC-AUC) on MR imaging and 0.8719 (SD, 0.0831) for cross-validation ROC-AUC on CT. When evaluated on the prospective test sets, our proposed model achieved a mean ROC-AUC of 0.7967 (SD, 0.0335) with a mean sensitivity of 0.7286 (SD, 0.1849) and specificity of 0.8462 (SD, 0.1216) for MR imaging and a mean ROC-AUC of 0.8051 (SD, 0.0377) with a mean sensitivity of 0.8615 (SD, 0.1131) and specificity 0.7500 (SD, 0.1054) for CT, respectively, representing the first classification of the first-pass effect from MR imaging alone and the first automated first-pass effect classification method in CT.

CONCLUSIONS

Results illustrate that both nonperfusion MR imaging and CT from admission contain signals that can predict a successful first-pass effect following endovascular thrombectomy using our deep learning methods without requiring time-intensive manual segmentation.

Observation of the Effect of Nursing BPR on Thrombolytic Efficacy and Prognosis of Patients with Cerebral Infarction Based on CT Images

Cerebral infarction has become the main cause of death among Chinese residents, especially ischemic cerebral infarction. The existing CT technology is not very effective for the detection of cerebral infarction, and some angiography has problems such as blurring and shadowing. In order to understand the treatment methods and effects of patients with cerebral infarction, this article observes the effect of nursing BPR on thrombolytic efficacy and prognosis of patients with cerebral infarction based on CT images. The patients were divided into thrombolytic group and nonthrombolytic group, and a simple rating scale was used to assess the motor function of the patients' limbs, and the stroke scale was used to assess the patient's neurological function. Compare the baseline data, the time of admission, 24 hours and 7 days, the scores before and after treatment, and the ratio between the two groups. According to the monitoring, record each time point. The analysis of the occurrence of primary endpoint was events and secondary endpoint events and risk factors affecting limb motor function. The results of the study found that, based on the computer scanning observation of nursing BPR, compared with the traditional model, the patient's bleeding was significantly reduced, and the time required for nursing was also reduced by more than 50% compared with the traditional model. Compared with the traditional nursing model, the satisfaction of patients with the BPR nursing model is nearly 40% higher than that of the traditional nursing model. This shows that the observation of thrombolytic effect in patients with cerebral infarction based on computed tomography and BPR nursing can produce good therapeutic effects.

Successful mechanical thrombectomy in acute ischemic stroke: revascularization grade and functional independence

Most studies define the technical success of endovascular thrombectomy (EVT) as a Thrombolysis in Cerebral Infarction (TICI) revascularization grade of 2b or higher. However, growing evidence suggests that TICI 3 is the best angiographic predictor of improved functional outcomes. To assess the association between successful TICI revascularization grades and functional independence at 90 days, we performed a systematic review and network meta-analysis of thrombectomy studies that reported TICI scores and functional outcomes, measured by the modified Rankin Scale, using the semi-automated AutoLit software platform. Forty studies with 8691 patients were included in the quantitative synthesis. Across TICI, modified TICI (mTICI), and expanded TICI (eTICI), the highest rate of good functional outcomes was observed in patients with TICI 3 recanalization, followed by those with TICI 2c and TICI 2b recanalization, respectively. Rates of good functional outcomes were similar among patients with either TICI 2c or TICI 3 grades. On further sensitivity analysis of the eTICI scale, the rates of good functional outcomes were equivalent between eTICI 2b50 and eTICI 2b67 (OR 0.81, 95% CI 0.52 to 1.25). We conclude that near complete or complete revascularization (TICI 2c/3) is associated with higher rates of functional outcomes after EVT.

Identification of successful cerebral reperfusions (mTICI ≥2b) using an artificial intelligence strategy

BACKGROUND

The modified thrombolysis in cerebral infarction (mTICI) scale is a widely used and validated qualitative tool to evaluate angiographic intracerebral inflow following endovascular thrombectomy (EVT). We validated a machine-learning (ML) algorithm to grade digital subtraction angiograms (DSA) using the mTICI scale.

MATERIALS AND METHODS

We included angiograms of identified middle cerebral artery (MCA) occlusions who underwent EVT. The complete DSA sequences were preprocessed and normalized. We created three convolutional neural networks to classify DSA into two outcomes, low- (mTICI 0,1,2a) and high-grade (mTICI 2b,2c,3).

RESULTS

We included a total of 234 angiograms in this study. The area under the receiver operating characteristic was 0.863 (95% CI 0.816-0.909), 0.914 (95% CI 0.876-0.951), and 0.890 (95% CI 0.848-0.932) for the anteroposterior (AP), lateral (L), and combined models, respectively, when dichotomizing outcomes into low and high grade. The models' area under the precision-recall curve was 0.879 (95% CI 0.829-0.930), 0.906 (95% CI 0.844-0.968), and 0.887 (95% CI 0.834-0.941) for the AP, L, and combined models.

CONCLUSION

In complete cerebral DSA, our angiography-based ML strategy was able to predict mTICI scores following EVT rapidly and reliably for MCA occlusions.

Deep Learning-Based Automated Thrombolysis in Cerebral Infarction Scoring: A Timely Proof-of-Principle Study

Background and Purpose

Mechanical thrombectomy is an established procedure for treatment of acute ischemic stroke. Mechanical thrombectomy success is commonly assessed by the Thrombolysis in Cerebral Infarction (TICI) score, assigned by visual inspection of X-ray digital subtraction angiography data. However, expert-based TICI scoring is highly observer-dependent. This represents a major obstacle for mechanical thrombectomy outcome comparison in, for instance, multicentric clinical studies. Focusing on occlusions of the M1 segment of the middle cerebral artery, the present study aimed to develop a deep learning (DL) solution to automated and, therefore, objective TICI scoring, to evaluate the agreement of DL- and expert-based scoring, and to compare corresponding numbers to published scoring variability of clinical experts.

Methods

The study comprises 2 independent datasets. For DL system training and initial evaluation, an in-house dataset of 491 digital subtraction angiography series and modified TICI scores of 236 patients with M1 occlusions was collected. To test the model generalization capability, an independent external dataset with 95 digital subtraction angiography series was analyzed. Characteristics of the DL system were modeling TICI scoring as ordinal regression, explicit consideration of the temporal image information, integration of physiological knowledge, and modeling of inherent TICI scoring uncertainties.

Results

For the in-house dataset, the DL system yields Cohen’s kappa, overall accuracy, and specific agreement values of 0.61, 71%, and 63% to 84%, respectively, compared with the gold standard: the expert rating. Values slightly drop to 0.52/64%/43% to 87% when the model is, without changes, applied to the external dataset. After model updating, they increase to 0.65/74%/60% to 90%. Literature Cohen’s kappa values for expert-based TICI scoring agreement are in the order of 0.6.

Conclusions

The agreement of DL- and expert-based modified TICI scores in the range of published interobserver variability of clinical experts highlights the potential of the proposed DL solution to automated TICI scoring.

A Machine Learning Approach to Predict Acute Ischemic Stroke Thrombectomy Reperfusion using Discriminative MR Image Features

Mechanical thrombectomy (MTB) is one of the two standard treatment options for Acute Ischemic Stroke (AIS) patients. Current clinical guidelines instruct the use of pretreatment imaging to characterize a patient's cerebrovascular flow, as there are many factors that may underlie a patient's successful response to treatment. There is a critical need to leverage pretreatment imaging, taken at admission, to guide potential treatment avenues in an automated fashion. The aim of this study is to develop and validate a fully automated machine learning algorithm to predict the final modified thrombolysis in cerebral infarction (mTICI) score following MTB. A total 321 radiomics features were computed from segmented pretreatment MRI scans for 141 patients. Successful recanalization was defined as mTICI score >= 2c. Different feature selection methods and classification models were examined in this study. Our best performance model achieved 74.42±2.52% AUC, 75.56±4.44% sensitivity, and 76.75±4.55% specificity, showing a good prediction of reperfusion quality using pretreatment MRI. Results suggest that MR images can be informative to predicting patient response to MTB, and further validation with a larger cohort can determine the clinical utility.

Interobserver Agreement in Scoring Angiographic Results of Basilar Artery Occlusion Stroke Therapy

BACKGROUND AND PURPOSE

The modified TICI Infarction grading system is a metric currently used to evaluate angiographic results of thrombectomy for large-vessel occlusion in ischemic stroke. Originally designed for evaluating MCA territories, it is currently used for other vessel occlusions, including the posterior circulation. We postulate that the modified TICI use for the posterior circulation is not accurate due to the different vascular territories supplied by vertebrobasilar vasculature, making grading more complex.

MATERIALS AND METHODS

We collected angiographic results from 30 patients who presented with acute posterior circulation occlusions between 2015 and 2018 and underwent thrombectomy in our institution. Eight observers were asked to evaluate the TICI scores before and after thrombectomy. The multirater statistics were computed using Fleiss κ analysis. Further data were collected regarding the potential brain territories at risk and the existence of atherosclerotic disease in the basilar artery.

RESULTS

The overall agreement κ reached 0.277 (SD, 0.013), which suggests a "fair" agreement among the raters. On average, 45% of observers achieved a high accuracy in predicting brain areas at risk of ischemia. As for the existence of basilar atherosclerotic disease, a high agreement (defined as at least 5 of 6 observers) was seen in 20 of the 30 patients.

CONCLUSIONS

Despite TICI being ubiquitous in stroke diagnostics, the high variability of posterior circulation TICI scores calls into question its use in these strokes. Other methods should be developed to assess recanalization in the posterior circulation.

Defining reperfusion post endovascular therapy in ischemic stroke using MR-dynamic contrast enhanced perfusion

OBJECTIVES

Cerebral blood flow (CBF) measurements after endovascular therapy (EVT) for acute ischemic stroke are important to distinguish early secondary injury related to persisting ischemia from that related to reperfusion when considering clinical response and infarct growth.

METHODS

We compare reperfusion quantified by the modified Thrombolysis in Cerebral Infarction Score (mTICI) with perfusion measured by MRI dynamic contrast-enhanced perfusion within 5 h of EVT anterior circulation stroke. MR perfusion (rCBF, rCBV, rTmax, rT0) and mTICI scores were included in a predictive model for change in NIHSS at 24 h and diffusion-weighted imaging (DWI) lesion growth (acute to 24 h MRI) using a machine learning RRELIEFF feature selection coupled with a support vector regression.

RESULTS

For all perfusion parameters, mean values within the acute infarct for the TICI-2b group (considered clinically good reperfusion) were not significantly different from those in the mTICI <2b (clinically poor reperfusion). However, there was a statistically significant difference in perfusion values within the acute infarct region of interest between the mTICI-3 group versus both mTICI-2b and <2b (p = 0.02). The features that made up the best predictive model for change in NIHSS and absolute DWI lesion volume change was rT0 within acute infarct ROI and admission CTA collaterals respectively. No other variables, including mTICI scores, were selected for these best models. The correlation coefficients (Root mean squared error) for the cross-validation were 0.47 (13.7) and 0.51 (5.7) for change in NIHSS and absolute DWI lesion volume change.

CONCLUSION

MR perfusion following EVT provides accurate physiological approach to understanding the relationship of CBF, clinical outcome, and DWI growth.

ADVANCES IN KNOWLEDGE

MR perfusion CBF acquired is a robust, objective reperfusion measurement providing following recanalization of the target occlusion which is critical to distinguish potential therapeutic harm from the failed technical success of EVT as well as improve the responsiveness of clinical trial outcomes to disease modification.

Reliability of the Modified TICI Score among Endovascular Neurosurgeons

BACKGROUND AND PURPOSE

The modified TICI score is the benchmark for quantifying reperfusion after mechanical thrombectomy. There has been limited investigation into the reliability of this score. We aim to identify intra-rater and inter-rater reliability of the mTICI score among endovascular neurosurgeons.

MATERIALS AND METHODS

Four independent endovascular neurosurgeons (raters) reviewed angiograms of 67 patients at 2 time points. κ statistics assessed inter- and intrarater reliability and compared raters'-versus-proceduralists' scores. Reliability was also assessed for occlusion location and by dichotomizing modified TICI scores (0-2a versus 2b-3).

RESULTS

Interrater reliability was moderate-to-substantial, weighted κ = 0.417-0.703, overall κ = 0.374 (P < .001). The dichotomized modified TICI score had moderate-to-substantial interrater agreement, κ statistics = 0.468-0.715, overall κ = 0.582 (P < .001). Intrarater reliability was moderate-to-almost perfect, weighted κ = 0.594-0.81. The dichotomized modified TICI score had substantial-to-almost perfect reliability, κ = 0.632-0.82. Proceduralists had fair-to-moderate agreement with raters, weighted κ = 0.348-0.574, and the dichotomized modified TICI score had fair-to-moderate agreement, κ = 0.365-0.544. When proceduralists and raters disagreed, proceduralists' scores were higher in 79.6% of cases. M1 followed by ICA occlusions had the highest agreement.

CONCLUSIONS

The modified TICI score is a practical metric for assessing reperfusion after mechanical thrombectomy, though not without limitations. Agreement improved when scores were dichotomized around the clinically relevant threshold of successful revascularization. Interrater reliability improved with time, suggesting that formal training of interventionalists may improve reporting reliability. Agreement of the modified TICI scale is best with M1 and ICA occlusion and becomes less reliable with more distal or posterior circulation occlusions. These findings should be considered when developing research trials.

Operator Versus Core Lab Adjudication of Reperfusion After Endovascular Treatment of Acute Ischemic Stroke

Background and Purpose- The modified Treatment In Cerebral Ischemia (mTICI) score is the standard method to quantify the degree of reperfusion after endovascular treatment in acute ischemic stroke. In clinical practice, it is commonly assessed by local operators after the procedure. In clinical trials and registries, mTICI is evaluated by an imaging core lab. The aim of this study was to compare operator mTICI with core lab mTICI scores in patients included in the MR CLEAN (Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands) Registry. Methods- All patients with an intracranial carotid or middle cerebral artery occlusion with anteroposterior and lateral digital subtraction angiography runs were included. Operators determined the mTICI score immediately after endovascular treatment. Core lab neuroradiologists were blinded to clinical characteristics and assessed mTICI scores based on pre- and postintervention digital subtraction angiography. The agreement between operator and core lab mTICI scores and their value in the prediction of outcome (score on modified Rankin Scale at 90 days) was determined. Results- In total, 1130 patients were included. The proportion of agreement between operator and core lab mTICI score was 56% (95% CI, 54%-59%). In 33% (95% CI, 31%-36%), mTICI was overestimated by operators. Operators reported a higher rate of successful reperfusion than the core lab (77% versus 67%; difference 10% [95% CI, 6%-14%]; P<0.001). In 252 (33%) of 763 patients scored as incomplete reperfusion by the core lab (mTICI <3), the local read was mTICI 3. Multivariable logistic regression models containing either core lab scored or operator scored successful reperfusion predicted outcome on the full (C statistic of both models: 0.76) or dichotomized modified Rankin Scale (modified Rankin Scale, 0-2; C statistic of both models: 0.83) equally well. Conclusions- Operators tend to overestimate the degree of reperfusion compared with the core lab although this does not affect the accuracy of outcome prediction.

Validation of the extended thrombolysis in cerebral infarction score in a real world cohort

Background

A thrombolysis in cerebral infarction (TICI) score of 2b is defined as a good recanalization result although the reperfusion may only cover 50% of the affected territory. An additional mTICI2c category was introduced to further differentiate between mTICI scores. Despite the new mTICI2c category, mTICI2b still covers a range of 50–90% reperfusion which might be too imprecise to predict neurological improvement after therapy.

Aim

To compare the 7-point “expanded TICI” (eTICI) scale with the traditional mTICI in regard to predict functional independence at 90 days.

Methods

Retrospective review of 225 patients with large artery occlusion. Angiograms were graded by 2 readers according the 7-point eTICI score (0% = eTICI0; reduced clot = eTICI1; 1–49% = eTICI2a, 50–66% = eTICI2b50; 67–89% = eTICI2b67, 90–99% = eTICI2c and complete reperfusion = eTICI3) and the conventional mTICI score. The ability of e- and mTICI to predict favorable outcome at 90days was compared.

Results

Given the ROC analysis eTICI was the better predictor of favorable outcome (p-value 0.047). Additionally, eTICI scores 2b50, 2b67 and 2c (former mTICI2b) were significantly superior at predicting the probability of a favorable outcome at 90 days after endovascular therapy with a p-value of 0.033 (probabilities of 17% for mTICI2b50, 24% for mTICI2b67 and 54% for mTICI2c vs. 36% for mTICI2b).

Conclusions

The 7-point eTICI allows for a more accurate outcome prediction compared to the mTICI score because it refines the broad range of former mTICI2b results.

eTICI reperfusion: defining success in endovascular stroke therapy

Background

Revascularization after endovascular therapy for acute ischemic stroke is measured by the Thrombolysis In Cerebral Infarction (TICI) scale, yet variability exists in scale definitions. We examined the degree of reperfusion with the expanded TICI (eTICI) scale and association with outcomes in the HERMES collaboration of recent endovascular trials.

Methods

The HERMES Imaging Core, blind to all other data, evaluated angiography after endovascular therapy in HERMES. A battery of TICI scores (mTICI, TICI, TICI2C) was used to define reperfusion of the initial target occlusion defined by non-invasive imaging and conventional angiography.

Results

Angiography of 801 subjects was available, including 797 defined by non-invasive imaging (154 internal carotid artery (ICA), 583 M1, 60 M2) and 748 by conventional angiography (195 ICA, 459 M1, 94 M2). Among 729 subjects in whom the reperfusion grade could be established, using eTICI (3=100%, 2C=90–99%, 2b67=67–89%, 2b50=50–66%) of the conventional angiography target occlusion, there were 63 eTICI 3 (9%), 166 eTICI 2c (23%), 218 eTICI 2b67 (30%), 103 eTICI 2b50 (14%), 100 eTICI 2a (14%), 19 eTICI 1 (3%), and 60 eTICI 0 (8%). Modified Rankin Scale shift analyses from baseline to 90 days showed that increasing TICI grades were linked with better outcomes, with significant distinctions between TICI 0/1 versus 2a (p=0.028), 2a versus 2b50 (p=0.017), and 2b50 versus 2b67 (p=0.014).

Conclusions

The benefit of endovascular therapy in HERMES was strongly associated with increasing degrees of reperfusion defined by eTICI. The eTICI metric identified meaningful distinctions in clinical outcomes and may be used in future studies and routine practice.

Systematic review and meta-analysis on outcome differences among patients with TICI2b versus TICI3 reperfusions: success revisited

OBJECTIVE

A reperfusion quality of thrombolysis in cerebral infarction (TICI)≥2b has been set as the therapeutic angiography target for interventions in patients with acute ischaemic stroke. This study addresses whether the distinction between TICI2b and TICI3 reperfusions shows a clinically relevant difference on functional outcome.

METHODS

A systematic literature review and meta-analysis was carried out and presented in conformity with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses criteria to test the primary hypothesis that TICI2b and TICI3 reperfusions are associated with different rates of modified Rankin Scale (mRS) ≤2 at day 90. Secondary endpoints included rates of haemorrhagic transformations, mortality and excellent functional outcome (mRS ≤1). Summary estimates of ORs (sOR) with 95% CI were calculated using the inverse variance heterogeneity model accounting for multiple true effect sizes.

RESULTS

Fourteen studies on 2379 successfully reperfused patients were included (1131 TICI3, 1248 TICI2b). TICI3 reperfusions were associated with higher rates of functional independence (1.74, 95% CI 1.44 to 2.10) and excellent functional outcomes (2.01, 95% CI 1.60 to 2.53), also after including adjusted estimates. The safety profile of patients with TICI3 was superior, as demonstrated by lower rates of mortality (sOR 0.59, 95% CI 0.37 to 0.92) and symptomatic intracranial haemorrhages (sOR 0.42, 95% CI 0.25 to 0.71).

CONCLUSION

TICI3 reperfusions are associated with superior outcome and better safety profiles than TICI2b reperfusions. This effect seems to be independent of time and collaterals. As reperfusion quality is the most important modifiable predictor of patients' outcome, a more conservative definition of successful therapy and further evaluation of treatment approaches geared towards achieving TICI3 reperfusions are desirable.