RESUMEN
Background In acute ischemic stroke (AIS), fluid-attenuated inversion recovery (FLAIR) is used for treatment decisions when onset time is unknown. Synthetic FLAIR could be generated with deep learning from information embedded in diffusion-weighted imaging (DWI) and could replace acquired FLAIR sequence (real FLAIR) and shorten MRI duration. Purpose To compare performance of synthetic and real FLAIR for DWI-FLAIR mismatch estimation and identification of patients presenting within 4.5 hours from symptom onset. Materials and Methods In this retrospective study, all pretreatment and early follow-up (<48 hours after symptom onset) MRI data sets including DWI (b = 0-1000 sec/mm2) and FLAIR sequences obtained in consecutive patients with AIS referred for reperfusion therapies between January 2002 and May 2019 were included. On the training set (80%), a generative adversarial network was trained to produce synthetic FLAIR with DWI as input. On the test set (20%), synthetic FLAIR was computed without real FLAIR knowledge. The DWI-FLAIR mismatch was evaluated on both FLAIR data sets by four independent readers. Interobserver reproducibility and DWI-FLAIR mismatch concordance between synthetic and real FLAIR were evaluated with κ statistics. Sensitivity and specificity for identification of AIS within 4.5 hours were compared in patients with known onset time by using McNemar test. Results The study included 1416 MRI scans (861 patients; median age, 71 years [interquartile range, 57-81 years]; 375 men), yielding 1134 and 282 scans for training and test sets, respectively. Regarding DWI-FLAIR mismatch, interobserver reproducibility was substantial for real and synthetic FLAIR (κ = 0.80 [95% CI: 0.74, 0.87] and 0.80 [95% CI: 0.74, 0.87], respectively). After consensus, concordance between real and synthetic FLAIR was almost perfect (κ = 0.88; 95% CI: 0.82, 0.93). Diagnostic value for identifying AIS within 4.5 hours did not differ between real and synthetic FLAIR (sensitivity: 107 of 131 [82%] vs 111 of 131 [85%], P = .2; specificity: 96 of 104 [92%] vs 96 of 104 [92%], respectively, P > .99). Conclusion Synthetic fluid-attenuated inversion recovery (FLAIR) had diagnostic performances similar to real FLAIR in depicting diffusion-weighted imaging-FLAIR mismatch and in helping to identify early acute ischemic stroke, and it may accelerate MRI protocols. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Carroll and Hurley in this issue.
Asunto(s)
Isquemia Encefálica , Accidente Cerebrovascular Isquémico , Accidente Cerebrovascular , Anciano , Isquemia Encefálica/diagnóstico por imagen , Isquemia Encefálica/terapia , Imagen de Difusión por Resonancia Magnética/métodos , Femenino , Humanos , Accidente Cerebrovascular Isquémico/diagnóstico por imagen , Imagen por Resonancia Magnética/métodos , Masculino , Reproducibilidad de los Resultados , Estudios Retrospectivos , Accidente Cerebrovascular/terapia , Factores de TiempoRESUMEN
BACKGROUND: Tandem basilar artery occlusions (tBAO) are defined as concomitant basilar artery and extracranial dominant vertebral artery occlusions. The prognosis of such tBAO treated by mechanical thrombectomy (MT) has been scantly reported. The purpose of our study was to compare the safety and effectiveness of MT for patients with tBAO compared to those with non-tandem basilar artery occlusions (ntBAO). PATIENTS AND METHODS: Retrospective analysis of a database of patients who underwent MT at two academic centres. All patients treated for BAO were retrieved. Patients with tBAO, defined as a concomitant BAO and extracranial vertebral artery (VA) occlusion or severe stenosis ≥70% (V1 or proximal V2 segment) were compared with patients with ntBAO. RESULTS: A total of 15 patients with tBAO and 74 patients with ntBAO were enrolled. Successful reperfusion (modified thrombolysis in cerebral infarction score ≥2b) was obtained in 73.3% versus 90.5% (ORâ¯=â¯0.29, 95%CI: 0.07-1.15), good clinical outcome (3-months modified Rankin scale ≤2) was reached by 26.7% versus 32.4% (ORâ¯=â¯0.76; 95% CI: 0.24-2.63) and mortality at 3-months was 46.7% versus 31% (ORâ¯=â¯1.94; 95%CI: 0.63-6) of patients with tBAO versus ntBAO, respectively. Two patients (13.3%) with tBAO and three (4%) with ntBAO had symptomatic intracranial haemorrhage (ORâ¯=â¯3.64; 95% CI: 0.55-24). CONCLUSION: Mechanical thrombectomy for patients with tandem basilar artery occlusion tends to be associated with lower rates of successful reperfusion and good clinical outcome, and higher rate of mortality. Larger multicentre studies are warranted to better precise the proper selection and management of such patients.
Asunto(s)
Trombectomía/métodos , Insuficiencia Vertebrobasilar/cirugía , Anciano , Angiografía de Substracción Digital , Arteria Basilar , Angiografía Cerebral , Femenino , Humanos , Masculino , Persona de Mediana Edad , Estudios Retrospectivos , Arteria Vertebral , Insuficiencia Vertebrobasilar/diagnóstico por imagen , Insuficiencia Vertebrobasilar/mortalidadRESUMEN
The reported incidence of persistent hypoperfusion despite complete recanalization as surrogate for impaired microvascular reperfusion (IMR) has varied widely among clinical studies, possibly due to differences in i) definition of complete recanalization, with only recent Thrombolysis in Cerebral Infarction (TICI) grading schemes allowing distinction between complete (TICI3) and partial recanalization with distal occlusions (TICI2c); ii) operational definition of IMR; and iii) consideration of potential alternative causes for hypoperfusion, notably carotid stenosis, re-occlusion and post-thrombectomy hemorrhage. We performed a systematic review to identify clinical studies that carried out brain perfusion imaging within 72 hrs post-thrombectomy for anterior circulation stroke and reported hypoperfusion rates separately for TICI3 and TICI2c grades. Authors were contacted if this data was missing. We identified eight eligible articles, altogether reporting 636 patients. The incidence of IMR after complete recanalization (i.e., TICI3) tended to decrease with the number of considered alternative causes of hypoperfusion: range 12.5-42.9%, 0-31.6% and 0-9.1% in articles that considered none, two or all three causes, respectively. No study reported the impact of IMR on functional outcome separately for TICI-3 patients. Based on this systematic review, IMR in true complete recanalization appears relatively rare, and reported incidence highly depends on definition used and consideration of confounding factors.
Asunto(s)
Isquemia Encefálica , Accidente Cerebrovascular , Humanos , Incidencia , Resultado del Tratamiento , Estudios Retrospectivos , Infarto Cerebral/etiología , Trombectomía/métodos , Reperfusión , Isquemia Encefálica/etiologíaRESUMEN
PURPOSE: To evaluate performance of synthetic and real FLAIR for identifying early stroke in a multicenter cohort. METHODS: This retrospective study was conducted using DWI and FLAIR extracted from the Endovascular Treatment in Ischemic Stroke image registry (2017-2021). The database was partitioned into subsets according to MRI field strength and manufacturer, and randomly divided into training set (70%) used for model fine-tuning, validation set (15%), and test set (15%). In test set, five readers, blinded to FLAIR sequence type, assessed DWI-FLAIR mismatch using real and synthetic FLAIR. Interobserver agreement for DWI-FLAIR rating and concordance between synthetic and real FLAIR were evaluated with kappa statistics. Sensitivity and specificity for identification of ⩽4.5 h AIS were compared in patients with known onset-to-MRI delay using McNemar's test. RESULTS: 1454 complete MRI sets (1172 patients, median (IQR) age: 73 years (62-82); 762 women) acquired on 125 MRI units were analyzed. In test set (207 MRI), interobserver reproducibility for DWI-FLAIR mismatch labeling was substantial for real and synthetic FLAIR (Fleiss κ = 0.79 (95%CI: 0.73-0.84) and 0.77 (95%CI: 0.71-0.82), respectively). After consensus, concordance between real and synthetic FLAIR was excellent (κ = 0.85 (95%CI: 0.78-0.92)). In 141 MRI sets with known onset-to-MRI delay, diagnostic performances for ⩽4.5 h AIS identification did not differ between real and synthetic FLAIR (sensitivity: 60/71 (85%) vs 59/71 (83%), p = .56; specificity: 65/70 (93%) vs 65/70 (93%), p > 0.99). CONCLUSION: A deep-learning-based FLAIR fine-tuned on multicenter data can provide comparable performances to real FLAIR for early AIS identification. This approach may help reducing MR protocol duration and motion artifacts.
RESUMEN
BACKGROUND: Recent studies in the general stroke population treated with endovascular treatment (EVT) reported that higher pre-treatment lesional volumes were independently associated with poor neurological outcome and functional dependence after stroke. However, it has been not evaluated in older patients. AIM: We test the association between the pre-treatment lesional volume on diffusion-weighted magnetic resonance imaging and relevant outcome measures in older adults with stroke treated with EVT. METHODS: We included consecutive older adults with stroke (⩾80 years old) treated with EVT in two academic comprehensive stroke centers. The association between pre-treatment lesional volume and relevant outcome measures (poor outcome (modified Rankin scale 4-6), 3-month mortality and symptomatic intracerebral hemorrhage (sICH)) was evaluated using univariate and multivariable models. RESULTS: Five hundred seventy-nine patients were included (mean age: 85.6 ± 4.1, median lesional volume was 10 ml; interquartile range: 3-30 ml). Pre-treatment lesional volume was associated with poor functional outcome (adjusted odds ratio (aOR): 1.87, 95% confidence interval (CI): 1.60-2.20, for +1 logarithmic increase of lesional volume), 3-month mortality (aOR: 1.50, CI: 1.28-1.76), and sICH (aOR: 1.67, CI: 1.27-2.20). A threshold lesional volume >35 ml predicted 90% of patients with poor functional outcome and a cut-off >51 ml predicted 90% of patients dead at 3 months. CONCLUSIONS: Pre-treatment lesional volume might contribute, in association with other relevant clinical features, to the selection of older stroke patients who will benefit from EVT.
Asunto(s)
Isquemia Encefálica , Procedimientos Endovasculares , Accidente Cerebrovascular , Humanos , Anciano , Anciano de 80 o más Años , Isquemia Encefálica/terapia , Procedimientos Endovasculares/métodos , Resultado del Tratamiento , Sistema de Registros , Infarto Cerebral/etiologíaRESUMEN
Machine Learning (ML) has been proposed for tissue fate prediction after acute ischemic stroke (AIS), with the aim to help treatment decision and patient management. We compared three different ML models to the clinical method based on diffusion-perfusion thresholding for the voxel-based prediction of final infarct, using a large MRI dataset obtained in a cohort of AIS patients prior to recanalization treatment. Baseline MRI (MRI0), including diffusion-weighted sequence (DWI) and Tmax maps from perfusion-weighted sequence, and 24-hr follow-up MRI (MRI24h) were retrospectively collected in consecutive 394 patients AIS patients (median age = 70 years; final infarct volume = 28mL). Manually segmented DWI24h lesion was considered the final infarct. Gradient Boosting, Random Forests and U-Net were trained using DWI, apparent diffusion coefficient (ADC) and Tmax maps on MRI0 as inputs to predict final infarct. Tissue outcome predictions were compared to final infarct using Dice score. Gradient Boosting had significantly better predictive performance (median [IQR] Dice Score as for median age, maybe you can replace the comma with an equal sign for consistency 0.53 [0.29-0.68]) than U-Net (0.48 [0.18-0.68]), Random Forests (0.51 [0.27-0.66]), and clinical thresholding method (0.45 [0.25-0.62]) (P < 0.001). In this benchmark of ML models for tissue outcome prediction in AIS, Gradient Boosting outperformed other ML models and clinical thresholding method and is thus promising for future decision-making.
Asunto(s)
Encéfalo/irrigación sanguínea , Encéfalo/diagnóstico por imagen , Imagen de Difusión por Resonancia Magnética/métodos , Infarto/diagnóstico por imagen , Accidente Cerebrovascular Isquémico/diagnóstico por imagen , Aprendizaje Automático/estadística & datos numéricos , Anciano , Anciano de 80 o más Años , Encéfalo/patología , Toma de Decisiones Clínicas , Femenino , Estudios de Seguimiento , Humanos , Infarto/patología , Accidente Cerebrovascular Isquémico/patología , Accidente Cerebrovascular Isquémico/terapia , Masculino , Persona de Mediana Edad , Valor Predictivo de las Pruebas , Pronóstico , Reperfusión/métodos , Estudios RetrospectivosRESUMEN
Despite early thrombectomy, a sizeable fraction of acute stroke patients with large vessel occlusion have poor outcome. The no-reflow phenomenon, i.e. impaired microvascular reperfusion despite complete recanalization, may contribute to such "futile recanalizations". Although well reported in animal models, no-reflow is still poorly characterized in man. From a large prospective thrombectomy database, we included all patients with intracranial proximal occlusion, complete recanalization (modified thrombolysis in cerebral infarction score 2c-3), and availability of both baseline and 24 h follow-up MRI including arterial spin labeling perfusion mapping. No-reflow was operationally defined as i) hypoperfusion ≥40% relative to contralateral homologous region, assessed with both visual (two independent investigators) and automatic image analysis, and ii) infarction on follow-up MRI. Thirty-three patients were eligible (median age: 70 years, NIHSS: 18, and stroke onset-to-recanalization delay: 208 min). The operational criteria were met in one patient only, consistently with the visual and automatic analyses. This patient recanalized 160 min after stroke onset and had excellent functional outcome. In our cohort of patients with complete and stable recanalization following thrombectomy for intracranial proximal occlusion, severe ipsilateral hypoperfusion on follow-up imaging associated with newly developed infarction was a rare occurrence. Thus, no-reflow may be infrequent in human stroke and may not substantially contribute to futile recanalizations.
Asunto(s)
Circulación Cerebrovascular/fisiología , Accidente Cerebrovascular/cirugía , Trombectomía/métodos , Anciano , Femenino , Humanos , Masculino , Persona de Mediana Edad , Estudios Retrospectivos , Resultado del TratamientoRESUMEN
The benefits of mechanical thrombectomy (MT) for patients with acute ischemic stroke (AIS) and a large ischemic core (LIC) at presentation are uncertain. We aimed to obtain up-to-date aggregate estimates of the outcomes following MT in patients with volumetrically assessed LIC. We conducted a Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA)-conformed, PROSPERO-registered, systematic review and meta-analysis of studies that included patients with AIS and a baseline LIC treated with MT, reported ischemic core volume quantitatively, and included patients with a LIC defined as a core volume ≥50 mL. The search was restricted to studies published between January 2015 and June 2020. Random-effects-meta-analysis was used to assess the effect of MT on 90-day unfavorable outcome (i.e., modified Rankin Scale [mRS] 3-6), mortality, and symptomatic intracranial hemorrhage (sICH) occurrence. Sensitivity analyses were performed for imaging-modality (computed tomography-perfusion or magnetic resonance-diffusion weighted imaging) and LIC-definition (≥50 or ≥70 mL). We analyzed 10 studies (954 patients), including six (682 patients) with a control group, allowing to compare 332 patients with MT to 350 who received best-medical-management alone. Overall, after MT the rate of patients with mRS 3-6 at 90 days was 74% (99% confidence interval [CI], 67 to 84; Z-value=7.04; I2=92.3%) and the rate of 90-day mortality was 36% (99% CI, 33 to 40; Z-value=-7.07; I2=74.5). Receiving MT was associated with a significant decrease in mRS 3-6 odds ratio (OR) 0.19 (99% CI, 0.11 to 0.33; P<0.01; Z-value=-5.92; I2=62.56) and in mortality OR 0.60 (99% CI, 0.34 to 1.06; P=0.02; Z-value=-2.30; I2=58.72). Treatment group did not influence the proportion of patients experiencing sICH, OR 0.96 (99% CI, 0.2 to 1.49; P=0.54; Z-value=-0.63; I2=64.74). Neither imaging modality for core assessment, nor LIC definition influenced the aggregated outcomes. Using aggregate estimates, MT appeared to decrease the risk of unfavorable functional outcome in patients with a LIC assessed volumetrically at baseline.
RESUMEN
BACKGROUND AND PURPOSE: A long clot, defined by a low (0-6) clot burden score (CBS) assessed by T2*-MR sequence, is associated with worse clinical outcome after intravenous thrombolysis (IVT) for acute ischemic stroke than is a small clot (CBS, 7-10). The added benefit of mechanical thrombectomy (MT) might be higher in patients with long clot. The aim of this pre-specified post hoc analysis of the THRombectomie des Artères CErebrales (THRACE) trial was to assess the association between T2*-CBS, successful recanalization and clinical outcome. METHODS: Of 414 patients randomized in the THRACE trial, 281 patients were included in this analysis. Associations between T2*-CBS and clinical outcome on the modified Rankin Scale (mRS) at 3 months were tested. RESULTS: High T2*-CBS, i.e., small clot, was associated with a shift toward better outcome on the mRS; proportional odds ratio (POR) per point CBS was 1.19 (95% confidence interval [CI], 1.05 to 1.34) in the whole population, 1.34 (95% CI, 1.13 to 1.59) in IVT group, and 1.04 (95% CI, 0.87 to 1.23) in IVTMT group. After adjustment for baseline prognostic variables, the effect of the full scale T2*-CBS was not statistically significant in the whole population and for the IVTMT group but remains significant for the IVT group (POR, 1.32; 95% CI, 1.11 to 1.58). CONCLUSION: s A small clot, as assessed using T2*-CBS, is associated with improved outcome and may be used as a prognostic marker. Despite the worst outcome with long clot, the relative benefit of MT over IVT seemed to increase with low T2*-CBS and longer clot.
RESUMEN
BACKGROUND: Susceptibility vessel sign (SVS) can be a useful MRI biomarker of an occlusion but its relationship with clinical outcomes of acute ischemic stroke (AIS) is yet to be fully elucidated. OBJECTIVE: To investigate SVS in relation to the clinical outcomes after mechanical thrombectomy using a stent retriever (SR) as first-line approach in patients with AIS. MATERIAL AND METHODS: We included patients with a first-line SR approach for anterior AIS from the the Contact Aspiration vs Stent Retriever for Successful Revascularization (ASTER) and THRombectomie des Artères CErebrales (THRACE) trials when both baseline imaging of SVS and 90-day modified Rankin Scale (mRS) scores were available. Patients were assigned to two groups based on the presence of an SVS (independent core laboratory), and the overall distributions of the mRS score at 90 days (shift analysis) and clinical independence (mRS score ≤2) were compared. RESULTS: 217 patients were included and SVS was diagnosed in 76.0% of cases (n=165, 95% CI 70.4% to 81.7%). After adjustment for potential confounders, SVS+ was significantly associated with 90-day mRS improvement (adjusted common OR=2.75; 95% CI 1.44 to 5.26) and favorable outcome (adjusted common OR=2.76; 95% CI 1.18 to 6.45). CONCLUSION: Based on results for patients of the ASTER and THRACE trials receiving first-line SR treatment, SVS was associated with lower disability at 3 months. Large prospective studies using MRI-based thrombus evaluation are warranted.