Abstract
Objectives
This study aimed to evaluate the factors associated with decreasing diffusion-weighted imaging (DWI) positive areas in patients with large early ischemic changes after mechanical thrombectomy (MT).
Materials and Methods
This retrospective single-center clinical study was conducted between January 2013 and December 2022. We included consecutive patients who underwent MT for acute large-vessel occlusion of the anterior circulation with low pretreatment DWI-Alberta Stroke Program Early Computed Tomography Scores (ASPECTS) (0–5), effective recanalization [thrombolysis in cerebral infarction (TICI) 2b or TICI3], and magnetic resonance imaging (MRI) acquired before and after MT. We measured the DWI-positive area volume before and after MT. The primary endpoint was the after/before-MT DWI-positive area–volume ratio.
Results
In total, 28 patients were included in this study. Eight patients (29%) had an after/before-MT DWI-positive area–volume ratio of <1. The median mean apparent diffusion coefficient (ADC) levels of the DWI-positive areas in the groups with a ratio of <1 or >1 were 717 × 10⁶ mm2/s and 637 × 106 mm2/s, respectively (p = 0.011). Multivariate logistic regression analysis showed that ADC level (OR, 1.020 [95% confidence intervals (CIs), 1.001–1.040]; p = 0.040) was an independent predictor of a decreased DWI-positive area after MT. There was a negative correlation between the mean ADC level and the after/before-MT DWI-positive area–volume ratio (p < 0.001, |ρ| = 0.650), and the mean pretreatment ADC cutoff level was 649 × 106 mm2/s (area under the curve (AUC) = 0.806) for predicting a volume ratio of <1.
Conclusions
The mean ADC level before-MT correlated with the after/before-MT DWI-positive area–volume ratio. A mean pretreatment ADC cutoff level of 649 × 106 mm2/s predicted a decreased DWI-positive area after MT.
Keywords: Large early ischemic changes, diffusion-weighted imaging, mean ADC level, mechanical thrombectomy
Introduction
A beneficial effect of mechanical thrombectomy (MT) has been shown in patients with acute ischemic stroke caused by large-vessel occlusion who had a large early ischemic change.1–3 However, the effect of MT was considered to be worse for patients with large early ischemic changes than for patients without. 4 Diffusion-weighted imaging (DWI) has been established as a useful modality for diagnosing acute ischemic stroke.5,6 Therefore, large early ischemic changes are typically accompanied by large DWI-positive areas. In patients with large early ischemic changes, DWI-positive areas after MT usually result in residual or enlarged areas even after effective recanalization; however, DWI-positive areas sometimes decrease.7,8 The DWI reversal volume at recanalization in patients with large early ischemic changes was previously found to be greater in patients with good outcomes than in those with poor outcomes.7,8 Therefore, analyzing the factors associated with DWI reversal volume in patients with large early ischemic changes might be useful in predicting the outcomes of these patients. This study aimed to evaluate factors associated with decreasing DWI-positive area volume after MT in patients with large early ischemic changes.
Material and methods
Patient population
This retrospective single-center clinical study was conducted between January 2013 and December 2022. We analyzed consecutive patients who underwent MT for acute large-vessel occlusion of the anterior circulation at our hospital. We used an intravenous recombinant tissue plasminogen activator (rtPA), as indicated. The indication of MT was determined by each operator. The inclusion criteria were as follows: (a) pretreatment low DWI-Alberta Stroke Program Early CT Score (DWI-ASPECTS: 0–5), (b) effective recanalization [thrombolysis in cerebral infarction (TICI) grade 2b or TICI grade 3] by MT, (c) magnetic resonance imaging (MRI) performed before and after MT, and (d) clinical follow-up with a 90-day modified Rankin Scale (mRS). We defined a 90-day mRS ≤ 3 as a good outcome. The National Institutes of Health Stroke Scale (NIHSS) was used to record evaluations before and 24 hours after MT. A follow-up MRI was performed approximately 24 hours after MT. The volume of the DWI-positive area was measured before and after MT, and the after/before-MT DWI-positive area–volume ratio was the primary endpoint. Age, sex, occluded vessel, mean apparent diffusion coefficient (ADC) level of all DWI-positive areas, recanalization status, use of the rtPA or not, time to recanalization, and outcome were analyzed to determine the relationship between after/before-MT DWI-positive area–volume ratio.
Mechanical thrombectomy
All procedures were performed on the patient under local anesthesia. We performed stent retriever thrombectomy (SRT) using a direct aspiration first-pass technique (ADAPT) or a combination of SRT and ADAPT. Treatment strategies and device selection were determined by each operator.
MRI protocol and data analysis
MRI was performed without contrast medium on a 1.5-T or 3.0-T MR system (MAGNETOM Symphony, Siemens Healthcare, Erlangen, Germany). The slice thicknesses of the DWI and ADC maps were 5 mm, with approximately 24 slices for 120-mm imaging. The following parameters were used to perform DW: repetition time, 4000–5000 ms; echo time, 58–94 ms; and flip angle, 90° (1.5 T) or 180° (3 T), with b values of 1000 s/mm2. The DWI-positive area was defined by two independent stroke specialists. The volume of the DWI-positive area and the mean ADC level of the DWI-positive area were measured using a workstation (SYNAPSE VINCENT imaging system; Fuji Medical Systems, Tokyo, Japan). We traced all slices of the contour of the DWI-positive area and made the 3D volume of the DWI-positive fields by integrating each slice. The volume of the extracted fields and the mean ADC level of these fields were calculated automatically using the workstation.
Statistical analyses
SPSS version 27 statistical software (IBM, Tokyo, Japan) was used to perform all statistical analyses. Values of p < 0.050 were accepted as indicating statistical significance. Univariate analysis between the two groups was performed using the Mann–Whitney U test or Fisher's exact probability test. Receiver operating characteristic (ROC) curve analysis and the area under the curve (AUC) were used to assess the threshold. In the multivariable analysis, we adjusted for potential imbalances adapted from the univariate analysis and clinical prognostic factors described in the previous literature 9 : onset to recanalization time, recanalization status, and the mean ADC level of all DWI-positive areas. Relationships were expressed as adjusted odd ratios (ORs) with corresponding 95% confidence intervals (CIs).
Results
Between January 2013 and December 2022, 1097 patients were admitted to Our Hospital with acute ischemic stroke. MT was performed in 125 patients with anterior large-vessel occlusions. Among them, 35 patients had low DWI-ASPECTS (0–5) and 28 (80%) patients achieved effective recanalization (TICI2b or TICI3). All 28 patients underwent MRI before and after MT, so 28 patients were finally enrolled in this study (Figure 1). The baseline characteristics and outcomes of the patients are presented in Table 1. Eight patients had decreased DWI high-intensity area volume. The median age was 78.5 years [interquartile range (IQR), 71.5–84.3]. The median volume of DWI-positive area was 153 ml (IQR, 109–204). The puncture to reperfusion time was 48 min (IQR, 37–53). Recirculation status resulting in TICI grade 3 was achieved in 16 patients (57%). The median NIHSS before-MT was 22 (IQR, 18–23).
Figure 1.
Flowchart of patient selection. A total of 1097 patients were admitted to our hospital with acute ischemic stroke. Mechanical thrombectomy (MT) was performed in 125 patients with anterior large-vessel occlusion. Among them, 35 patients had a low diffusion aspect value of <5, and 28 (80%) patients achieved effective recanalization (thrombolysis in cerebral infarction (TICI)2b or TICI3). All 28 patients underwent magnetic resonance imaging before and after MT, so 28 patients were finally enrolled in this study.
Table 1.
Baseline characteristics of patients.
| Volume of DWI-positive area | ||||
|---|---|---|---|---|
| All cases (n = 28) | Decrease (n = 8) | Increase (n = 20) | p-value | |
| Age, years, median (IQR) | 78.5(71.5–84.3) | 78.5(74.3–80.3) | 78.5(69.8–85) | 0.791 |
| Male, no., (%) | 21(75) | 5(63) | 16(80) | 0.371 |
| Occluded vessel, no., (%) | ||||
| ICA | 17(6) | 6(75) | 11(55) | 0.422 |
| MCA | 11(39) | 2(25) | 9(45) | |
| Occlusion side, no., (%) | ||||
| Right | 16(57%) | 3(37%) | 13(65%) | 0.231 |
| Left | 12(43%) | 5(63%) | 7(35%) | |
| DWI ASPECTS, median, (IQR) | ||||
| Pre-MT | 3(2–4) | 3(1–3) | 3(3–4) | 0.211 |
| Post-MT | 3(2–4) | 4(3–4) | 3(1–4) | 0.312 |
| Volume of DWI-positive area, ml, median, (IQR) | ||||
| Pre-MT | 153(109–204) | 146(116–192) | 153(107–206) | 0.790 |
| Post-MT | 168(102–238) | 103(79–140) | 192(131–299) | 0.030* |
| Volume ratio (Post-MT/Pre-MT), (IQR) | 1.25(0.97–1.48) | 0.78(0.47–0.91) | 1.32(1.22–1.52) | <0.001* |
| Mean ADC, ×106 mm2/s, median, (IQR) | 649(592–688) | 717(656–801) | 637(584–641) | 0.011* |
| Onset to recanalization, min, median, (IQR) | 253(190–408) | 227(180–319) | 276(201–464) | 0.290 |
| rtPA, no., (%) | 7(25) | 3(38) | 4(20) | 0.370 |
| Recanalization status, no., (%) | ||||
| TICI2b | 12(43) | 2(25) | 10(50) | 0.401 |
| TITI3 | 16(57) | 6(75) | 10(50) | |
| Symptomatic bleeding, no., (%) | 3(11) | 2(25) | 1(5) | 0.190 |
| NIHSS, median, (IQR) | ||||
| Pre-MT | 22(18–23) | 21(18–23) | 22(18–24) | 0.951 |
| Post-MT | 19(15–23) | 18(11–21) | 21(16–23) | 0.363 |
| 90-Day mRS ≦ 3, no., (%) | 6(21) | 4(50) | 2(10) | 0.040* |
DWI = diffusion-weighted image; ICA = internal carotid artery; MCA = middle cerebral artery; ASPECTS = Alberta Stroke Program Early CT Score; MT = mechanical thrombectomy; ADC = apparent diffusion coefficient; TICI = thrombolysis in cerebral infarction; NIHSS = National Institutes of Health Stroke Scale; mRS = modified Rankin Scale.
There were no significant differences in patient background (age, sex, included vessel, laterality, DWI-ASPECTS before-MT, volume of the DWI-positive area before-MT, and time to onset of recanalization) between the two groups with ratios <1 or >1. There were significant differences in the postoperative volume of the DWI-positive area (103 ml vs. 192 ml, p = 0.030), preoperative mean ADC of all DWI-positive areas (717 × 106 mm2/s vs. 637 × 106 mm2/s, p = 0.011), after/before-MT volume ratios (0.78 vs. 1.32, p < 0.001), and the percentage of good outcomes (50% vs. 10%, p = 0.040) between the two groups. In the multivariable analysis, the mean ADC level of the DWI-positive area was associated with a decrease in the DWI-positive area (OR 1.020, 95%CI 1.001–1.040, p = 0.040) (Table 2). In the ROC curve analysis, the mean ADC threshold level of the after/before-MT DWI-positive area–volume ratio <1 was 649 × 106 mm2/s (sensitivity of 0.875 and specificity of 0.650, and AUC = 0.806) (Figure 2). The mean ADC level of the DWI-positive area and the after/before-MT DWI-positive volume ratio were negatively correlated (p = 0.036, |ρ| = 0.460) (Figure 3). The details of a representative case involving an 80-year- old man are shown in Figure 4.
Table 2.
Multivariate analysis of volume reduction in the DWI-positive area after MT.
| Volume of DWI-positive area | Adjusted OR (95% CI) | p-value | ||
|---|---|---|---|---|
| Decrease (n = 8) | Increase (n = 20) | |||
| Onset to recanalization, min, median, (IQR) | 227(180–319) | 276(201–464) | 1.000(0.994–1.006) | 0.985 |
| Recanalization status, TICI3, no., (%) | 6(75) | 10(50) | 1.693(0.200–14.356) | 0.629 |
| Mean ADC, ×106 mm2/s, median, (IQR) | 717(656–801) | 637(584–641) | 1.020(1.001–1.040) | 0.040* |
Figure 2.
Receiver operating characteristic (ROC) curve analysis of the mean apparent diffusion coefficient (ADC) levels of volume reduction in the diffusion-weighted imaging (DWI)-positive area after mechanical thrombectomy. We established the optimal cutoff value to include the point on the ROC curve closest to (0, 1) and from the Youden index. The threshold of the ADC level of the DWI-positive area–volume reduction was 649 × 10⁶ mm²/s (sensitivity of 0.875, specificity of 0.650, area under the ROC curve = 0.806).
Figure 3.
Relationship between the mean apparent diffusion coefficient (ADC) level area and volume ratio. The horizontal dotted line represents a volume ratio of 1, and the vertical dotted line represents the cutoff value of the ADC level of the diffusion-weighted imaging (DWI)positive area reduction after mechanical thrombectomy (MT) (649 × 10⁶ mm²/s). The mean ADC level of all DWI-positive areas and the after/before-MT DWI-positive area–volume ratio were negatively correlated (p = 0.036, |ρ| = 0.460).
Figure 4.
Representative case. An 80-year-old man presented with an internal carotid artery occlusion. (A) Premechanical thrombectomy (MT) magnetic resonance imaging showed diffusion-weighted imaging (DWI) positivity in the left hemisphere, and the DWI-Alberta Stroke Program Early CT Score (ASPECTS) was 1. (B) We traced a slice of the contour of the DWI-positive area in the apparent diffusion coefficient (ADC) map of pre-MT. (C) We traced all of the slices of the contour of the DWI-positive area and created the 3D volume of the DWI-positive fields by integrating each slice on the workstation. (D) The volume and mean ADC level of the 3D volume were calculated automatically on the workstation. In this case, the volume of the DWI-positive area was 205 ml, and the mean ADC level was 841 × 10⁶ mm2/s. (E) We performed MT, and the recanalization status of thrombolysis in cerebral infarction 3 was achieved. (F) We traced a slice of the contour of the DWI-positive area in the DWI post-MT. (G) We traced all slices of the contour of the DWI-positive area post-MT and made the 3D volume of the DWI-positive fields by integrating each slice at the workstation. In this case, the volume of the DWI-positive area was 87 ml (post-MT/pre-MT volume ratio = 0.42).
Discussion
We showed that in MT for patients with large early ischemic changes, the high mean ADC level of the DWI-positive area was a factor that decreased the volume of DWI-positive areas after MT. The threshold for the mean ADC level at which the volume of the DWI-positive region decreased was 649 × 106 mm2/s. Our study is unique because we enclosed the entire DWI-positive region and examined the mean ADC level. The mean ADC level for the entire DWI-positive region can be easily measured, which is useful in actual emergencies. The mean ADC reflected the depth of the entire cerebral ischemic area, including the penumbra and ischemic core. The ADC level was considered to represent the ischemic depth, and ischemic core thresholds have been reported variously.10,11 Archana Purushotham et al. reported that the optimal threshold for the identification of an ischemic core was an ADC of 620 × 106 mm2/s (sensitivity 0.69, specificity 0.78). 11 Gwak et al. 12 reported that a low ADC 540/ADC 620, which may reflect less severe ischemic stress inside a diffusion lesion, may help identify patients who would benefit from MT despite having a large ischemic core. 11 The mean relative ADC ratio in the DWI reversible area compared with the contralateral normal site ranged from 0.7 to 0.81.13–16
R.N. Sener reported that the ADC value of normal brain cortex was approximately 750 × 106 mm2/s and white matter was 840 × 106 mm2/s. 17 Therefore, our result appears to be reasonable because the threshold for the mean ADC level at which the volume of the DWI-positive region decreased was found to be 649 × 106 mm2/s. Labeyrie et al. reported that DWI reversal was strongly associated with early neurological improvement. 7 Panni et al. reported that DWI reversal after MT in patients with DWI-ASPECTS 0–5 significantly influenced clinical outcomes. 8 In our study, the percentage of good outcome cases was significantly higher in the group with a decreased volume of DWI high-intensity areas. Therefore, DWI reversal may be important for the clinical outcomes of patients with large early ischemic changes.
Complete reperfusion and shorter imaging time to recanalization previously were found to be independently associated with DWI reversal in patients with acute ischemic stroke who received MT.9,18 In our study, these two factors showed a trend but were not significantly different. The small sample size may have affected the results. Regarding DWI reversal in patients with large early ischemic changes, it is possible that early complete recanalization and a high mean ADC level may have even greater effects. Therefore, further analysis is required.
A systematic review and meta-analysis of MT in patients with low ASPECTS scores concluded that patients might become ambulatory and have an acceptable risk of hemorrhage.19,20 MT is cost effective compared with medical management alone in patients with low ASPECTS. 21 Therefore, the use of MT for patients with large early ischemic changes is expected to increase in the future. Even in patients with the same DWI-ASPECTS, the ischemic depth may differ depending on the mean ADC value of the DWI-positive area. Our method, which measures mean ADC levels, is expected to detect better indications among low DWI-ASPECTS cases.
Limitations
This study has several limitations. First, blurring may occur when the ischemic region is defined despite being evaluated by independent stroke specialists. Second, a decreased volume of the DWI-positive areas after MT was associated with a good outcome, but it is difficult to prove that brain tissue in the reversal region is capable of functional recovery. 22 Third, the study sample was small, and MR images were not acquired before and after MT in some patients. Finally, this study used a retrospective cohort design; therefore, future prospective studies with larger cohorts are warranted to confirm these results.
Conclusion
The mean ADC level before-MT correlated with the after/before-MT DWI-positive area–volume ratio. The mean pretreatment ADC cutoff value of 649 × 106 mm2/s predicted a decreased DWI-positive area after MT.
Acknowledgments
The authors would like to thank Enago (www.enago.jp) for the English language review.
Footnotes
Author contributions: Ryosuke Shintoku did conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Roles/Writing original draft; and Writing review & editing. Tatsuya Shimizu did investigation; Data curation; Supervision. Masanori Aihara: Data curation; Supervision. Hirofumi Asano did data curation and investigation. Rei Yamaguchi and Haruka Tsuneoka did data curation. Hiroya Shimauchi-Ohtaki did data curation and Methodology. Masahiko Tosaka did supervision and Methodology. Yuhei Yoshimoto did project administration and supervision;
Data availability: The data that support the findings of this study are openly available.
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethics approval: This study was reviewed and approved by the Institutional Review Board of Gunma University Graduate School of Medicine (HS2022-145). Instead of obtaining informed consent from all study subjects, we disclosed information on how to opt out of this study on our homepage.
Funding: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Grants-in-Aid from the Japanese Society for the Promotion of Science (JSPS) (grant number 22K16676).
ORCID iDs: Ryosuke Shintoku https://orcid.org/0000-0003-4749-3765
Tatsuya Shimizu https://orcid.org/0000-0002-2367-3481
Hirofumi Asano https://orcid.org/0000-0002-5509-1638
References
- 1.Yoshimura S, Sakai N, Yamagami H, et al. Endovascular therapy for acute stroke with a large ischemic region. N Engl J Med 2022; 386: 1303–1313. [DOI] [PubMed] [Google Scholar]
- 2.Sarraj A, Hassan AE, Abraham MG, et al. Trial of endovascular thrombectomy for large ischemic strokes. N Engl J Med 2023; 388: 1259–1271. [DOI] [PubMed] [Google Scholar]
- 3.Huo X, Ma G, Tong X, et al. Trial of endovascular therapy for acute ischemic stroke with large infarct. N Engl J Med 2023; 388: 1272–1283. [DOI] [PubMed] [Google Scholar]
- 4.Inoue M, Olivot JM, Labreuche J, et al. Impact of diffusion-weighted imaging Alberta stroke program early computed tomography score on the success of endovascular reperfusion therapy. Stroke 2014; 45: 1992–1998. [DOI] [PubMed] [Google Scholar]
- 5.Schellinger PD, Bryan RN, Caplan LR, et al. Evidence-based guideline: the role of diffusion and perfusion MRI for the diagnosis of acute ischemic stroke. Chin J Cerebrovasc Dis 2010; 7: 668–672. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Iris EC, Brian T, Haoyue Z, et al. Automated estimation of ischemic core volume on noncontrast-enhanced CT via machine learning. Interv Neuroradiol 2022: 159101992211454. doi: 10.1177/15910199221145487 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Labeyrie MA, Turc G, Hess A, et al. Diffusion lesion reversal after thrombolysis: a MR correlate of early neurological improvement. Stroke 2012; 43: 2986–2991. [DOI] [PubMed] [Google Scholar]
- 8.Panni P, Lapergue B, Maïer B, et al. Clinical impact and predictors of diffusion weighted imaging (DWI) reversal in stroke patients with diffusion weighted imaging Alberta Stroke program early CT score 0–5 treated by thrombectomy: diffusion weighted imaging reversal in large volume stroke. Clin Neuroradiol 2022; 32: 939–950. [DOI] [PubMed] [Google Scholar]
- 9.Yoo J, Choi JW, Lee SJ, et al. Ischemic diffusion lesion reversal after endovascular treatment. Stroke 2019; 50: 1504–1509. [DOI] [PubMed] [Google Scholar]
- 10.Meng X, Fisher M, Shen Q, et al. Characterizing the diffusion/perfusion mismatch in experimental focal cerebral ischemia. Ann Neurol 2004; 55: 207–212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Purushotham A, Campbell BCV, Straka M, et al. Apparent diffusion coefficient threshold for delineation of ischemic core. Int J Stroke 2015; 10: 348–353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gwak DS, Choi W, Shim DH, et al. Role of apparent diffusion coefficient gradient within diffusion lesions in outcomes of large stroke after thrombectomy. Stroke 2022; 53: 921–929. [DOI] [PubMed] [Google Scholar]
- 13.Fiehler J, Knudsen K, Kucinski T, et al. Predictors of apparent diffusion coefficient normalization in stroke patients. Stroke 2004; 35: 514–519. [DOI] [PubMed] [Google Scholar]
- 14.Chung JW, Kim KJ, Noh WY, et al. Validation of FLAIR hyperintense lesions as imaging biomarkers to predict the outcome of acute stroke after intra-arterial thrombolysis following intravenous tissue plasminogen activator. Cerebrovasc Dis 2013; 35: 461–468. [DOI] [PubMed] [Google Scholar]
- 15.Moghadam SE, Ebrahimi SN, Salehi P, et al. Matthias Hamburger3 and EJ. HHS public access. Physiol Behav 2017; 176: 139–148.28363838 [Google Scholar]
- 16.Shinoda N, Hori S, Mikami K, et al. Utility of relative ADC ratio in patient selection for endovascular revascularization of large vessel occlusion. J Neuroradiol 2017; 44: 185–191. [DOI] [PubMed] [Google Scholar]
- 17.Sener RN. Diffusion MRI: apparent diffusion coefficient (ADC) values in the normal brain and a classification of brain disorders based on ADC values. Comput Med Imaging Graph 2001; 25: 299–326. [DOI] [PubMed] [Google Scholar]
- 18.Sakamoto Y, Kimura K, Shibazaki K, et al. Early ischaemic diffusion lesion reduction in patients treated with intravenous tissue plasminogen activator: infrequent, but significantly associated with recanalization. Int J Stroke 2013; 8: 321–326. [DOI] [PubMed] [Google Scholar]
- 19.Baig AA, Bouslama M, Turner RC, et al. Mechanical thrombectomy in low Alberta stroke program early CT score (ASPECTS) in hyperacute stroke—a systematic review and meta-analysis. Br J Radiol 2023; 96: 20230084. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hassan K, Gautam A, Sherief G, et al. Endovascular thrombectomy for ischemic stroke with large core volume: an updated, post-TESLA systematic review and meta-analysis of the randomized trials. Interv Neuroradiol. 2023. doi: 10.1177/15910199231185738 [DOI] [PubMed] [Google Scholar]
- 21.Moreu M, Scarica R, Pérez-García C, et al. Mechanical thrombectomy is cost-effective versus medical management alone around Europe in patients with low ASPECTS. J Neurointerv Surg 2023; 15: 629–633. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Goyal M, Ospel JM, Menon B, et al. Challenging the ischemic core concept in acute ischemic stroke imaging. Stroke 2020; 51: 3147–3155. [DOI] [PubMed] [Google Scholar]