Automated Perfusion Calculations vs. Visual Scoring of Collaterals and CBV-ASPECTS

Abstract

Purpose Use of automated perfusion software has gained importance for imaging of stroke patients for mechanical thrombectomy (MT). We aim to compare four perfusion software packages: 1) with respect to their association with 3‑month functional outcome after successful reperfusion with MT in comparison to visual Cerebral Blood Volume - Alberta Stroke Program Early CT Score (CBV-ASPECTS) and collateral scoring and 2) with respect to their agreement in estimation of core and penumbra volume. Methods This retrospective, multicenter cohort study (2015–2019) analyzed data from 8 centers. We included patients who were functionally independent before and underwent successful MT of the middle cerebral artery. Primary outcome measurements were the relationship of core and penumbra volume calculated by each software, qualitative assessment of collaterals and CBV-APECTS with 3‑month functional outcome and disability (modified Rankin scale >2). Quantitative differences between perfusion software measurements were also assessed. Results A total of 215 patients (57% women, median age 77 years) from 8 centers fulfilled the inclusion criteria. Multivariable analyses showed a significant association of RAPID core (common odds ratio, cOR 1.02; p = 0.015), CBV-ASPECTS (cOR 0.78; p = 0.007) and collaterals (cOR 0.78; p = 0.001) with 3‑month functional outcome (shift analysis), while RAPID core (OR 1.02; p = 0.018), CBV-ASPECTS (OR 0.77; p = 0.024), collaterals (OR 0.78; p = 0.007) and OLEA core (OR 1.02; p = 0.029) were significantly associated with 3‑month functional disability. Mean differences on core estimates between VEOcore and RAPID were 13.4 ml, between syngo.via and RAPID 30.0 ml and between OLEA and RAPID −3.2 ml. Conclusion Collateral scoring, CBV-ASPECTS and RAPID were independently associated with functional outcome at 90 days. Core and Penumbra estimates using automated software packages varied significantly and should therefore be used with caution.

Acute reperfusion therapies for acute ischemic stroke patients with unknown time of symptom onset or in extended time windows: an individualized approach

“…We have shown that reported differences in imaging interpretation are important and appear to limit the use of the DAWN and DEFUSE 3 criteria on core volumes derived from other programs, as they might substantially overestimate the ischemic core volume and lead to the inappropriate disqualification of patients who are eligible for MT. 108 It is clear that the generic term ‘perfusion imaging’ cannot be applied to different software that provide different results for the same patient. Many studies of perfusion imaging in AIS may use software that is suboptimal to the one used in positive published RCTs.…”

Machine Outputs Must Be Checked

“…

We read with interest the multicenter manuscript of Psychogios et al in which they reported on the comparison of infarct core and tissue at risk maps generated by four different vendors as well as visual Cerebral Blood Volume-Alberta Stroke Program Early CT Score (CBV-ASPECTS) and visually assessed collateral scores [1]. They related the maps of 182 patients undergoing mechanical thrombectomy (MT) and receiving a TICI 2b, 2c or III reperfusion to the clinical outcome assessed with the modified Rankin score (mRS) and the functional disability defined as mRS > 2.

…”

“…They calculated mean differences between RAPID (iSchemaView Inc, Menlo Parc, CA, USA) and other software packages and illustrated them with Bland-Altman plots. They concluded that the infarct core defined by the RAPID software correlates best with the clinical outcome whilst VEOcore (VEObrain GmbH, Freiburg, Germany) and syngo.via (Siemens Healthineers AG, Erlangen, Germany) overestimate the infarct core and Olea (OLEA medical Inc., La Ciotat, France) underestimates it [1].The message is clear but can we trust it? In the manuscript the authors clearly state that out of 215 cases 33 cases have been excluded from the final analysis due to "... technical failure of at least 1 perfusion software"; however, if we take a look at the Bland-Altman plot of RAPID-VEOcore (only available in the Supplemental Material) there is a striking outlier in the infarct core volume difference of around -2131 mL (which is distinctly larger than an entire brain).…”

“…They calculated mean differences between RAPID (iSchemaView Inc, Menlo Parc, CA, USA) and other software packages and illustrated them with Bland-Altman plots. They concluded that the infarct core defined by the RAPID software correlates best with the clinical outcome whilst VEOcore (VEObrain GmbH, Freiburg, Germany) and syngo.via (Siemens Healthineers AG, Erlangen, Germany) overestimate the infarct core and Olea (OLEA medical Inc., La Ciotat, France) underestimates it [ 1 ].…”

“…Thus, a single nonplausible outlier caused a significant difference and led to an erroneous conclusion that the VEOcore software is inappropriate for treatment decisions beyond the 6h window [ 1 ].…”