A New Automated Software for Enhanced Computed Tomography Perfusion Volume Assessment

Objective

To evaluate the predictive value of UKIT, a novel automated computed tomography perfusion (CTP) software, for infarct outcomes in acute ischemic stroke (AIS) and compare its perfusion volume assessments with those of MIStar.

Methods

Data from AIS patients were collected across 1 Chinese hospital, with all participants undergoing CTP prior to reperfusion therapy. The concordance of CTP measures between the two softwares was assessed using Spearman rank correlation, Bland-Altman plots, and kappa tests. For AIS patients who received endovascular treatment (EVT) and achieved complete recanalization, consistency between the ischemic core volume as determined by CTP and the ground truth was assessed using Spearman rank correlation.

Results

In the included 278 patients, the ischemic core and hypoperfusion volumes measured by UKIT correlated strongly with MIStar's measurements (ischemic core: r = 0.982, intraclass correlation coefficient = 0.902, P < 0.001; hypoperfusion: r = 0.979, intraclass correlation coefficient = 0.956, P < 0.001). Excellent agreement was observed between the two softwares according to imaging criteria in EXTEND (kappa = 0.73, P < 0.001) and DEFFUSE 3 (kappa =0.73, P < 0.001). Among the 103 patients who underwent EVT and achieved full recanalization, 86 (83.5%) underwent diffusion-weighted imaging within 24 hours post-EVT. Ischemic core volume assessed by both UKIT and MIStar was found to be in close agreement with the ground truth (UKIT: r = 0.695, P < 0.001; MIStar: r = 0.721, P < 0.001).

Conclusions

The automated CTP software UKIT reliably assesses ischemic core and hypoperfusion volumes, aiding in the clinical selection of AIS patients for reperfusion therapy.