Large Differences in Pulsatility and Resistance of the Middle Cerebral Artery Assessed With Transcranial Doppler Versus Transcranial Color-Coded Doppler: Impact of Post-Processing Approaches

Background At present, many clinical examinations of the cerebral circulation include a non-imaging transcranial Doppler (TCD) assessment. While post-hoc correction of Doppler angles has been attempted by imaging probes through a temporal window, measurement error may not be fully compensated and may translate into inaccurate quantification of clinically-meaningful parameters, such as pulsatility index (PI). Therefore, the purpose of this study was to compare the traditional TCD with transcranial color-coded Doppler (TCCD). Methods For this pilot study in 123 healthy volunteers, middle cerebral artery (MCA) measurements of peak-systolic velocity (PSv), end-diastolic velocity (EDv), PI and resistance index (RI) of the MCA were compared between conventional non-imaging TCD and TCCD. Data were analyzed with paired, two-tailed t -tests, Bland–Altman limits of agreement and polynomial second-order regression analysis. Post-hoc custom signal post-processing was used exploratively to attempt to interpret observed differences. Results After angle-correction with TCCD, PSv, EDv, PI and RI were different from TCD (all p < 0.0001), and the variances of PSv, PI and RI were greater with TCCD ( p = 0.01, p = 0.0003, p = 0.005, respectively). Bland–Altman analyses revealed non-linear bias for most parameters, with an average mean difference of +7.28 cm/s in EDv. Manual post-processing of the TCD EDv signal revealed a discrepancy of +7.36 cm/s. Conclusion Compared with conventional TCD, TCCD generated a 33% higher PI and 18% higher RI, largely due to differences in EDv and likely explained by differences in signal post-processing. The greater variances observed with TCCD distinguish it from traditional TCD and may present an opportunity to provide a more sensitive marker of cerebral hemodynamics and clinical categorization of patients. Collectively, the findings suggest a much-needed reconsideration of ultrasound-based cerebrovascular assessments.