Robust cerebrovascular blood velocity and flow rate estimation from 4D‐CTA. Issue 5 (23rd March 2019)

• Record Type: Journal Article
• Authors: Bouillot, Pierre; Brina, Olivier; Chnafa, Christophe; Cancelliere, Nicole M.; Vargas, Maria I.; Radovanovic, Ivan; Krings, Timo; Steinman, David A.; Pereira, Vitor M.
• Is Part Of: Medical physics. Volume 46:Issue 5(2019)
• Page Start: 2126
• Page End: 2136
• Publication Date: 2019-03-23
• Subjects: blood flow -- cerebrovascular disease -- CT angiogaphy; Medical physics -- Periodicals; Medical physics; Geneeskunde; Natuurkunde; Toepassingen; Biophysics; Periodicals; Periodicals; Electronic journals; 610.153
• DOI: 10.1002/mp.13454
• Languages: English
• ISSNs: 0094-2405

Abstract

Abstract : Purpose: Blood velocity and flow rate information may be helpful for a wide variety of applications, but it often requires separate acquisitions. The dynamic information combined with the high spatial resolution of four‐dimensional computed tomography angiography (4D‐CTA) offers the possibility to quantify blood flow simultaneous to vascular anatomy. Methods: A 4D‐CTA clinical protocol with a novel dedicated postprocessing were validated in vitro in a patient‐specific model, and tested in a pilot study of six patients. Blood flow was assessed in both internal carotid (ICAs) and vertebral (VAs) arteries by analyzing spatial displacement of contrast agent in the form of time‐intensity curves (TICs). Unlike previous approaches, it does not require any a priori assumptions about TIC shape, but rather computes mean velocity and flow rates from the spatial displacement of the TICs along the automatically segmented vessels. Results: In vitro experiments showed good agreement between 4D‐CTA and flowmeter measurements under steady and pulsatile flow conditions. In vivo measurements exhibited large interpatient variability of the TIC shapes, from which blood flow rates could nevertheless be successfully measured in all patients and investigated vessels. On average, measured flow rates were 3.2 ± 0.7 ml/s (in ICAs) and 1.3 ± 0.8 ml/s (in VAs) consistent with previous reference standards. Contrary to our novel approach, which considered the full TIC shape, application of previous time‐to‐peak analyses based on idealized TIC shapes showed limited reliability. Conclusions: We demonstrate the high potential of 4D‐CTA for assessing blood velocity and flow rate in addition to anatomical evaluation. The wide variety of TIC shapes encountered in vivo highlights the importance of an adaptive TIC analysis as proposed in the present work.