Imaging
Joseph F. Whitehead, MS (he/him/his)
Graduate Research Assistant
University of Wisconsin-Madison
Disclosure information not submitted.
Sarvesh Periyasamy, PhD
Medical Student and Post-doctoral Researcher
University of Wisconsin School of Medicine and Public Health
Paul F. Laeseke, MD, PhD
Assistant Professor
University of Wisconsin
Martin G. Wagner, PhD
Assistant Professor
University of Wisconsin Madison
Michael A. Speidel, PhD
Associate Professor
University of Wisconsin-Madison
Quantitative DSA (qDSA) enables blood velocity quantification from high frame rate DSA imaging. To improve clinical translatability, there is a need to reduce qDSA radiation dose. This work evaluates a new interleaved angiography technique that provides high frame rate, lower dose images for blood velocity quantification while simultaneously providing higher dose images at the same frame rate as conventional DSA for visualizing vessel morphology.
Materials and Methods:
Interleaved acquisition was implemented on a Siemens Artis zee system with custom software for frame-by-frame modulation of radiation dose. The interleaving protocol consisted of nominal 27 frame/s (fps) imaging, with most frames acquired at low dose except for a high dose image every 10th frame (2.7 fps). Accuracy of qDSA velocity was evaluated in a vessel phantom. Images were acquired with 5 mean flow rates [7.5 - 14.2 mL/s] and 3 protocols (27 fps DSA and interleaving with 70% and 89% dose rate reduction relative to 27 fps DSA) and 8 repeat acquisitions per state. An ultrasound (US) flow probe provided gold standard velocity measurements. To evaluate angiographic quality, contrast-to-noise ratios (CNRs) of the gastroduodenal artery (chosen for consistent temporal signal) were measured in porcine hepatic angiograms (n=4) using an interleaved protocol and conventional 3 fps DSA.
Results:
In the phantom study, linear regression showed strong agreement between US and qDSA velocity for each interleaving protocol (n=40, Pearson’s r: [0.98 - 0.99]). For the lowest dose protocol, the slope was 1.03 (95% CI: [1.0, 1.1]) with a bias of 4.7 cm/s (95% CI: [2.2, 7.2]). The 27 fps interleaved dose rate was 1.3 mGy/s, which was 9.4x lower than the 12.6 mGy/s for 27 fps DSA. For reference, the DSA dose rate would have been 1.4 mGy/s for a conventional 3 fps DSA. In vivo, the CNRs in the high dose interleaved images and conventional DSA images were 6.5 ± 0.5 and 7.2 ± 0.5.
Conclusion:
The proposed interleaved angiography technique enables simultaneous acquisition of intra-procedural blood velocity and vessel morphology at dose rates comparable to conventional DSA. This could facilitate clinical translation of qDSA for standardizing treatment endpoints in blood flow altering procedures.