PurposeTo determine the feasibility of using a 2-dimensional quantitative digital subtraction venography (qDSV) technique that employs a temporally modulated contrast medium injection to quantify blood velocity in phantom, normal, and stenotic porcine iliac vein models. Materials and MethodsBlood velocity was calculated using qDSV following temporally modulated pulsed injections of iodinated contrast medium and compared with Doppler ultrasound (US) measurements (phantom, in-line sensor; in vivo, diagnostic linear probe). Phantom evaluation was performed in a compliant polyethylene tube phantom with simulated venous flow. In vivo evaluation of qDSV was performed in normal (n = 7) and stenotic (n = 3) iliac vein models. Stenoses were created using endovenous radiofrequency ablation, and blood velocities were determined at baseline, after stenosis, after venoplasty, and after stent placement. ResultsIn the phantom model, qDSV-calculated blood velocities (12–50 cm/s) had very strong correlations with US-measured velocities (13–51 cm/s) across a range of baseline blood velocities and injection protocols (slope, 1.01–1.13; R2 = 0.96–0.99). qDSV velocities were similar to US velocities regardless of injection method: (a) custom injector, (b) commercial injector, or (c) hand injection. In the normal in vivo model, qDSV-calculated velocities (5–18 cm/s) had strong correlation (slope, 1.22; R2 = 0.90) with US-measured velocities (3–20 cm/s). In the stenosis model, blood velocities at baseline, after stenosis, after venoplasty, and after stent placement were similar on qDSV and US at all time points. ConclusionsVenous blood velocity was accurately quantified in a venous phantom and in vivo porcine models using qDSV. Intraprocedural changes in porcine iliac vein blood velocity were quantified with qDSV after creation of a stenosis and subsequently treating it with venoplasty and stent placement.