Magnetic resonance velocity mapping.

Magnetic resonance velocity mapping is achieved by encoding velocity at each point in a tomographic imaging plane in the phase of the magnetic resonance signal. Although this can be achieved with almost any imaging sequence, cine gradient echo sequences are particularly suitable because of the high signal from blood and the ability to repeat the sequence rapidly to form a cine image. The technique has been shown to be accurate by in vitro and in vivo validation, with flow measurements in the great vessels having an accuracy of approximately 6%. A potential problem arises from loss of signal from turbulent blood flow, but this can be overcome with the use of even echo rephasing and echo times below 5 ms. Using such sequences, velocities of up to 6 m s-1 have been measured clinically and pressure gradients across valves as great as 16 kPa (120 mmHg) can be computed. Clinical application has centred on the measurement of flow in the pulmonary circulation and in shunts and conduits in patients with congenital heart disease. Other applications include the measurement of valvular regurgitation and stenosis, and flow in coronary artery bypass grafts. Flow in native coronary arteries has been measured but the technique requires further development before such measurements can be considered reliable. Real time imaging using echo planar velocity mapping has been achieved and it is hoped that this will make coronary artery flow measurements more robust.