Distal popliteal arterial variations may influence the success of femorodistal popliteal and tibial arterial reconstructions. Two patients whose bypass procedures were initially unsatisfactory because of a poor choice for anastomosis stimulated a review of variations in the distal popliteal artery in 1000 femoral arteriograms. The popliteal arterial anatomy could be assessed in 605 extremities and the tibial arterial anatomy in 495 extremities. Seventy-five variant cases were identified. Normal branching of the popliteal artery was present in 92.2%. Among the 7.8% incidence of variants, the majority (72%) were either high origin of the anterior tibial artery or a trifurcation pattern. Of variant patterns to the foot (5.6%), the most common was that in which the supply to the distal posterior tibial artery arose from the peroneal artery. We propose a unified classification of the popliteal and tibial arterial variations that encompasses both anatomic areas. Variant arterial supply to the foot can be suspected when the infrapopliteal vessels show a hypoplastic or aplastic anterior or posterior tibial artery and compensatory hypertrophy of the peroneal artery. Knowledge of these variants is important to angiographers and vascular surgeons.

To overcome the problems associated with gradient-echo (GRE) magnetic resonance (MR) angiography ("bright blood" imaging) and "black blood" imaging with presaturated spin-echo (SE) pulse sequences, the authors devised a new approach for black blood imaging. Their method, selective preinversion fast imaging with steady precession (turboFISP), uses a segmented GRE sequence for fast data acquisition. Nulling of vascular signal results, and stationary tissue appears bright. The method was compared with flow-compensated GRE imaging in a phantom and with GRE imaging and presaturated SE imaging in seven healthy volunteers and nine patients with various cardiac diseases. With phantoms, the selective preinversion turboFISP sequence produced better flow contrast than did GRE sequences. Selective preinversion turboFISP was often superior to SE imaging for depicting vessel lumina, particularly in patients with slowly flowing blood. Arteries appeared dark in selective black blood angiograms, but veins did not. Selective preinversion turboFISP can be used with bright blood GRE imaging to depict vessel lumina, and its capability for image acquisition within a breath hold and with cardiac gating minimizes artifacts from respiration and motion of the vessel wall.

The accurate assessment of pulmonary perfusion is especially important in the evaluation of patients with suspected pulmonary embolism, a common and potentially lethal disorder that can be treated by aggressive anticoagulation. In this study, we demonstrate for the first time the use of MR to image pulmonary perfusion in humans by using dynamic imaging after contrast administration. The technique, which uses an inversion recovery turbo FLASH sequence with ultrashort TE (1.4 ms) and 1-s temporal resolution, was tested in a series of eight healthy subjects and in a porcine model of pulmonary embolism. After the administration of gadopentetate dimeglumine in humans and animal models, there was serial enhancement of the systemic veins, right atrium, right ventricle, and pulmonary arteries. The pulmonary arterial tree was visualized beyond the segmental branches, followed by a gradual diffuse increase in signal intensity of the lung parenchyma over a period of 4.0-7.0 s. Pulmonary circulation times ranged from 3.0-3.4 s. Whereas a high dose (20 or 40 ml) of contrast agent tended to produce the most intense parenchymal enhancement, a low dose (5 ml) was best for showing recirculation. In the animal model, a perfusion defect due to a pulmonary embolus was clearly shown and confirmed by cine angiography. It is concluded that MRI of lung perfusion is feasible. With further development, perfusion MRI could eventually have a significant clinical role in the diagnostic evaluation of pulmonary embolism.