MRI of the pulmonary veins: Comparison between 3D MR angiography and T1-weighted spin echo

Objective: The purpose was to determine the ability of three-dimensional (3 D) magnetic resonance (MR) angiography to depict normal pulmonary veins in comparison with spin-echo MR imaging. Materials and Methods: MR imaging of 40 patients with cardiovascular diseas e were reviewed. Patients with known pulmonary venous abnormalities were ex cluded. Using a standard GE 1.5-T magnet, axial T1-weighted spin-echo 5-mm- thick contiguous slices and 3D MR angiography (contiguous slice thickness o f 2.5-3.5 mm, 20-30 cc of gadolinium bolus at 1-1.5 cc/sec, 32-43-second br eath-hold, coronal and sagittal plane acquisition) were evaluated retrospec tively on separate occasions by two experienced radiologists. Multiplanar i maging projection was used for the identification of pulmonary veins. Each lung was considered to have two drainage veins: a superior vein and an infe rior vein. identification of a pulmonary vein was made by visualizing a con nection with the left atrium. Results: 143 pulmonary veins (87.5% +/- 5.2) were identified at the level o f the left atrium on TI-weighted spin-echo images, and 157 (98.1% +/- 1.9) were identified on 3D MR angiography (p < 0.01). Overall we identified by T 1-weighted spin-echo imaging 36 right upper, 38 right lower, 27 left upper, and 38 left lower pulmonary veins. By 3D MR angiography, we identified 38 right upper, 40 right lower, 39 left upper, and 40 left lower pulmonary vei ns. All four pulmonary veins were detected in 22 patients on spin-echo imag ing (55%) and in 37 patients (92.5%) on 3D MR angiography (chi = 3.81, p < 0.05). Conclusion: A significant difference is demonstrated between 3D MR angiogra phy and spin-echo MR imaging in identifying normal pulmonary veins. MR angi ography provides a complete view of normal pulmonary venous anatomy and cou ld be a valuable tool for the assessment of abnormal pulmonary venous drain age.