Secondary Logo

Institutional members access full text with Ovid®

Pulmonary Artery Measurements in Pulmonary Hypertension: The Role of Computed Tomography

Mahammedi, Abdelkader MD*; Oshmyansky, Alexander MD, PhD*; Hassoun, Paul M. MD; Thiemann, David R. MD‡,§; Siegelman, Stanley S. MD*

doi: 10.1097/RTI.0b013e318271c2eb
Original Articles

Purpose: We studied the relationship between pulmonary artery diameter (PAD) as measured on computed tomography (CT) and pulmonary artery pressure (PAP) with the specific goal of assessing the reliability of various measurements on high-resolution chest CT as predictors of pulmonary hypertension (PH).

Materials and Methods: In a preliminary study we determined the method of measuring the main PAD (mPAD) that best correlated with PAP. Using this approach we measured mPAD on CT and correlated the data with PAP obtained from right heart catheterization in 298 patients with known PH and in 102 controls. Various metrics were analyzed for their specificity and sensitivity as screening measurements for PH.

Results: The mean PAD and mPAD/ascending aorta diameter (AAD) ratio were found to have the highest correlation with PAP (r=0.51 and 0.53, respectively; P<0.001). A threshold of mPAD>29.5 mm was found to be 70.8% sensitive and 79.4% specific for PH, and an mPAD threshold >31.5 mm had a sensitivity and specificity of 52.0% and 90.2%, respectively. An mPAD/AAD ratio >1 was found to be 70.8% sensitive and 76.5% specific for PH. There was no significant correlation between mPAD and body surface area or age (r=0.04 and 0.07, respectively). A strong statistically significant difference (P<0.0001) was found between mPAD and mPAD/AAD ratio between controls and the PH group.

Conclusion: mPAD and mPAD/AAD ratio may be used to detect PH in patients of any age or with any body surface area.

*Russell H. Morgan Department of Radiology and Radiological Science

Pulmonary and Critical Care Medicine

Departments of Medicine

§Epidemiology, Division of Health Sciences Informatics of the Johns Hopkins Medical Institutions, Baltimore, MD

The authors declare no conflicts of interest.

Reprints: Stanley S. Siegelman, MD, Russell H. Morgan Department of Radiology and Radiological Science, 601 N, Caroline St Suite 4214, Johns Hopkins Outpatient Center, Baltimore, MD 21287 (e-mail:

© 2013 Lippincott Williams & Wilkins, Inc.