Semi-automated Quantification of Lung Density on Chest CT Used as a Predictive Biomarker of Pulmonary Venous Hypertension. Academic Article uri icon

Overview

abstract

  • RATIONALE AND OBJECTIVES: We sought to determine if lung densities derived from computed tomography scans could be used to identify patients with pulmonary venous hypertension (Group II pulmonary hypertension [PH]), and to compare the performance of this metric with previously described metrics. MATERIALS AND METHODS: Patients were retrospectively included from a single-center cohort of patients with aortic stenosis being evaluated for transcatheter aortic valve replacement from April 2009 to July 2014. Fifty-four patients met inclusion criteria. Thirty-three had PH (pulmonary arterial pressure [PAP] ≥25 mmHg). Thirty-two had Group II PH (pulmonary capillary wedge pressure [PCWP] ≥15 mmHg). Mean lung density (mLD) was measured from chest computed tomography scans using semi-automated techniques. Aortic diameter (mAo) and main pulmonary artery diameter (mPA) were measured manually. These metrics were correlated with PAP and PCWP values. RESULTS: mLD was significantly correlated with PCWP (R = 0.45, P = .0006) and significantly higher in patients with elevated PCWP (P = .006). mPA was weakly correlated with PCWP (R = 0.28, P = .04), but not significantly different in patients with elevated PCWP. mPA/mAo was not significantly correlated with PCWP, nor was it significantly different in patients with elevated PCWP. mLD, mPA, and mPA/mAo were all significantly correlated with PAP and were significantly higher in patients with PH. CONCLUSIONS: Of all metrics, only mLD was significantly correlated with PCWP and served to differentiate patients with elevated and normal PCWP. As such, mLD may contribute to a noninvasive biomarker of pulmonary venous hypertension.

publication date

  • May 18, 2016

Research

keywords

  • Hypertension, Pulmonary
  • Lung
  • Tomography, X-Ray Computed

Identity

Scopus Document Identifier

  • 84969199665

Digital Object Identifier (DOI)

  • 10.1016/j.acra.2016.03.018

PubMed ID

  • 27209264

Additional Document Info

volume

  • 23

issue

  • 8