Relation of mean pixel intensity to concentration of sonicated albumin microspheres: effects of ultrasound system settings. Academic Article uri icon

Overview

abstract

  • Myocardial contrast echocardiography (MCE) is a recently developed technique to assess myocardial perfusion. Previous studies have shown that MCE frequently underestimates coronary reserve as compared with other methods. Previous reports have suggested that some of the limitations of MCE are due to the nonlinear processing algorithms of the ultrasound systems. Therefore we designed an in vitro model to study the effects of various processing algorithms and levels of acoustic power (AP) on the relation between concentration of echocontrast and mean pixel intensities. A beaker containing a commercially produced, diluted solution of sonicated albumin microspheres was imaged with a commercial Hewlett Packard ultrasound system, and mean pixel intensities were determined with an off-line computer system. The solution was imaged at a wide range of APs (40-10 dB) and time gain compensation (40-60 dB). Concentration versus intensity curves were generated using all possible combinations of compression and processing algorithms available on the ultrasound system used. Contrast effect diminished rapidly when exposed to high AP. Lower powers had less effect on contrast intensity duration, with no effect seen below 20 dB. Changes in time gain compensation did not affect contrast intensities. Regardless of the processing algorithm examined, the relation between concentration of echocontrast and mean pixel intensities was near linear in the first two to three concentrations; thereafter, the curves flattened during increasing concentrations of contrast. At higher concentrations, acoustic shadowing produced a decrease in pixel intensities.(ABSTRACT TRUNCATED AT 250 WORDS)

publication date

  • November 1, 1991

Research

keywords

  • Albumins
  • Diagnosis, Computer-Assisted
  • Echocardiography
  • Image Enhancement

Identity

Scopus Document Identifier

  • 0026248783

Digital Object Identifier (DOI)

  • 10.1002/clc.4960141706

PubMed ID

  • 1764837

Additional Document Info

volume

  • 14

issue

  • 11 Suppl 5