Abstract
Purpose :
The presence of vitreous inhomogeneities can result in scattering of both light and ultrasound. We recently demonstrated that quantitative ultrasound methods characterizing echodensities in the vitreous of a clinical population were correlated with contrast sensitivity. In this study we compare the capabilities of three ultrasound probes for detection of vitreous inhomogeneities in an in vitro model.
Methods :
We injected 0.1, 0.25 or 0.5ml of blood-mimicking fluid (BMF-US, Shelley Medical Imaging Technologies, London, ON) at concentrations of 25, 50 and 75% into the central vitreous of fresh ex vivo pig eyes using an 18g needle inserted through the optic nerve. Three eyes were examined at each concentration and volume. Eyes were then imaged using a Quantel Aviso ultrasound system modified for acquisition of phase-resolved echo data using 10 and 20 MHz probes, and by a prototype 20 MHz annular array. We performed calibrated spectrum analysis to assess the relative performance of each device.
Results :
Signal-to-noise ratios (SNR) were 19.1±3.5, 4.4±1.5 and 15.2±3.2dB for 10, 20 and annular array probes respectively. Midband fit, a measure of backscatter amplitude, was correlated with concentration, but not injected volume. Pixel count (the number of pixels exceeding the noise level) was correlated with volume (p<.001) for all probes.
Conclusions :
The 10 MHz probe had the highest SNR, possibly due to the effect of attenuation at 20 MHz. The 20 MHz annular array, however, was almost as sensitive as the single-element 10 MHz probe, most likely due to the higher SNR resulting from the extended depth-of-field produced by synthetic focusing. The use of ultrasound offers an objective means for quantifying vitreous inhomogeneities and may contribute to management in patients with PVD, floaters and other functional disturbances of the vitreous structure.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.