Reference Phantom Method for Ultrasonic Imaging of Thin Dynamic Constructs
Publication Name
Ultrasound in Medicine and Biology
Abstract
Quantitative ultrasound has a great potential for the non-destructive evaluation of tissue engineered constructs, where the local attenuation and the integrated backscatter coefficient (IBC) can be used for monitoring the development of biological processes. The local determination of both parameters can be achieved using the reference phantom method (RPM). However, its accuracy can be affected when evaluating constructs of evolving sound speed, attenuation and thickness, for example, when evaluating biodegradable hydrogels developing neocartilage. To assess the feasibility of using the RPM under such dynamic conditions while employing a 50-MHz transducer, we conducted a series of experiments on 3-mm-thick acellular hydrogels laden with microspheres. The ultrasonic evaluation procedure used was validated by detecting and compensating for large attenuation variations occurring in the construct, up to 20-fold with respect to the reference phantom, with estimations errors below 1%. We found that sound speed mismatch does not affect the local attenuation estimation, but causes a strong diffraction effect by reducing the backscatter intensity. Such intensity reduction was compensated by determining the IBC percentage change (IBCΔ) as function of sound speed mismatch with respect to the reference phantom (ΔSS), with the equation IBCΔ = (0.63 ± 0.07) ΔSS + (8.54 ± 0.76) 10–3 ΔSS2. The investigated ΔSS interval was up to 120 m/s and using two different concentrations of microspheres, with estimation errors below 7% relative to the construct's actual IBC. Finally, we found that the spectral difference method is sufficient to measure within a few millimetres in depth mismatch, and when combining with sound speed mismatch, we found negligible additional effects. These results pave the way for the use of a generic reference phantom for the evaluation of thin dynamic constructs, thus simplifying the need for using different phantoms depending on the construct's properties.
Open Access Status
This publication is not available as open access
Volume
47
Issue
8
First Page
2388
Last Page
2403
Funding Number
CE 140100012
Funding Sponsor
Australian National Fabrication Facility