3-D High-Frequency Ultrasound Improves the Estimation of Surface Properties in Degenerated Cartilage

Abstract

High-frequency ultrasound (US) surface parameters are well known to be sensitive to degenerative changes in cartilage tissue, but estimates deteriorate if the sample is inclined. We propose 3-D US to precisely estimate the local surface and inclination. For this purpose, the most common ultrasonic surface parameters ultrasound roughness index and integrated reflection coefficient were extended to 2-D surface measurements. Tissue-mimicking phantoms and human cartilage samples with varying degrees of degeneration were measured using a 40-MHz transducer. Characteristic inclination dependencies of the parameters aided in the distinction between specular reflected or backscattered signal origins and allowed a restriction to suitable local inclinations. In the application to cartilage, comparisons with histologic grading (structural Mankin-score) depicted a statistically significant (p < 0.05) increase of US roughness index for scores larger than 0 and decrease of integrated reflection coefficient for scores larger than 1. The presented findings will increase the reliability of ultrasonic surface parameters and can in principal be applied in vivo.