In this study, the reflection coefficients of the cartilage tissue and the roughness index of the cartilage surface were quantitatively described using the ultrasound microscopic imaging, and the relationship between the quantitative acoustic parameters and the severity of OA cartilage qualitatively assessed by histology was explored.
The results of our histological assessment indicate that the pathological symptoms in OA cartilage become more obvious and serious with the increase of OA grade. The cartilage surface became more uneven, the fibrillation of the superficial tissue more serious, and toluidine blue staining faded or totally disappeared from the surface to the deep layers as OA grade increased. The cartilage thickness decreased in the grade 3 cartilage samples. Previous studies reported consistent results that the proteoglycan content of articular cartilage changed first in the early OA, decreasing gradually from the surface to the deeper layers with the degeneration [22, 23].
It has been well known that the earliest signs of OA include the loss of proteoglycans in the superficial layer and the disruption of the superficial collagen network, leading to fibrillation in the surface and softening in the superficial tissue [24–26]. High-frequency UBM has been demonstrated its useful application in detecting the early damage in the articular cartilage tissue due to the ability of ultrasound to penetrate into the tissue . Responding to the changes in the surface roughness and the internal compositions of the OA cartilage, the ultrasound reflection and scattering in the tissue increased and thus the enhanced echoes reflected from the internal tissue were demonstrated in the UBM image . The finding is consistent with the results of other studies [12, 27]. These studies, however, did not provide any quantitative criteria for the ultrasonic diagnosis of OA cartilage.
In engineering, the roughness index is mainly used to describe the small valley and pitch conditions on the material surface, as known as micro-roughness. Recently, it has been used to describe the roughness of the cartilage surface in the evaluation of cartilage degeneration. Our results of the surface roughness calculation show that the roughness of the cartilage surface increased with the OA grade, indicating that the cartilage surface became rough and uneven with the development of OA. Similar results have been obtained in previous studies of cartilage degradation . The statistical results in this study showed significant differences in URI between the grade 3 cartilage and the cartilage with grade 1 and grade 2 (p < 0.05), but an insignificant change between the grade 1 cartilage and the grade 2 cartilage (p > 0.05), showing that cartilage degradation progressively developed from quantitative change to qualitative change. It was found that the surfaces of the cartilage samples with grade 1 and 2 started to become rough with a significant increase in URI compared with normal cartilage, but no significant change between groups of grade 1 and grade 2. However, a sharp deterioration occurred in the grade 3 cartilage samples with a great increase in URI to 193.27 ± 75.06 μm. The results show that the surface roughness index URI could be used to distinguish early OA and mid OA, but has no ability to accurately distinguish the grades of early OA (i.e. grade 1 and grade 2).
Two possible reasons may explain the decrease in the reflection coefficient of the cartilage surface with the OA grade. First, the increase of surface roughness resulted in diffuse reflection and consequently decreased the amplitude of the echoes. Secondly, while OA occurred, the cartilage surface was softened. The compositions and structure of articular cartilage gradually changed from the surface to the deep layer. As more transmitted ultrasonic energies were absorbed, the reflection coefficient decreased. Contrary to the reflection coefficient of the surface, the reflection coefficient of the cartilage-bone interface increased with OA grades (Figure 4c). It was revealed that the bottom reflection coefficient of the grade 1 cartilage changed insignificantly compared with normal cartilage while that of OA cartilage with grade 2 and 3 increased significantly. The OA cartilage with grade 3 had a significant increase compared with normal cartilage and early OA cartilage (grade 1 and 2).
Therefore, it might be speculated that the reflection coefficient of the cartilage surface could be a more sensitive index to distinguish the early OA grades than the surface roughness index and the reflection coefficient of the cartilage-bone interface. The integrated analysis of these three parameters in diagnosis of cartilage degeneration not only evaluates the surface morphology (surface roughness), but also assesses the impact of the composition changes by measurement of the reflection coefficients of both the cartilage surface and the cartilage-bone interface. Thus, more accurate diagnostic results may be obtained.
Two limitations of this study require further investigations. First, the OA grades were only evaluated subjectively by three pathologists in this study. The quantitative analysis of changes in compositions and structural parameters such as PG content and cartilage thickness and the sequent study in their relationships to ultrasound parameters are needed. Secondly, the numbers of the samples with different OA grades were uneven, especially the number of OA cartilage with grade 3 was small (N = 6). The small number of samples may have some impacts on the statistical results. Thus further studies with large number of samples are needed.