Frontiers in High-resolution ultrasound imaging: unlock the potential of Ultrasound in providing quantitative biomarkers in Musculoskeletal Radiology (ENHANCE-US)

Introduction: Recent technological advancements in ultrasound (US) equipment, with the introduction of ultra-high frequency probes and the progressive refinement of image-processing algorithms, open new perspectives for the extrapolation of quantitative information from musculoskeletal structures with US. Technological improvement of Elastography and Doppler techniques now enable the calculation of quantitative parameters reflecting, amongst others, tissue perfusion and structure with a much higher accuracy and reproducibility compared to the past. A standardized scanning technique and the use of high-end systems may enhance the quality of images and reduce the presence of artifacts, potentially allowing the analysis of images through Artificial Intelligence (AI) empowered algorithms and the extrapolation of innovative biomarkers. The project ENHANCE-US aimed at exploring the potential of US in providing quantitative biomarkers in Neuromuscular, Rheumatologic, and Orthopaedic diseases. Methods: Four main lines of research were pursued in the three years of the PhD program, respectively addressed at exploring the potential of US in extrapolating biomarkers in the context of: i) muscles degeneration and sarcopenia; ii) rheumatologic conditions; iii) peripheral nerves pathologies; iv) orthopaedic diseases. The research activity is carried out through sequential steps, consisting of: i) developing a standardized scanning technique for examining the questioned musculoskeletal structure and extrapolating new quantitative biomarkers; ii) investigating the Interobserver and Intraobserver variability of US in measuring the specific biomarker through the calculation of the Interclass Correlation Coefficient (ICC); iii) comparing US biomarkers with other gold-standard biomarkers of the same disease; iv) evaluating potential clinical applications of the new biomarker. Results: In a study on cadavers and volunteers, high-resolution US was for the first time demonstrated able to recognize and measure the marginal mandibular branch of the facial nerve (MMN). The MMN could be observed stepping over the mandible body and pointing further anterior in the direction of the innervated muscles. At the point where the facial artery crosses the MMN, the mean diameter of the nerve was 0.69±0.07 mm (95% confidence interval [CI]: 0.66-0.72). The MMN was amenable to US for a segment of 3.4±1.2 cm (95% CI: 2.9-4). The mean distance between the point where the MMN steps over the mandibular body and the gonion was 3.3 ± 0.6 cm (95% CI: 3–3.6). In three patients with clinically evident MMN neuropathy US was able to identify definite signs of nerve damage. Regarding US reproducibility in quantitative analysis of muscles and nerves, through the adoption of a standardized scanning technique, US was demonstrated to be reliable in measuring the thickness and the cross-sectional area of the biceps brachii and the rectus femoris in cadaver, showing almost perfect intraobserver and interobserver agreement (ICC 0.93-0.99, p<.0001). Similarly, semiquantitative US scoring of muscle tropism appeared highly reproducible amongst different observers (ICC 0.85-1). In addition, an optimal (ICC = 0.84; 95% CI, 0.75 to 0.90, p<.0001) and good (ICC = 0.79; 95% CI, 0.69 to 0.87, p<.0001) intra and interobserver agreement in the US measurement of the Recurrent Motor Branch (RMB) of the median nerve diameter were demonstrated in a cohort of 50 healthy volunteers. In this group, the mean nerve diameter was 0.69 mm (SD 0.14). US showed an optimal concordance with gold-standard modalities in muscle biomarkers calculation. In cadaveric specimens, we demonstrated a high concordance between US and histologic measurement of muscles thickness and cross-sectional areas, as well as in the semiquantitative grading of muscle tropism (ICC = 0.92, p<.0001). On the contrary, in a second study we were not able to demonstrate a significant concordance between the diameter of the RMB of the median nerve as measured with US and clinical and electrophysiological measures of nerve pathology in atients with carpal tunnel syndrome (CTS). The potential application of new US biomarkers in pathologic conditions was explored in two studies. In a study on Systemic Sclerosis patients, microvascular flow analysis performed with High-resolution US was able to disclose statistically significant alterations in the terminal arteries of the nailfold. In detail, patients demonstrated an increased resistive index calculated in the nailfold (dRI [SD] = 0.65 [0.14] vs dRI [SD] = 0.57 [0.11]; P = .0115) compared to age and sex-matched controls. Such a difference was not found when the resistive index was calculated in a more proximal vessel such as the proper digital arteries (pRI [SD] = 0.76 [0.11] vs pRI [SD] = 0.73 [0.12]; P = .359), suggesting that Systemic Sclerosis histopathological alterations with capillaries thrombosis and hyalinization may cause blood flow modifications that are detectable only by studying the terminal arteries. In the study regarding RMB neuropathy in CTS, US was able to disclose a statistically significant enlargement of the RMB of the median nerve in patients compared to controls (0.97mm vs. 0.69mm, P < .0001). Conclusion: The research lines pursued during the three years of my PhD program aimed at identifying the main obstacles to the implementation of quantitative musculoskeletal US in clinical practice and investigating potential solutions. Data from both standard (e.g., grey scale, Doppler) and radiofrequency-based (e.g., backscatter) US techniques were acquired and analyzed, in accordance with the purpose of exploring the whole panel of opportunities for biomarker extraction offered by this modality. Further research is required for the complete integration of quantitative musculoskeletal US in the upcoming era of personalized medicine.