Background: To investigate the effectiveness of platelet-rich plasma (PRP) injection among patients with rotator cuff tendinopathy in comparison with sham injection, no injection, or physiotherapy alone. Methods: From the earliest records to November 1, 2018, all published or unpublished randomized controlled trial (RCTs) comparing PRP injection with a control group (sham injection, no injection, or physiotherapy alone) in patients with rotator cuff tendinopathy were included. Eligible trials were included from the search results of electronic databases including PubMed, EMBASE, Scopus, The Cochrane Library, as well as the bibliographies of relevant trials. Five RCTs were enrolled in our meta-analysis. Two authors independently assessed the quality of RCTs with the Cochrane risk of bias tool. We designated pain reduction as the primary outcome and functional improvement as the secondary outcome. Standardized mean difference (SMD) was applied for random-effect meta-analysis. Results: In the short term (3–6 weeks) and medium term (12 weeks), the effectiveness of PRP injection and control group was indistinguishable in terms of both outcomes (pain reduction and functional improvement). Nevertheless, PRP injection led to significant long-term (>24 weeks) pain relief (SMD: 0.42, 95% confidence interval (CI): 0.12–0.72, without heterogeneity). For functional improvement in the long term, PRP injection was not more effective than the control group (SMD: 1.20, 95% CI: −0.20–2.59, with heterogeneity). Conclusions: PRP injection may provide benefit over the control group (sham injection, no injection, or physiotherapy alone) in reducing pain at long-term follow-up for patients with rotator cuff tendinopathy.

There is an emerging trend to employ dynamic sonography in the diagnosis of entrapment neuropathy, which exhibits aberrant spatiotemporal characteristics of the entrapped nerve when adjacent tissues move. However, the manual tracking of the entrapped nerve in consecutive images demands tons of human labors and impedes its popularity clinically. Here we evaluated the performance of automated median nerve segmentation in dynamic sonography using a variety of deep learning models pretrained with ImageNet, including DeepLabV3+, U-Net, FPN, and Mask-R-CNN. Dynamic ultrasound images of the median nerve at across wrist level were acquired from 52 subjects diagnosed as carpal tunnel syndrome when they moved their fingers. The videos of 16 subjects exhibiting diverse appearance and that of the remaining 36 subjects were used for model test and training, respectively. The centroid, circularity, perimeter, and cross section area of the median nerve in individual frame were automatically determined from the inferred nerve. The model performance was evaluated by the score of intersection over union (IoU) between the annotated and model-predicted data. We found that both DeepLabV3+ and Mask R-CNN predicted median nerve the best with averaged IOU scores close to 0.83, which indicates the feasibility of automated median nerve segmentation in dynamic sonography using deep learning.