Automated Segmentation of Median Nerve in Dynamic Sonography Using Deep Learning: Evaluation of Model Performance.

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.

Effect of Perineural Injection with Different Dextrose Volumes on Median Nerve Size, Elasticity and Mobility in Hands with Carpal Tunnel Syndrome.

This study aimed to investigate the effect of different injectate volumes on ultrasonographic parameters and the correlation to clinical outcomes under perineural dextrose injection (PDI). In this post hoc analysis of the randomized, double-blinded, three-arm trial, ultrasound-guided PDI with either 1 mL, 2 mL, and 4 mL 5% dextrose water was administered, respectively, in 14, 14, and 17 patients. Ultrasound outcomes included mobility, shear-wave elastography (SWE), and cross-sectional area (CSA) of the median nerve; clinical outcomes were Visual Analog Scale (VAS) and Boston Carpal Tunnel Questionnaire (BCTQ) score. Outcomes were measured before injection, and after injection at the 1st, 4th, 12th, and 24th week. For ultrasound outcomes, CSA decreased significantly from baseline data at all follow-up time-points in the 2 mL group (p = 0.005) and the 4 mL group (p = 0.015). The mean change of mobility from baseline showed a greater improvement on the 4 mL group than the other groups at the 1st week post-injection. For clinical outcomes, negative correlation between the VAS and mobility at the 1st (p = 0.046) and 4th week (p = 0.031) post-injection in the 4 mL group were observed. In conclusion, PDI with higher volume yielded better nerve mobility and decreased CSA of median nerve, but no changes of nerve elasticity.