Dataset Information

Automated Segmentation of the Median Nerve in the Carpal Tunnel using U-Net.

ABSTRACT: Nerve area and motion in carpal tunnel syndrome (CTS) are currently under investigation in terms of prognostic potential. Therefore, there is increasing interest in non-invasive measurement of the nerve using ultrasound. Manual segmentation is time consuming and subject to inter-rater variation, providing an opportunity for automation. Dynamic ultrasound images (n = 5560) of carpal tunnels from 99 clinically diagnosed CTS patients were used to train a U-Net-shaped neural network. The best results from the U-Net were achieved with a location primer as initial region of interest for the segmentations during finger flexion (Dice coefficient = 0.88). This is comparable to the manual Dice measure of 0.92 and higher than the resulting automated Dice measure of wrist flexion (0.81). Although there is a dependency on image quality, a trained U-Net can reliably be used in the assessment of ultrasound-acquired median nerve size and mobility, considerably decreasing manual effort.

SUBMITTER: Festen RT

PROVIDER: S-EPMC8169596 | biostudies-literature | 2021 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Automated Segmentation of the Median Nerve in the Carpal Tunnel using U-Net.

Festen Raymond T RT Schrier Verena J M M VJMM Amadio Peter C PC

Ultrasound in medicine & biology 20210427 7

Nerve area and motion in carpal tunnel syndrome (CTS) are currently under investigation in terms of prognostic potential. Therefore, there is increasing interest in non-invasive measurement of the nerve using ultrasound. Manual segmentation is time consuming and subject to inter-rater variation, providing an opportunity for automation. Dynamic ultrasound images (n = 5560) of carpal tunnels from 99 clinically diagnosed CTS patients were used to train a U-Net-shaped neural network. The best result ...[more]

PMID: 33931286

Similar Datasets

Project description:BackgroundThere is little consensus on ultrasound (US) normative values of cross-sectional area of median nerve (MN-CSA) in carpal tunnel syndrome (CTS) because of its dependency on anthropometric parameters. We aim to propose a novel anthropometric-independent US parameter: MN-CSA/flexor radialis carpi CSA (FCR-CSA) ratio ("Nerve Tendon Ratio", NTR), in the diagnosis of clinically and electrodiagnostic (EDS)-defined CTS.Methods74 wrists of 49 patients with clinically defined CTS underwent EDS (scored by the 1-5 Padua Scale of electrophysiological severity, PS) and US of carpal tunnel with measurement of MN-CSA (at the carpal tunnel inlet), FCR-CSA (over scaphoid tubercle) and its ratio (NTR, expressed as a percentage). US normality values and intra-operator agreement were assessed in 33 healthy volunteers.ResultsIn controls, the mean MN-CSA was 5.81 mm2, NTR 64.2%. In 74 clinical CTS, the mean MN-CSA was 12.1 mm2, NTR 117%. In severe CTS (PS > 3), the mean MN-CSA was 15.9 mm2, NTR 148%. In CTS, both MN-CSA and NTR correlated with sensitive conduction velocity (SCV) (p < 0.001), distal motor latency (DML) (p < 0.001) and PS (p < 0.001), with a slight superiority of NTR vs. MN-CSA when controlled for height, wrist circumference and weight. In CTS filtered for anthropometric extremes, only NTR maintained a correlation with SCV (p = 0.023), DML (p = 0.016) and PS (p = 0.009). Diagnostic cut-offs were obtained with a binomial regression analysis. In those patients with a clinical diagnosis of CTS, the cut-off of MN-CSA (AUROC: 0.983) was 8 mm2 (9 mm2 with highest positive predictive value, PPV), while for NTR (AUROC: 0.987), the cut-off was 83% (100% with highest PPV). In patients with EDS findings of severe CTS (PS > 3), the MN-CSA (AUROC: 0.876) cut-off was 12.3 mm2 (15.3 mm2 with highest PPV), while for NTR (AUROC: 0.858) it was 116.2% (146.0% with highest PPV).ConclusionsNTR can be simply and quickly calculated, and it can be used in anthropometric extremes.

Dataset Information

Automated Segmentation of the Median Nerve in the Carpal Tunnel using U-Net.

Publications

Automated Segmentation of the Median Nerve in the Carpal Tunnel using U-Net.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets