Evaluating Deep Learning-Based Nerve Segmentation in Brachial Plexus Ultrasound Under Realistic Data Constraints

TL;DR

This work addresses ultrasound-based brachial plexus nerve localization under realistic data constraints, proposing a U-Net segmentation framework evaluated across datasets from two ultrasound machines. It systematically analyzes dataset composition and annotation strategy, showing that combining data from multiple devices provides regularization for weaker acquisition sources but may not beat best single-source performance, and that multi-class annotation can substantially reduce nerve segmentation accuracy. A key finding is a positive correlation between nerve size and segmentation quality (Pearson $r=0.587$, $p<0.001$), highlighting smaller nerves as a major challenge. The results offer practical guidance for deploying robust ultrasound nerve segmentation systems, balancing data diversity with domain-specific targeting, and motivating further improvements in context-aware learning and clinical validation.

Abstract

Accurate nerve localization is critical for the success of ultrasound-guided regional anesthesia, yet manual identification remains challenging due to low image contrast, speckle noise, and inter-patient anatomical variability. This study evaluates deep learning-based nerve segmentation in ultrasound images of the brachial plexus using a U-Net architecture, with a focus on how dataset composition and annotation strategy influence segmentation performance. We find that training on combined data from multiple ultrasound machines (SIEMENS ACUSON NX3 Elite and Philips EPIQ5) provides regularization benefits for lower-performing acquisition sources, though it does not surpass single-source training when matched to the target domain. Extending the task from binary nerve segmentation to multi-class supervision (artery, vein, nerve, muscle) results in decreased nerve-specific Dice scores, with performance drops ranging from 9% to 61% depending on dataset, likely due to class imbalance and boundary ambiguity. Additionally, we observe a moderate positive correlation between nerve size and segmentation accuracy (Pearson r=0.587, p<0.001), indicating that smaller nerves remain a primary challenge. These findings provide methodological guidance for developing robust ultrasound nerve segmentation systems under realistic clinical data constraints.

Figures (6)

• Figure 1: Representative ultrasound images of the brachial plexus with expert annotations for artery, vein, nerve, and muscle.
• Figure 2: Comparison of ultrasound images from (a) ultrasound-1 (SIEMENS ACUSON NX3 Elite) and (b) ultrasound-2 (Philips EPIQ5).
• Figure 3: Distribution of image metrics (brightness, contrast, tonal richness, and sharpness) across the two ultrasound machines.
• Figure 4: The U-Net architecture used for nerve segmentation, featuring skip connections for boundary preservation.
• Figure 5: Experimental setup for combined dataset training and cross-machine evaluation.