Brachial Plexus Nerve Trunk Segmentation Using Deep Learning: A Comparative Study with Doctors' Manual Segmentation
Yu Wang, Binbin Zhu, Lingsi Kong, Jianlin Wang, Bin Gao, Jianhua Wang, Dingcheng Tian, Yudong Yao
TL;DR
This study addresses the challenge of accurately identifying brachial plexus trunks during ultrasound-guided nerve block anesthesia by creating a public BPUS dataset (BPSegData) and developing BPSegSys, a deep-learning segmentation system. Using Att U-Net with a tailored loss function and CLAHE-based preprocessing, the authors achieve doctor-level trunk segmentation performance and demonstrate that BPSegSys can assist clinicians, improving identification accuracy and showing potential for training applications. Key contributions include ground-truth trunk labels from three doctors, 10-fold cross-validated evaluation across two devices, and evidence that algorithmic assistance reduces inter-reader variability. The work advances automated nerve trunk segmentation for UGNB and suggests practical routes for integrating AI assistance into clinical workflows.
Abstract
Ultrasound-guided nerve block anesthesia (UGNB) is a high-tech visual nerve block anesthesia method that can observe the target nerve and its surrounding structures, the puncture needle's advancement, and local anesthetics spread in real-time. The key in UGNB is nerve identification. With the help of deep learning methods, the automatic identification or segmentation of nerves can be realized, assisting doctors in completing nerve block anesthesia accurately and efficiently. Here, we establish a public dataset containing 320 ultrasound images of brachial plexus (BP). Three experienced doctors jointly produce the BP segmentation ground truth and label brachial plexus trunks. We design a brachial plexus segmentation system (BPSegSys) based on deep learning. BPSegSys achieves experienced-doctor-level nerve identification performance in various experiments. We evaluate BPSegSys' performance in terms of intersection-over-union (IoU), a commonly used performance measure for segmentation experiments. Considering three dataset groups in our established public dataset, the IoU of BPSegSys are 0.5238, 0.4715, and 0.5029, respectively, which exceed the IoU 0.5205, 0.4704, and 0.4979 of experienced doctors. In addition, we show that BPSegSys can help doctors identify brachial plexus trunks more accurately, with IoU improvement up to 27%, which has significant clinical application value.