The purpose of the study is to develop and validate an algorithm based on deep neural networks (DNNs) to identify interscalene brachial plexus on ultrasonography automatically.
The investigators plan to develop a deep learning-based network to automatically identify interscalene brachial nerves on ultrasound images. The trained model will be validated on an independent dataset. The performance of the network will also be compared against practicing anesthesiologists.
Study Type
INTERVENTIONAL
Allocation
NA
Purpose
DIAGNOSTIC
Masking
NONE
Enrollment
1,126
the participants will be placed in the supine position, with head turned slightly away from the operating side and arms beside the body. The operator will identify right and left interscalene brachial plexuses by ultrasound equipment (Sonosite EDGE or GE LOGIQ e). Clear images and videos of brachial plexus will be captured and saved.
Huashan Hospital
Shanghai, Shanghai Municipality, China
The distance of the lateral midpoints of the nerve sheath contours
between model predictions and the ground truth; between nonexpert anesthesiologist predictions and the ground truth
Time frame: immediately after the procedure
Accuracy, Sensitivity and specificity
Accuracy, Sensitivity and specificity of the network and nonexpert anesthesiologists
Time frame: immediately after the procedure
The percentage of the intersection over union
between model predictions and the ground truth; between nonexpert anesthesiologist predictions and the ground truth
Time frame: immediately after the procedure
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.