Multi-target recognition method and system for brachial plexus ultrasonic images

Abstract

The invention discloses a multi-target recognition method and system for brachial plexus ultrasonic images. The method comprises the steps of acquiring a feature map, obtaining a multi-scale feature map, generating feature maps with different acceptance domains, obtaining a feature map matrix with different scale comparison features, obtaining a feature map matrix after weight endowing, taking each pixel point in the weighted feature map matrix as a central point to generate a region of interest, performing binary classification and detection bounding box regression on the region of interest,aligning the selected regions to form a feature map with a fixed size, classifying the regions of interest in the feature map aligned to the fixed size, segmenting a mask, detecting a bounding box, and outputting to complete multi-target identification of the brachial plexus nerve ultrasonic image. According to the method, an anesthetist can be helped to quickly and accurately identify the required multi-target tissue from the brachial plexus ultrasonic image.

Description

Technical field [0001] The invention relates to the field of image recognition, in particular to a method and system for multi-target recognition of brachial plexus ultrasound images. Background technique [0002] Brachial plexus block (PNB) is a local anesthesia commonly used in surgical procedures. In PNB surgery, the anesthesiologist will first use ultrasound equipment to detect and locate the brachial plexus in ultrasound imaging, and then perform an anesthesia operation for the patient through local anesthesia of the nerve. The brachial plexus and peripheral nerve branches have small targets and dynamically change during ultrasound visualization. It is difficult to identify. The anesthesiologist needs to have sufficient knowledge of the structure of the nerve and the surrounding anatomical tissues (muscle tissue, arterial blood vessels, venous blood vessels) Understand and have rich clinical experience and medical knowledge. Due to the difficulty in identifying nerves, the...

AI Technical Summary

Problems solved by technology

[0008] In view of the above-mentioned deficiencies in the prior art, the present invention provides a multi-target recognition method and system for brachial plexus ultrasound images to solve the problem of poor automatic recognition of brachial plexus ultrasound images

Images