Highlighting Nerves in Images for Ultrasound Guided Regional Anesthesia

Abstract

Surgery is a treatment given to the patient, when patient is having injuries in the body parts. While doing the surgery the patient will undergo pain and discomfort, to reduce the pain of patient anesthesia is given before surgery. There are many types of anesthesia procedure called local, regional and general. Local anesthesia is given to specific part of the body; general anesthesia is where the patient is completely senseless. Regional anesthesia is procedure, where anesthetist does injection near a class of nerves to numb the area of human body that requires surgery. There is difference in structure of every person. This makes doctor hard to detect nerve in internal organ. Ultrasound image provide information about structure of body which can be used for regional anesthesia. Ultrasonography is the safest method compared to the other imaging techniques like Xray, CT, and MRI; because it is not based on ionizing radiation. In this paper, we simplify the issue of detection of nerve in ultrasound image. For input, image such as ultrasound images of brachial plexus are used. Initially in preprocessing, filtering technique is used to remove noise from ultrasound image and then segmentation is carried out using active contour model. Performance of segmentation is evaluated using dice coefficient and Jaccard coefficient. The nerve location in ultrasound image is detected to help the doctors in giving regional anesthesia. © 2020 IEEE.