Title
NerveGPS: A Novel Decision Support System for Ultrasound Nerve Block Guidance
Abstract
Peripheral regional nerve blockade (nerve block) is used in anesthesia to administer anesthesia to certain regions of the body. During nerve block the anesthetic is injected directly into or around the area surrounding the nerve. One of the most critical components of ultrasound-guided nerve blocks is the detection of the nerve within the ultrasound image. This study presents a software-based solution for the detection of nerves within the ultrasound feed. The software that was developed used image processing tools to detect the location of the sciatic nerve at the popliteal fossa within an ultrasound feed. For the study, 100 images were obtained from authors. Then, 2 anesthesiologists with experience in ultrasound-guided nerve blocks were asked to identify the contour of the nerve in the obtained images. The study used two criteria for evaluating the software: 1) the location of the software determined nerve center fell within the manually defined contour and 2) the percentage of overlap between the software and manually detected areas. A test was successful if the percentage of overlap between automatic and manual detection was a minimum of 95% for a circle of 0.4cm diameter centered at the software detected nerve center. For the 100 image sample, the nerve was detected within the manually detected area in 99% of the cases, with an overlap of 98% for a 0.4cm diameter circle. Time to detect the nerve ranged from 0.95s to 1.6s. These preliminary results show that the nerve detection software provides a reliable target for needle insertion. Future work will extend the software to different nerves of interest.
Year
DOI
Venue
2014
10.1109/CBMS.2014.34
CBMS
Keywords
Field
DocType
ultrasound, decision support, nerve block, sciatic nerve, ultrasound imaging,decision support system,anesthesia,biomedical imaging,decision support systems,neurophysiology,decision support,ultrasound
Biomedical engineering,Medical imaging,Computer science,Image processing,Popliteal fossa,Software,Artificial intelligence,Surgery,Ultrasound image,Computer vision,Nerve block,Sciatic nerve,Ultrasound
Conference
ISSN
Citations 
PageRank 
2372-9198
0
0.34
References 
Authors
0
7