Challenge: Nerve blocks have revolutionized anesthesia and surgery as well as chronic pain management. They provide better pain relief, significantly reduce the need for opioids, thus impacting on the problem of opioid addiction, speed up recovery from surgery, as well as shorten hospital stay. However, they require special training as well as regular skills maintenance. Very few anesthesiologists have this formal training and are currently able to perform all 58 described nerve blocks.
Solution: Artificial intelligence (AI) offers a solution to the improvement of the nerve block procedure: it can reduce inter-user variability, thus improving the quality of care, lower the complication rate, make these blocks more accessible to all anesthesiologists, and reduce costs. The goal of this project is to develop a marketable product to assist clinicians in performing nerve blocks. By using modern deep learning and computer vision techniques, several technologies will be developed within this project, including a software to identify nerves in ultrasound images with high accuracy and a decision tool to guide users to perform nerve blocks by providing clinical suggestions on medication type and dose for a given clinical scenario.
Expected Achievements /Impacts: Compared with other research groups focused on nerves, the developed marketable product will be a fully integrated product, a ‘complete package’ that combines the nerve recognition software and the decision support system for the clinician’s guidance. This technology will have an immediate impact on healthcare quality and costs. Moreover, the project will allow clinicians to develop dual competencies: in AI and anesthesia. This will increase Quebec’s visibility in AI and catalyze innovation in medicine. In addition, this project will strongly contribute to the development of the newly Quebec-based company Divocco AI.
