Bringing The Anatomy Textbook to Real-time Handheld Ultrasound
ePoster
Authors: Kris Dickie, Clarius Mobile Health Ben Kerby, Clarius Mobile Health
Handheld ultrasound is being broadly and rapidly adopted by clinical specialities that struggle to keep pace with the required education that goes hand-in-hand with using and interpreting diagnostic ultrasound. As artificial intelligence technological dominance grows, and access to clinical images expands through open platforms and cloud based archiving solutions, new functionality can be rapidly developed to help aid in the training and education of new ultrasound users. Many existing AI training tools apply labels on the ultrasound image or may colour-code regions of the image with an opacity to see through to the greyscale ultrasound underneath, however these solutions do not help correlate anatomy to typical medical school training resources and materials, thus potentially creating a critical gap in understanding for those that have not undertaken an ultrasound fellowship or a degree clinical ultrasonography. Through the use of AI and novel GPU processing, our aim was to help bridge that gap by creating useful tools for multiple clinical specialities that require a deeper insight into the understanding of real-time ultrasound imaging. Specifically, by taking an AI segmented output on the image and applying a texturised image to help characterise the image from a pictorial anatomical level, we can provide better real-time training while clinicians perform their scans.