Computer Vision News - January 2020

2 Summary Medical Imaging 1 Kristin tells us that the SUSI workshop was the first time people had gathered more generally across the ultrasound domain, including general imaging, women’s health imaging, hand-held support care imaging and cardiovascular ultrasound. It was a chance to present the latest works in an arena where they hadn’t typically had much coverage. Participants could see where different technologies overlapped and leverage the synergies between modalities. There are pros and cons of ultrasound, but thanks to the latest advancements in medical imaging and image processing technologies, Kristin says we’re starting to work towards more of the spatial resolution that is typically seen in CT and MRI. With greater use of AI to augment the imaging, we’re seeing much better image quality from ultrasound. However, there’s been a lack of standardisation in ultrasound compared to other imaging modalities. With CT and MRI, there’s quite a standard way to acquire the images, but with ultrasound images you have to find an acoustic window between, for example, the ribs or other obstructing objects in the foreground. Thanks to developments in terms of guidance and standardisation within Smart Ultrasound Kristin McLeod is the AI and data science team leader at the cardiovascular ultrasound division of GE Healthcare. She was also co-organiser of the Smart Ultrasound Imaging (SUSI) workshop at MICCAI 2019 in Shenzhen. She speaks to us about recent and future developments in the ultrasound area. with Kristin McLeod