MIDL Vision 2020

Women in Science 20 Qi Dou is currently an Assistant Professor in the Department of Computer Science and Engineering at The Chinese University of Hong Kong. She is co-affiliated with T Stone Robotics Institute at the university. Qi, can you tell us more about your work? I am a new faculty here. I just joined starting in February of this year. I have been involved in the research field focusing on AI for medical image analysis for the past five years. What I focus on is automated detection of the lesions on medical images, for instance, CT andMRI images. Also, I do tasks of segmentation for lesions, which can be used for the quantification of the disease, for the diagnosis part and intervention. That is very meaningful. In terms of the technology part, applying computer vision methodologies into the medical image analysis area is not straightforward because these two areas have many differences. Specialized techniques need to be tailored and adapted to be fitted into the specific problems in the healthcare domain, especially medical images. Part of my work is to do this kind of technological development for medical data. One representative work is about 3D Convolutional Neural Networks. It was my first project in PhD studies five years ago. I think that work was very meaningful as we tried to figure out how to more effectively extract the high dimension of the volumetric information in the medical data, because that was different from 2D images or 3D images. Then we continued to go deeper into 3D Convolutional Neural Networks. I’m very excited to do more work in computer robotic surgery, like minimally invasive surgery, so that procedures can be less invasive and safer, and the patient can have a better recovery. In minimally invasive surgery, the surgeons face more challenges than in open surgery. What I’m driving now is to bring more intelligence into the robotic surgery “I felt I had the capability to learn the skills.”