The Medical Electronic Command Center (MedECC) infographic by Alice Chao
I have been leading a team of fantastic collaborators on a telemedicine crisis management web application for the COVID-19 pandemic. Our platform, “MedECC”, is targeted at low SES hospitals that must cope with local spikes of cases. MedECC has many benefits for these stressed hospitals thanks to countless volunteer hours worked by the team. We are happy to have been able to share our work through the Computing Research Association (CRA) Virtual Conference 2020. Founded in 1972, the CRA strives to enhance innovation in computer science research and facilitates contributions from industry, government and academia. You can see our 3 minute lightning talk under the social monitoring category here.
View from the Vancouver Conference Center during NeurIPS 2019
I was fortunate to have my work on an AI based Patient Specific Physical Simulation method for accurate surgical image guidance accepted for the Black in AI workshop at NeurIPS 2020. It was my first time attending the conference which was an absolutely amazing experience! I met a great many brilliant scientists specializing in a diverse range of topics and hopefully forged lasting new connections. With 16,000+ attendees and the research community being small I also bumped into many old friends and it was great to catch up. I certainly hope to have the chance to present new work next year!
Between August and September 2019 I had the pleasure of working with Professor Nabil Alsharafa and his fantastic research team in the HAbits Lab. The HAbits lab is working on the cutting edge of computer assisted preventative medicine with a focus on applied AI and “mHealth” IoT applications. My stint with the lab was both enjoyable and productive, with one R21 grant submission and three paper submissions including a successful workshop submission to NeurIPS on AI enhanced patient specific physical models for AR surgical image guidance. We also had an accepted research poster at SAGES 2020 on novel methods for training open surgery technique using portable electronic devices.
Dalian University of Science and Technology International Campus
From June until early August I was lucky enough to research in the Computer Science department of the Dalian University of Technology in northern China. 2019 is auspicious for DUT which celebrates its 70th birthday this year. I was generously hosted by Professor Na Lei who extended every hospitality to make my stay very pleasant. It was a productive couple of months where I submitted three first author papers. One was accepted at the “International Symposium on Visual Computing” (ISVC 2019) with the other two currently under review.
I was very fortunate to be invited to the Department of Chemical and Biological Engineering at the Hong Kong University of Science and Technology to give a research talk in Feb 2019. The seminar outlines novel Augmented Reality (AR) techniques that can be applied to patient specific surgical image guidance and is titled:
“Enriching Augmented Reality Assisted Surgical Image Guidance With Efficient Patient Specific Physical Simulation“
View from Ocean Park Tower, Hong Kong
Talk Abstract
“Rapid advances in computing performance and novel sensors have opened many exciting avenues in medical technology research. Advanced imaging techniques have found many medical applications and Augmented Reality (AR) is a case in point. In this talk we look towards the future of image guided surgical interventions with a focus on AR applied to liver procedures. Although recent breakthroughs such as Minimally Invasive Robotic Surgical tools (MIRS) have dramatically increased possible surgical precision, this potential is greatly underutilized because of limited MIRS video camera visual systems. Ideally, the surgeon would use super hero like see-through vision to precisely guide their tools instead of guessing what lies beneath the surface of the organs they see. For example in conventional liver procedures it is difficult for surgeons to avoid unnecessary damage to the internal vessels which are hidden beneath the livers visible surface. We will discuss how AR can offer a see-through style image guidance solution and the steps needed project hidden landmarks such as blood vessels onto the surgical scene. We will drill down into the details of especially difficult problems and how our research innovations have made great strides to address them. In particular we focus on computationally efficient techniques for: Tissue surface tracking with real time non-rigid 3D registration. Our work opens new opportunities to explore AR speed and accuracy trade-offs and we close with a discussion on our ongoing and future work in this space.”
The Hong Kong University of Science and Technology
HKUST campus
Founded in 1991, HKUST has rapidly matured into a prestigious University famed for its high quality research. Situated in Kowloon bay, HKUST boasts a beautiful and unique campus that descends from lush green hills into the ocean.
Hello, I’m Michael Barrow, a researcher currently based in San Diego California.
I have many interests but recently I have been focusing on medical technology applications. In particular my PhD thesis has focused on novel computationally efficient surgical image guidance. Right now I’m at a turning point in my work. Looking ahead I am exploring other applications of technology to medicine and social problems. Although some exciting projects are on the horizon, I am certainly open to interesting opportunities in my areas of interest and expertise.
Below are some selected research projects I’ve worked on
Vision Chip
I worked on a novel computer vision hardware accelerator designed to “Register” 3D surfaces in real time. Real time 3D surface registration is useful for tracking soft object surfaces over time (for example tissue in surgical procedures). This used hardware-software co-design to implement a novel “TRW-S” based global registration.
Previously I was co-director and helped develop curriculum for the Early Scholars Research Program (ERSP). ERSP is a novel undergraduate program aiming to retain Computer Science majors identified as “at risk” of not matriculating. ERSP focuses on Under Represented Minorities which have higher attrition rates.
Some ERSP students (see above) and I worked to create a data set of canned surgical 3D video. The intended use case was to aid in developing high quality algorithms for accurate Mixed Reality image guidance in surgery.
Prior to my PhD, I worked on a hardware image compression energy reduction technique for image sensors. The work focused on novel near sensor computing techniques to reduce the energy cost of image acquisition in mobile devices.
