Computer Vision News - June 2021

329 Julie Shah than the human expert that diagnoses cancer. That can help fill in the gaps of the errors that a machine will naturally make, which will be different than people. Think of it like a swiss cheese model! [laughs] Each of us has our biases and errors that we make. If we put it together the right way, we can cover each other's gaps. That’s how you can achieve more effective capability. We have a really interesting turning point in surgery with the introduction of the new robotic surgery systems, the Da Vinci and other systems. They are still primarily tele-operated by the surgeons. The surgeon is sitting at a console, removed from the patient. They are basically controlling every joint of that robot through this fancy, remote control interface. They might be in another continent. Yes, with a time delay. There are researchers that are working very actively on that very remote form of conducting surgery. There are many interesting challenges on the stability of the control system under time delay. These new surgical systems introduce a computer between the surgeon and the patients. With surgical robots, you have a computer between the surgeon and the patient. This is the beginning of an enormous opportunity in enhancing how we go about surgery, improving safety and improving outcomes. That’s quite exciting. Computer vision is a key enabling technology in that direction of work. Did you do any research in this area yourself? My own work in healthcare focuses more on supporting decision-making. For the past few years, I’ve worked on a project with Beth Israel Deaconess Hospital in Boston. We studied the decision-making strategies of very senior nurses on a labor delivery floor. These nurses are called charge nurses or resource nurses. They’re like the nurse managers of the labor floor. It was very interesting for me to learn and understand that these nurses are essentially doing the job of an air traffic controller. They’re deciding which patient goes to which room,