Computer Vision News - June 2019

35 Computer Vision News Women in Science The first international workshop on surgical data science was created in 2016 when we saw more and more success stories in all sorts of fields of data science, for example, radiological data science. We realized that we don’t see these success stories in surgery. We figured that we should try to advance this field, identify the challenges, and define what we actually mean by “surgical data science” (in comparison to biomedical data science). Together with my colleagues, Stefanie Speidel and Pierre Jannin , we organized the first workshop where we invited leading people from the field of computer- assisted surgery, robotics, etc. We had an interactive workshop to define what we mean by surgical data science, what the problems are, and how to move the field forward. At the beginning of the event, we performed an anonymous voting: we had about 70 or 80 participants and we asked them for their opinions and why they were attending. There was quite a big portion of people that said: “ I’m here because my supervisor made me [attend the workshop]! ” [ smiles ] After the workshop, when we asked people anonymously again, everyone said that they would like to do it again. There was 100% agreement. Most of them voted for meeting annually or every two years. That was a success for us. People enjoyed the workshop after having attended it: and the workshop resulted in a Nature Biomedical Engineering paper 2017. Now we will have the second edition of the workshop. Personally, I don’t know what the low hanging fruits related to surgical data science are. Nobody seems to know. A reason may be that surgical data science is comparatively complex from an infrastructure point of view. In radiological data science, for example, you have digital data anyway. And you have the radiologist who makes a diagnosis. All you have to do is place the algorithm between the two, right? And you can annotate the data that is digitized anyway to train the algorithm. After that, you can show a new radiological image to the algorithm to support the physician. In surgery, the situation is so much more complex. It starts with the fact that most of the information that is used by the physician is not digitized. Also, there is a whole team involved: the surgeons, the patient, the anesthesia team, the nurses, and maybe robots. All of them are involved in manipulating the patient and even making decisions Photo: Dave Dove Lena Maier-Hein “ A good network of friends is so helpful when combining family and science! ” “ I’m here because my supervisor made me! ” “ In surgery, the situation is so much more complex… ”