Computer Vision News - May 2016

Application Triton - Gauss Surgical COMPUTER VISION NEWS 4 We all know that blood transfusions save many live s and most of us know somebody who can personally testify in favour of this statement. What is less known is the other side of the coin: many transfusi ons are performed even though they a re not needed, carrying with them un necessary potential risks of reactions , errors, infection, respiratory dis eases and even cancer recurrence. So me studies estimate that up to 60% of RBC (red blood cells) transfusions a re inappropriate , at a very high cost for the health system, the medical pe rsonnel and the patients. Why does this happen? Believe it or not, it seems that most decisions to transfuse are taken following visual estimation of blood loss (EBL) during surgery, whic h proves to be highly inaccurate. Th e Joint Commission , a recognized no t-for-profit organization having as a mission to improve the public’s heal th and safety, has mandated that blood loss be quantified at the end of every procedure and listed in the patient’s medical records . Nonetheless, no measurement technique exists besides the visual estimation in use since the beginning of medicine. Without an adequate measure of lost blood, it is difficult to take a proper decision regarding whether a blood transfusion is necessary and to what extent. Physicians currently rely on proxy metrics and visual assessments, which are inaccurate and subjective. During surgery, blood loss is typically captured by sponges made of absorbent fabric and by suction canisters, clear buckets that collect fluids suctioned from the patient during the surgical procedure. Physicians looking at the sponges to determine how saturated and soaked they are end up debating the extent of the blood loss and the final decision comes at the conclusion of some sort of negotiation. We decided to discuss those findings with Siddarth Satish, founder and CEO of Gauss Surgical , a medical technology company based in Los Altos, California. Satish, trained as a chemical engineer and having worked years in biophysics and computational modelling, remembers how puzzled he was when he had the opportunity to observe a surgery for the first time: is it possible that in the era of robotic surgery, we are still guesstimating a key medical parameter like blood loss? It seemed clear to him that this practice, never solved so far, represented a key opportunity to improve patient safety and quality, while at the same time reducing the costs of healthcare. The belief that computer vision techniques would provide the computational solution to calculating blood loss led to the development of Triton, a mobile platform for real- time monitoring of surgical blood loss , having four FDA clearances along with the approval for CE Mark. “ The Triton solution is a mixture of computer vision and mobile devices ”