AI for Robotic Surgery: Tse'ela Mida of Intuitive

In this episode

• da Vinci doesn't operate on its own — the surgeon sits at a console and controls four robotic arms; the robot doesn't tire or tremble and adds degrees of freedom and 3D vision, which translates into fewer complications, shorter hospital stays and lower costs.
• Wherever there's data there's an opening for AI: every robotic surgery is recorded automatically, producing video and robot data that are used to train models that come back to the surgeon to help them train and improve.
• Success metrics range from adoption and utilization (did the surgeon even connect to the system and access their data) to proving clinical value to the hospital — hospitalization days, bleeding events, readmissions — versus open and laparoscopic surgery.
• Medical regulation is about efficacy and safety, and a product that can't prove both "has no justification"; AI-specific regulation demands transparency — how the model was trained, how bias is prevented and how it's monitored — and the clinical decision stays with the doctor.
• Medical data is a barrier to entry for new players; Tse'ela built a "Data Factory" — a heterogeneous team with medical consultants and a clinical team that undergoes intensive training to label anatomy and surgical steps, both for model training and for quality control.
• The product skills she considers most important: strong communication at eye level and in each side's language (developers vs. doctors), the ability to move people to action, and focus — separating the essential from the secondary and prioritizing.