AI-based assessment of surgical skills

More than a million jobs are created in Switzerland each year. The skill of a surgeon has a direct impact on the outcome of the operation. Training and experience play a part, as well as instant fatigue and other influential traits. At the moment, skill is proven by experts, either directly during work or by evaluating a video film. This procedure is very expensive and only a few experts are available. Furthermore, the assessment may vary and is not always fully reproducible. For some time, attempts have been made to automate and challenge the assessment of surgeons ’skills.

Proof of feasibility

The main outcome of the study is the confirmation of the basic feasibility of artificial intelligence-based (AI) assessment of a surgeon’s skill in the context of a surgical procedure. The AI ​​used in the study identified good or moderate surgical skill with 87 percent accuracy. This can be considered a very good find. Lead author Joël Lavanchy explains: “What was amazing was the high level of algorithm accuracy with the chosen method. Our method of assessing surgical skills is based on instrument movement analysis. . Surgical instruments were identified using computer algorithms and their movement was analyzed. During the period. “

An innovative, three-tier approach with AI

The research team used a newly developed three-step approach. The study was based on 242 videos of laparoscopic gallbladder removal procedures. The first step was to identify the instruments used. For this purpose, a controversial neural network (CNN) was trained to recognize the instruments. In the second step, the movements were studied, and their patterns were extracted. In the third step, the extracted movement patterns were linked to results graded by experts using linear regression.

There is a need for a more comprehensive database and in-depth training on algorithms

This study is an important first step towards evaluating surgical performance. More in-depth measures are needed before the technology can be applied in clinical practice. For one thing, the AI ​​algorithms need to be trained on a broader database to further recognize the recognition of instruments. For another thing, additional surgeries need to be explored and, in the medium term, videos of open surgeries can be treated as well as procedures in addition to the abdominal area.

Dr. Enes Hosgor, co – author of the study leading the AI ​​division at caresyntax, a medical technology company headquartered in Berlin and Boston, categorized the results as follows: “AI has been used mostly to This is to assess the skill of surgery based on surgical videos.In the future, the use of AI can solve problems at several levels: it is available on demand peri-operatively (without reliance on a few difficult-to-find experts); it aims to use algorithm-driven standards; it is comparable at a cross-departmental level as well as level of surgeon and therefore could provide significant support to decision – making processes at certification centers. “

AI at a medical location in Bern: CAIM as an opportunity

The project provides an important signal for the development of the use of AI in future medicine. In the future, it will move from the recent evaluation of imaging products to the provision of expert systems. Professor Guido Beldi, head of the study, explains: “The first step is the study. Now that we have demonstrated basic capability, we can start design support systems that support Surgeons during work. For example, they will be notified when fatigue is detected, thus helping to prevent complications. “

The promotion of AI at the Bern medical center will contribute to further development when the Center for Artificial Intelligence in Medicine (CAIM) opens here on March 19th.