Over one million operations are carried out in Switzerland annually. The skill of a surgeon has a direct effect on the operation result. Experience and training, as well as temporary fatigue and other influencing factors, collectively play a crucial role.
Currently, skill is validated by experts, either by assessing video footage or directly at the time of an operation. But this method is very expensive and only a few experts are available. Furthermore, the evaluation is likely to differ and is not invariably completely reproducible. For some time, efforts have been made to objectify and automate the evaluation of the skills of surgeons.
Proof of Feasibility
The main outcome of the research work is the evidence of the underlying possibility of artificial intelligence (AI)-based evaluation of a surgeon’s skill in the framework of a surgical procedure. The AI utilized in the analysis determined good or moderate surgical skill with a precision of 87%. This could be regarded as an excellent outcome.
What was surprising was the high degree of algorithms' accuracy with the selected method. Our method of assessing surgical skills is based on the analysis of instrument movement. Surgical instruments were identified using computer algorithms and their movement was analyzed during the time period.
Joël Lavanchy, Study Lead Author, Bern University Hospital
Innovative, Three-Stage Approach with AI
A newly developed, three-stage method was utilized by the researchers. The analysis was based on 242 videos of procedures that involved laparoscopic removal of the gallbladder. The initial step was to identify the devices used.
To achieve this objective, the researchers trained a convolutional neural network (CNN) to identify the instruments. In the next step, the movements were examined, and their patterns were obtained. In the final step, the movement patterns, which were extracted, were correlated with the rating results by experts through linear regression.
Broader Database and in-Depth Training of Algorithms is Needed
The new study represents a major first step toward evaluating surgical performance. However, more detailed steps are required before AI technology can be utilized in clinical settings.
On the one hand, there is a need for AI algorithms to be trained on a wider database to additionally enhance the recognition of instruments. On the other hand, more surgeries need to be examined and, in the medium term, videos of open surgeries and also procedures, in addition to the abdominal region, had to be tackled.
AI has mainly been used thus far to identify instruments or specific surgical phases. In our study, we now assess surgical skill based on surgical videos. In the future, the use of AI can solve problems at multiple levels: it is available on-demand peri-operatively (not dependent on a few hard-to-find experts); it is objective using algorithm-driven standards.
Dr Enes Hosgor, Study Co-Author and Leader of AI division, Caresyntax
Dr Hosgor added, “it is comparable at a transregional level as well as surgeon level and could thus provide important support for decision-making processes at certification institutes.”
Caresyntax, a medical technology company headquartered in Boston and Berlin, categorizes the outcomes.
AI at Medical Location in Bern: CAIM as an Opportunity
The study offers a significant indication of the upcoming development of the use of AI in the field of medicine. Going forward, it will move from the recent assessment of image material to the provision of expert systems.
The study is a first step. Now that we have demonstrated the fundamental feasibility, we can start planning assistance systems that will support surgeons during operations. For example, they will be alerted when fatigue is detected, thereby helping to prevent complications.
Guido Beldi, Professor and Head, Bern University Hospital
Lavanchy, J. L., et al. (2021) Automation of surgical skill assessment using a three-stage machine learning algorithm. Scientific Reports. doi.org/10.1038/s41598-021-84295-6.