A team of researchers were able to develop a machine-learning algorithm to create a system that accurately rates a surgeon's suturing skill level. The team from the University of Georgia published their study to the preprint server "arXiv."

To begin the development of the system, the researchers filmed 41 surgeons and nurses as they sutured test boards made of foam. Each of the subjects wore accelerometers on their hands to catch all of their movements, TechXplore reported.

Afterwards, the team showed the videos to a clinician who provided the ratings for the skill level of the subjects. The video was then fed to a computer with machine-learning algorithm along with the clinician's scores. This gave the system a basis on how to rate the work.

Lastly, the researchers removed the clinician's scores from the system and tasked the computer to rate the surgeon's suturing skill level by itself. The new system was found to be 93.2 percent accurate in matching the rating of the clinician's scores.

According to New Scientist, the team found that the best surgeons, or those with high suturing skill levels, moved their hands in smooth synchronization and consistency with every stitch. Those who had low scores moved in an erratic and less predictable manner, probably similar to Doctor Strange after his car accident.

Aneeq Zia, who was from the Georgia Institute of Technology and was part of the study, noted that including accelerometer data actually had less impact on the results than what they expected. It even made the system a little less accurate at rating the task. This could be because the video included information about how surgeons moved their hands and tools but accelerometers could only track movements of their hands.

Zia added that they hope later versions of the system can give trainee surgeons better feedback on their skills, specifically in suturing. This could also remove the need of having an expert observe them every time since these doctors are busy with their patients as well.