Robotic surgery training has long depended on human experts to judge surgical skill, a process that is subjective and hard to scale. A new preprint on arXiv introduces SurGPT, a foundation model designed to handle both skills assessment and gesture recognition in a multi-task framework.
The authors argue that automating these tasks could remove bottlenecks in surgical education. SurGPT is presented as a novel approach, though the abstract does not include experimental results or comparisons with existing methods.
As a preprint, the work has not yet undergone peer review. Its significance lies in the potential to standardise evaluation and provide scalable feedback in surgical training.