Today’s robots are improving surgical precision and enhancing patient recovery. Future developments, like machine learning and augmented reality have the potential to “transform surgical practice as we know it1.” This recent review examines the latest innovations in surgical robots, and their current and future role in orthopaedics1.
Robots can be classified in one of three categories
1) Computer-assisted surgical systems (CAS) – computer guided navigation, which provides the surgeon with additional data to better inform their surgical decision making1
2) Robot-assisted systems (RAS) – an extension of CAS to provide “precision-guided robotic instrumentation”. RAS enhance precision by constraining the surgeons’ actions to their pre-set targets. They remain under the surgeons’ control, but help the surgeon to accurately achieve their planned cuts and implant alignment1
3) Autonomous systems – perform the surgery independently of the surgeon. Once the surgeon has set up the workflow and parameters for the surgery, the robot completes the surgery (bone cuts) autonomously1
Why is robot-assisted surgery growing so quickly?
The recent commercial success of the RAS was attributed to evidence that they can improve surgical workflows and outcomes. The authors cited a meta-analysis by Argawal et. al, which observed that robot-assisted total knee arthroplasty improved HSS and WOMAC scores when compared to conventional instruments1. They also noted that RAS has been shown to increase precision, reduce the incidence of mechanical malalignment, and improve patient recovery times – enhancing the overall patient experience1. It was acknowledged, however, that not all the literature has shown significant differences between RAS and conventional instruments; leaving the authors to speculate that other factors – like patient and surgeon preference for robotic knee arthroplasty – are also driving the adoption of RAS1.
The future of robotics: machine learning
The most common application of Machine Learning (ML) in medicine is in classification analysis. This is a way of predicting an outcome from a rigid set of choices1. While the data analysis in this process is very complex, the outcome tends to be a simple question – given some specific information, can the data be separated into distinct categories1? In knee arthroplasty, such a classification analysis might help to separate patients indicated for partial knee replacement from those that are not. The ‘decision’ function is learnt from patterns observed from previous examples. Once learnt, these patterns are used to make future predictions1.
While current applications are limited, ML is expected to play a big role in the future of robotics1. Pre- and intra-op it will help surgeons with real-time diagnosis, predicting the prognosis of treatment, and making recommendations for the next surgical steps1. Post-operatively it can track outcomes and ‘learn’ lessons that can be used to further refine ML and inform surgeon training1.
The future of robotics: Augmented Reality
Surgical robots will also leverage advances in Virtual and Augmented Reality1. While Virtual Reality (VR) has been primarily used as a training tool, it has been suggested that Augmented Reality (AR) might have a wider application in RAS1. By overlaying information onto reality, AR can be used intraoperatively to virtually indicate relevant surgical sites/landmarks, and provide real-time information about the technical steps of the procedure or the plan for a specific patient1. It could also “serve as the visual interface for ML driven surgical recommendations” in the future1.
Potential challenges to the adoption of RAS
- Team dynamics in the OR – where the theatre team and the way that they interact and assist the surgeon during surgery change with the introduction of RAS1
- Learning curve – while the learning curve for performing robotic-assisted surgeries is unknown, and may vary by manufacturer/work flow, surgeons and theatre teams should anticipate taking some time to adjust to RAS1
- Cost-effectiveness – as with all emerging technologies there is a need for more data to prove out the cost-effectiveness of RAS1
While more research into the cost-benefit of RAS in knee arthroplasty is needed, the review concluded that: “…the age of robotics is coming and surgical units should begin now to re-align training practices and establish expertise in readiness to welcome the incoming era1.”