At the renowned AI for Surgery lab in Baltimore, Maryland, a STAR was born. The Smart Tissue Autonomous Robot (STAR) was brought to artificial life by Axel Krieger and colleagues at Johns Hopkins University. STAR can perform complex keyhole anastomosis surgery with minimal human intervention with astounding accuracy, helping to overcome the challenges of these delicate procedures during which two pieces of small intestine are sewn into a continuous section.
According to Krieger, the robot uses a combination of novel suturing tools, imaging systems, machine learning algorithms, and robotic controls. STAR isn’t meant to replace human surgeons – it’s designed to be incorporated into the surgical workflow, enhancing surgical consistency from patient to patient. For now, it’s an impressive research project, but it illustrates the direction of travel.
Fast-forward a decade, and imagine, under the supervision and guidance of a surgeon, sophisticated robots that could plan and complete autonomous surgeries based on an individuals’ pathology and needs. The robots could revise plans in real-time, responding to complications and optimizing outcomes. However, a world where constellations of STARs operate under human supervision looks very different to the current surgical model. And if the anticipated time horizon is even vaguely true, it’s a world we need to start preparing for today.
A shrinking timeframe, fuelled by AI
Autonomous surgical robots promise to improve consistency, patient outcomes, and access to standardized surgical techniques. By 2033, the global market for autonomous surgical robotics is expected to reach $11.07 billion. While widespread autonomous surgery (particularly soft tissue surgery) is further down the line, it’s not as distant as some might think. In the last three decades, led by the more predictable environment of orthopaedics, researchers and developers have taken tentative steps towards autonomous surgery. But in the past couple of years, the rapid maturity of AI solutions, from every-day Gen AI applications to self-learning systems, has moved the dial. These technologies, alongside robust communications networks, are long-term enablers for autonomous surgery. Large Language Models (LLMs), for example, can now watch video content, draw inferences, and replicate human behaviours.
The knowledge, tools, and confidence already exist to support autonomous systems at scale. However, with the anticipated timeframe for autonomous surgery shrinking daily, healthcare and medtech organizations are being prompted to consider entirely new directions.
Critical structures, frameworks, and incremental steps are needed to build a world where robots perform automated procedures in harmony with human surgeons. And if the timeframe is 10 to 20 years rather than 50, there are fundamental considerations for decision-makers today. On the cusp of so much transformative change, what happens next? What far-reaching implications will autonomous systems bring? How can the global healthcare landscape prepare for an autonomous surgical future?
Understanding the impact of autonomous surgery
Research suggests that robots could perform surgical tasks 50 times faster than a human surgeon. If the work of 50 surgeons can be completed by a single robot, human professionals could be freed to focus on critical event resolution, the most complex procedures, and accelerated training and competence at global scale.
Faster and more accurate surgery could improve accessibility for the half of the world that currently lacks consistent, quality healthcare. Automated solutions could take telesurgery to a whole new level, aided by lightning-fast global communication so human professionals don’t need to be physically present alongside systems.
Procedure innovation is another advantage. Most surgical instruments are designed for use by humans with two arms, two hands, and two eyes. But autonomous systems could use multiple tools to complete several tasks at once, massively changing the nature of procedures. For example, while removing a piece of cancerous tissue, surgeons currently use a stapling device that lays two straight lines of metal staples to help close the excision. If an autonomous robot was released from the constraints of human control, perhaps metal staples – prone to tearing – might be supplanted by a super-fine dissolvable suture line, woven in situ. Like hemming a skirt, the cut profile could follow the shape of the tissue, supporting a faster, more effective procedure.
All these changes impact cost dynamics. For instance, if a robot can perform faster, streamlined surgery without extensive human guidance, as surgeons get redeployed and volumes increase, the resource balance changes. Do robotics providers charge the same fees as consultants, or does a competitive market dramatically drive down procedure costs? And if it does, the decision on whether and when to operate changes too. This could open the door to earlier and more widespread intervention.
The regulatory landscape
As promising as this future is, autonomous surgical tools need new guardrails and careful regulation. The autonomous future requires a regulatory approval process built to accommodate iterative, near-continuous learning. Let’s say a surgical tool uses a black box AI product to gather sensitive data during procedures and then applies the insights without explaining how. If a mistake harms a patient, there’s no way to track how or why. Not only this, but if the insights feed a shared database, the mistake could be multiplied.
Currently, regulators check updates to systems before release. But in a world of continuous-learning machines, a new model is needed. The recent release of guiding principles around Good Machine Learning Practice (GMLP) is just the beginning. Imagine a ‘dynamic validation system’ for AI platforms that checks whether insights are safe and worth implementing. Constant data-harvesting would need daily if not hourly validation cycles, calling for new processes, deeper integration and even a new management function. Chief Information Officers may need to work hand in glove with Chief Technology Officers to take on an important gatekeeping role, forging close partnerships with regulators.
Mapping the autonomous future
Between now and a fully autonomous future lie many smaller steps, but if organizations aren’t planning for this eventuality, are they risking disruption from other players? Stop-gap solutions are important, but if fully autonomous surgery is one to two decades away, is it right to invest in systems which may be obsolete within a handful of years? How do R&D decision-makers place the right bets, and how can they ensure their organization is ready?
To prepare for an autonomous surgical future, medtech leaders should explore the practical implications of potential scenarios, and structure investment accordingly. The key is to develop a living roadmap that drives decisions on a day-by-day basis, drawing on macro- and micro-trends. Conducting these assessments as part of the day job will indicate the actions needed to prepare for various possible futures. With a roadmap in place, companies can monitor, update, and optimize their strategies.
When it comes to autonomous surgery, one crucial early-stage action is exploring how to leverage AI and feed it with relevant data from disparate parts of the care pathway. The surgeon isn’t just responding to what they see, they are informed by patient history and personal experience. Accessing and understanding the relevance of this data ecosystem will help build effective models, giving context to the decision engines that will ultimately control outcomes. Armed with data, digital twins can be used to predict how systems will respond and help to optimize algorithms in a safe environment, offline from the patient.
These foundational steps need to be linked within an ecosystem of joined-up stakeholders. Medtech companies have already forged partnerships with major tech players, while other companies are actively pushing medical AI accelerator programs in the healthcare space. The wave of activity will continue.
Through a blend of horizon-scanning and road mapping, strategists in the medtech and healthcare sectors can prepare for the ‘reverse ripple effect’ of tomorrow’s changes on today’s decisions. These methodologies should be as central to a company’s operation as monthly accounts – especially given the speed of technological evolution.
Editor’s note: Neither the author, nor his company is affiliated with any of the entities mentioned in this article.
Photo: gorodenkoff, Getty Images
Alistair Fleming brings over 25 years’ experience in the field of MedTech, delivering ground-breaking solutions to clients. Working with an array of technologies, including imaging, surgical robotics and molecular diagnostics, Alistair has helped develop solutions for lung cancer, orthopedics, general surgery, urology, gynecology and diabetes in the US, UK, Germany and Japan.
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.
Source link
