HMN 2026: How Can virtual reality train surgeons? A 2,000-year-old experiment reveals what’s missing

As medical schools increasingly turn to virtual reality, 3D models and digital simulations, a new correspondence in Nature Medicine argues that one essential part of clinical training remains difficult to digitize: the hands-on judgment physicians develop through real-life, mentored practice.

Led by Prof. Orly Lewis of the Hebrew University of Jerusalem, the correspondence highlights a growing challenge in modern medical education. While digital tools can teach anatomy, procedures and decision-making with increasing sophistication, they often struggle to transmit what clinicians call “tacit learning,” the physical intuition needed to know how much pressure to apply, where to place one’s hands, when tissue feels wrong or when to stop.

To expose this gap, Lewis and a multidisciplinary team in Jerusalem turned to an unexpected source: Galen of Pergamum, the influential Roman physician of the second century CE. The team attempted to recreate a dissection of the abdominal wall and peritoneum of a female pig specimen by following instructions from the second book of Galen’s “Anatomical Procedures.”

Galen’s text is highly detailed. It instructs readers, for example, to cut superficially around the umbilicus until the underlying tissue appears “whiter” than surrounding structures. But when the researchers tried to follow the ancient directions in practice, they encountered a familiar modern problem: The instructions were explicit, but the skill was not.

Why written detail falls short

What does “superficially” mean under a scalpel? How much pressure is enough to expose tissue without damaging or distorting it? How should the hand move when resistance changes? These are the kinds of judgments that cannot be fully captured in a manual, a checklist or even a sophisticated simulation.

“Even Galen’s highly detailed written instructions weren’t enough to guide us, and today’s augmented reality simulations face the same challenge,” said Lewis, principal investigator of the ATLOMY project at the Hebrew University of Jerusalem.

“We can program the visible steps of a procedure, but simulations still struggle to transmit the physical intuition of an experienced physician.”

The stakes for medical training

The correspondence argues that this problem has become especially urgent as medical training increasingly relies on digital resources while opportunities for physical dissection, bedside teaching and supervised clinical practice decline in some settings.

In contemporary competency-based training, complex procedures are often broken down into discrete steps that can be taught, measured and assessed. But the authors warn that this approach may overlook the unspoken elements of expertise traditionally passed from teacher to student through observation, imitation, correction and repeated hands-on practice.

Whether a trainee is placing a catheter, making an incision, palpating tissue or navigating anatomical variation, clinical competence depends not only on knowing what to do, but on developing the sensory judgment to do it safely and effectively.

A call for better tools

The authors do not argue against digital innovation in medical education. Instead, they call for a more rigorous approach to designing these tools. If virtual reality, augmented reality and digital simulations are to replace or supplement hands-on training, they must identify and incorporate the tacit dimensions of medical skill acquisition, including touch, pressure, resistance, timing and embodied judgment.

The lesson from Galen, the authors suggest, is both ancient and urgently modern: Medical knowledge has never been transmitted by words or images alone. Even the most detailed instructions leave something out—and that missing knowledge may be exactly what future doctors need most.

The content is provided for information purposes only.