Techno-Urology
Meeting

Over the years, the Techno-Urology Meeting has explored how emerging technologies can make surgery increasingly personalised, precise and responsive to the characteristics of each individual patient.

Three-dimensional anatomical models were first introduced to help surgeons understand complex anatomy and plan the operative strategy before entering the operating room. These models subsequently evolved into patient-specific digital twins, immersive collaborative environments and augmented-reality tools designed to bring anatomical information closer to the surgical field.

Artificial intelligence represented the next step in this progression. During the previous editions, AI was used to analyse clinical and anatomical data, generate a patient-specific surgical profile and support the selection of the most appropriate operative strategy. This concept became the central theme of TUM26, “The AI-guided Surgical Profiling Era,” and was applied across different robotic procedures and platforms.

TUM27 aims to move beyond the idea of AI as a single assistant providing recommendations. The new objective is to create an integrated and interactive environment in which different AI tools, patient-specific data and visualisation technologies work together throughout the surgical workflow.

This transition is captured by the concept of the Agentic Surgery Ecosystem: an ecosystem in which artificial intelligence can understand the surgeon’s request, identify the appropriate digital resource, perform a predefined action and return the required information in a clinically meaningful form.

At the centre of this ecosystem is the Virtual Bioengineer.

Starting from the patient’s CT or MRI images, an AI-powered workflow can support the rapid generation of a detailed three-dimensional model, with the organ, tumour, arterial and venous anatomy, collecting system and other relevant structures segmented and organised as independent digital components. The aim is to progressively reduce the complex technical work traditionally required to produce patient-specific models, moving towards a streamlined process in which an anatomically detailed digital twin can be created and made available to the surgical team with minimal manual interaction.

However, the real innovation does not lie only in the automated creation of the model. It lies in how the surgeon can interact with it.

Through a conversational interface, the Virtual Bioengineer is designed to receive questions or commands in natural language and translate them into dynamic actions on the three-dimensional anatomy.

The surgeon may ask the system to rotate the model, isolate a specific artery, hide the renal parenchyma, highlight the relationship between the tumour and the collecting system, display a vascular territory or recreate the anatomical perspective expected during a particular surgical step.

The response is therefore not limited to a textual explanation or a predefined video: it becomes an immediate visual interaction with the patient-specific anatomy.

The same interaction can accompany different phases of the procedure. During preoperative planning, it can help the surgical team explore anatomy, compare approaches and define the operative strategy. During surgery, it can retrieve and display the anatomical information required at a specific moment, supporting orientation and communication within the operating room.

The development pathway discussed with the bioengineering team therefore combines automated model generation, granular anatomical labelling, real-time 3D manipulation, natural-language interaction and vocal feedback within a single environment.

The visible interface may be represented by a digital avatar acting as an AI copilot, while the underlying agent coordinates the tools and data required to answer each request. Its role is to make advanced bioengineering support faster, more accessible and more responsive to the surgeon’s needs, while preserving human supervision and clinical responsibility.

TUM27 will therefore explore the evolution from AI copilot to the Agentic Surgery Ecosystem: a connected environment in which artificial intelligence no longer simply analyses the surgical case, but helps generate, organise and interactively present patient-specific information across the entire surgical workflow.

The surgeon no longer adapts to the 3D model. The 3D model responds to the surgeon.

Francesco Porpiglia.