From Data to Decision

In the rapidly evolving landscape of vascular surgery, a silent revolution is taking place—one that shifts the paradigm from tradition-bound decision-making to a future where data and digital intelligence steer the course of patient care. The traditional reliance on the surgeon’s experience and clinical intuition, while invaluable, is no longer sufficient to meet the growing demands for precision, safety, and personalisation. As Professor Robert Bruce Salter pertinently stated, “Decisions are more important than incisions.” This philosophy is more relevant today than ever, as Artificial Intelligence (AI) and Digital Twin (DT) technologies reshape how vascular interventions are planned, executed, and evaluated.

At the same time, healthcare systems around the world are under increasing financial pressure, facing mounting costs and limited resources. There is a growing imperative to improve efficiency without compromising patient outcomes. In this context, leveraging AI and DT is not just a technological upgrade it is a necessity to optimise workflows, reduce unnecessary procedures, and deliver value-based care.

The Limits of Human-Centric Decision-Making

Historically, vascular interventions have been planned using a combination of anatomical imaging viewed through conventional medical imaging software, patient history, and the surgeon’s personal expertise. While this method has served well, it is inherently limited. Even the most seasoned practitioners are constrained by human cognitive capacity when interpreting complex 3D vascular structures and predicting dynamic outcomes. Variability in patient anatomy, comorbidities, and procedural complexities introduces uncertainty that is often underestimated.

In this high-stakes environment, relying solely on surgical experience is no longer adequate. With rising patient expectations, the need for standardisation, and increasingly complex devices, surgeons must augment their skills with digital tools that reduce variability and improve outcomes. Moreover, such tools offer the ability to enhance the reproducibility of procedures across institutions and practitioners, ensuring that high standards of care are consistently maintained.

Enter the Digital Twin: Precision at a Patient-Specific Level

A Digital Twin in medicine is a virtual replica of a patient’s anatomy, derived from high-resolution imaging data and enriched by real-time physiological and clinical information. When coupled with AI, these Digital Twins become predictive, allowing simulations of different procedural strategies and their likely outcomes.

In vascular interventions, this means simulating blood flow, stent-graft behavior, and potential complications such as endoleaks (persistent blood flow outside the stent graft) or thrombosis—all before a single incision is made. This empowers clinicians to make data-backed decisions, optimise device sizing and positioning, and minimise intraoperative surprises. In addition, clinicians can test several treatment scenarios to find the optimal balance between risk and efficacy.

Digital Twin technology transforms preoperative planning into a dynamic, patient-specific process. Each decision becomes a calculated move, grounded in simulated outcomes and enriched with contextual data.

AI-Powered Insights and Predictive Indicators

The integration of AI brings unprecedented power to the Digital Twin. Machine learning models, trained on thousands of previous cases, can recognise patterns, highlight risk factors, and predict procedural outcomes with increasing accuracy.

A compelling example is the Endoleak Risk Index (ERI), which estimates the likelihood of an endoleak following an endovascular aneurysm repair (EVAR). EVAR is a minimally invasive procedure to repair abdominal aortic aneurysms. The ERI, calculated from the Digital Twin and powered by AI analytics, allows clinicians to address complications in advance by adjusting device selection or strategy.

These indicators embody the shift from reactive to proactive medicine—where the aim is not only to treat complications but also to anticipate and prevent them. Digital Twin and AI technologies enable prediction of adverse events, prevention through customised planning, and personalisation by tailoring interventions to each patient’s unique anatomy and physiology.

Crucially, these tools are not a replacement for surgical judgment but a powerful augmentation—enhancing surgeons’ capacity to make informed, patient-specific decisions. The surgeon remains at the center of the process but is now supported by a network of computational insights, transforming decision-making into a collaborative human-machine effort.

Generative AI and the Future of Preoperative Planning

Looking ahead, Generative AI holds transformative potential. Unlike traditional AI models that classify or predict, Generative AI creates, producing new hypotheses, simulating scenarios, and helping design personalised treatment pathways.

Generative agents synthesise clinical knowledge, imaging, and patient data to propose optimised procedures—reducing cognitive load and enhancing preoperative planning. For instance, while planning an EVAR, a generative AI assistant could calculate the ERI, recommend stent graft configurations, forecast the impact of alternative trajectories, and propose mitigation strategies for anatomical constraints.

These agents continuously learn from new cases, guidelines, and outcomes. They can serve as virtual planning assistants—reviewing imaging, flagging anomalies, summarising history, and generating personalised preoperative reports. This improves accuracy, consistency, and relieves cognitive burden on surgical teams.

Generative AI can also facilitate cross-disciplinary collaboration. By translating complex datasets into visual dashboards or natural language summaries, it empowers not only surgeons but also anesthesiologists, nurses, and even patients to better understand and engage with the treatment plan. This democratisation of data aligns perfectly with the participatory pillar of 4P medicine.

In essence, Generative AI helps turn complex data into actionable knowledge. It serves as a responsive co-pilot, ensuring every decision is grounded in evidence, simulation, and predictive intelligence.

Companion Algorithms: The Future of Medical Devices

The next frontier lies in merging hardware and software. Devices such as stent grafts and catheters are increasingly designed alongside companion algorithms—intelligent software that analyses anatomy and hemodynamics, recommends configurations, and predicts outcomes.

This hybrid approach turns passive medical hardware into active decision-making participants. Manufacturers, regulators, and clinicians alike now recognise that safety and efficacy depend on integrating data intelligence into device use. Regulatory bodies may soon require simulation-backed validation of devices, pushing the entire industry toward more intelligent, accountable designs.

A New Standard of Care: Digital Preoperative Planning

Digital preoperative planning is becoming a requirement, not a luxury. Hospitals worldwide are adopting simulation platforms that reconstruct vasculature from DICOM images (standard medical imaging format), simulate interventions, and offer risk assessments in secure, intuitive interfaces.

This enables reproducibility, reduces procedure time, lowers costs, and enhances patient experiences. From training to informed consent, digital planning adds value across the care continuum. It also supports more rigorous postoperative evaluation by enabling comparisons between predicted and actual outcomes, thereby closing the feedback loop.

Moreover, this approach supports the personalisation pillar of 4P medicine. Since no two anatomies are identical, procedures should be customised. Digital tools allow individualised planning that considers anatomy, comorbidities, and lifestyle. The result is not only a better clinical outcome but also improved patient satisfaction and trust.

Robotics: The Next Step in the Digital Evolution

As digitalisation deepens, robotics enters the scene. Robotic-assisted interventions, once limited to urology or orthopedics, are emerging in vascular procedures. But robotics alone is insufficient.

Robots require precise instructions, which Digital Twin simulations can provide. Combining AI, Digital Twins, and robotics creates an ecosystem where procedures are guided by real-time, patient-specific digital models.

Robotic systems can also improve dexterity, reduce fatigue, and offer new access pathways that would be challenging with manual instruments. As robotics becomes more integrated with planning software, it will further streamline the surgical workflow, reduce variability, and contribute to better long-term outcomes.

This intelligent fusion of human expertise and computational power promises safer, faster, and more effective interventions. It represents not a replacement of skill but an evolution of the surgical craft.

Barriers to Adoption and the Path Forward

Nevertheless, significant challenges remain. In particular, regulatory frameworks are lagging behind the pace of innovation, and the upcoming AI Act in Europe is unlikely to resolve this gap. Validating AI and Digital Twin technologies requires access to large-scale datasets and long-term clinical trials. Additionally, integration with hospital IT systems is complex and demands secure, standardised communication protocols.

Training and cultural change are essential. Surgeons must view digital tools as allies that enhance their expertise, not as replacements. Institutions should invest in technology, change management, and digital literacy. Bridging the gap between engineers and clinicians is also critical to ensure that these tools are not only powerful but also user-friendly and aligned with real-world clinical workflows. Universities must adapt their curricula to prepare the next generation of surgeons for this new state-of-the-art environment.

Yet, the path is clear. Digitalization in vascular surgery is not a trend—it is a transformation toward a predictive, personalised, and more effective healthcare system. As the cost of computing power decreases and clinical validation accelerates, broader access and adoption are inevitable.

Conclusion

Vascular surgery stands at the threshold of a new era. We can no longer rely solely on human cognition, no matter how skilled. The demand for precision and the complexity of patient care require more.
With AI-powered Digital Twins, we simulate and decide before we act. Predictive indicators like the ERI are entering standard practice. Companion algorithms enhance device use. Robotics is poised to extend surgical capabilities beyond human limits.

Digital intelligence is not replacing the surgeon; it is enhancing every decision they make. This is the future of vascular interventions—a convergence of human and machine, of simulation and judgment, delivering on the promise of truly personalised medicine.

As Professor Salter said, “Decisions are more important than incisions.”

--Issue 06--