NASA was the first entity to use digital twin technology to simulate and predict physical systems. When an oxygen tank exploded on the Apollo 13 spacecraft, NASA used physical replicas of the spacecraft to simulate the damage.
Today, digital twins are more sophisticated, receiving data from multiple sources to accurately mirror real-world operations, analyze behaviors, and deliver predictive insights in industries such as manufacturing, construction, communications, and energy. Digital twin technology in healthcare has lagged behind, but it is gaining traction. From surgical planning to precision medicine, the technology is being tested in medical schools, clinical trials, and physician training. However, more discussion is needed to understand how the technology will fit into routine medical scenarios.
How do medical digital twins work?
Digital twin technology in healthcare creates a virtual replica of a patient, organ, system, or process from multiple data sources such as images, electronic health records, and wearables. The virtual twin uses the data to simulate health scenarios, predict outcomes, and personalize treatments.
Artificial intelligence (AI) uses the data to build a model of a system, organ, or patient. Once the foundational replica is in place, the AI application can continuously update the twin based on patient data. Some applications may even include lifestyle or environmental influences that impact a patient’s health.
Physicians can use the twin to assess treatment options, evaluate medication or dosage, and map out complex procedures before interacting with the patient. These simulations can help clinicians decide on a course of action and treatment plan that minimizes medical guesswork.
Most digital twins receive data from their physical counterparts, but some twins can send and receive data. A bidirectional data flow allows the AI-based application to return insights, such as suggesting a safer surgical pathway or alternative drug dosage.
Are doctors on board with digital twin technology?
Based on a Sermo poll, 54% of physician respondents had no experience at all with digital twin technology. About 18% were familiar with the term but did not understand how it worked, and another 20% had only read about it. For some, such as one general practitioner, digital twins seem “sound promising, especially for complex cases”; however, routine use appears years away.
Regardless of familiarity, physicians could foresee applications for digital twins in simulating complex cases, predicting future outcomes, and personalizing care. An emergency medicine physician described the technology as digital avatars that were “designed to analyze, simulate, and predict individual health outcomes, moving healthcare toward proactive, personalized, and predictive treatment.”
Although healthcare professionals perceive the technology positively, many are still unsure of its real-world applications.
Clinical applications of digital twins in healthcare
When asked where digital twins would be most useful, 37% of respondents were unsure. The remaining physicians considered the following applications as possible uses for digital twins, provided the appropriate safeguards are in place.
Predicting treatment responses for complex conditions
In a Stanford University School of Medicine–led clinical study involving 50 men with intermediate-risk prostate cancer, researchers evaluated Avenda Health’s AI-based modeling platform, Unfold AI, which functions similarly to a digital twin by integrating biomarkers, biopsy results, imaging, and clinical data to generate a patient-specific 3D cancer map. The study showed that margins defined using this AI model completely encapsulated clinically significant cancer in 80 % of cases, compared with 56 % using conventional margins, suggesting that personalized mapping can reduce the risk of missing tumor tissue during treatment planning.
Personalizing medication dosing and selection
Precision medicine aims to tailor diagnostic, prognostic, and therapeutic decisions to the individual patient rather than the population average. Digital twin technology supports this approach by integrating biological markers, genetic data, lifestyle factors, and environmental exposures into a dynamic computational model of the patient. These models draw on heterogeneous data sources—including electronic health records, longitudinal clinical data, and real-world behavioral inputs—to generate statistically robust simulations. In clinical practice, digital twins can be used to predict disease progression, simulate medication dosing and treatment responses, and compare therapeutic options before intervention. This enables more informed clinical decision-making and supports optimized, individualized care pathways, particularly for patients with complex or chronic conditions.
Surgical planning or procedural simulation
Providing surgeons with graphical detail for surgical procedures requires extensive computational power and efficient graphics processing. Nvidia, the leading graphics chip manufacturer, participated in a digital twin project for surgical preparation. The process creates photorealistic anatomical models from CT data using organ segmentation, mesh refinement, and texturing. These digital human twins could help surgeons to be better prepared for procedures, reduce surgery times (and therefore, also surgeon fatigue), identify safe surgical pathways, and improve patient safety.
Modeling disease progression for chronic conditions
Duke’s Center for Computational and Digital Health Innovation has created digital twins that simulate the impact of medical interventions on chronic conditions such as cardiovascular disease. The application uses the virtual patient replicas to compress time, demonstrating how treatments can affect future outcomes. Physicians can see how a disease progresses without treatment and compare the outcomes of different treatment options. It allows physicians to personalize treatment for patients suffering from chronic diseases.
Medical education and training
As more technology enters healthcare, digital twins become excellent tools for medical professionals to refine their skills. With digital twins, surgeons can practice before implementing new technology. Twins enable medical students to experience the impact of medication and dosage on patient outcomes, building clinicians’ proficiency in precision medicine. While digital twins have the potential to improve surgical and clinical outcomes, one Sermo member noted that simulations are still not the same as facing a patient and all of the complications that they bring with them, a sentiment others shared.
Physician concerns with digital twins in healthcare
Physicians share similar concerns to others evaluating the use of digital twins in their field of expertise. They wonder about data accuracy and privacy, legal implications, and model validation. Unlike other industries, healthcare’s concerns carry added weight as they involve life-changing decisions. The primary concerns, as shown in a Sermo poll, were the accuracy and reliability of simulations (30%) and the need for real-world validation (24%).
Reliability and accuracy of simulations
Because accurate simulations depend on accurate data, selecting data sources is a crucial factor in digital human twin creation. Each instance should stipulate the data sources for full transparency, and simulations must be free of AI biases to ensure outcomes are representative of the patient’s genetic, social, and environmental influences. Without these assurances, physicians may question the technology’s usefulness.
Data privacy and patient consent
Who owns the digital twin? If patients consent to using their data to construct digital twins, do they lose control of their information? Could their data be used numerous times by multiple physicians without their knowledge? Is it possible for digital human twins to become the Henrietta Lacks of the digital world? As a general practitioner noted, “There’s still so much we don’t fully understand about the limits and ethical side of this technology.”
Integration into workflows and EHRs
Digital twins extract data from multiple sources, but they rarely transmit data outside their application. Establishing interfaces to EHR systems to access comprehensive patient information will require significant development time. Without integration, physicians may need to enter treatment data in multiple places, adding to their workload.
Legal or liability issues
Beyond the question of ownership, healthcare must address liability concerns. What happens if a digital twin error results in a poor outcome? What responsibility does the patient have in providing accurate and complete information? Could twins be used to illustrate why a course of action is appropriate for insurance companies? What access, if any, should insurers have to a patient’s digital twin?
Digital twin technology requires a large platform with massive data storage and computing power. Securing that data from unauthorized access becomes a shared responsibility as physicians create twins from personal patient data and platforms access multiple sources for comparative data.
Lack of real-world validation
How will doctors know the technology works? What standards and testing protocols should be in place to validate a digital twin application? These are questions still largely unanswered. As an enthusiastic pulmonologist on Sermo noted, “digital twins could really change the way we approach patient care, but I’m curious how accurate and reliable these models will be in real-world clinical settings.”
With limited current clinical applications, real-world validation remains an open question.
What does the future hold for digital twin technology?
Almost half (45%) of the respondents in a Sermo poll wanted to see the technology in practice before committing to its use. Others wanted to see supporting data, while 23% would use the tools with strong human oversight. Another 21% were skeptical of its clinical value.
The viability of digital twin technology varies by specialty. For example, IBM Watson encountered numerous challenges in delivering a digital twin to revolutionize cancer treatment. IBM’s Watson for Oncology was to bridge the gap between cutting-edge research and clinical practice by identifying personalized treatment options. However, the project was unable to accommodate the complexities of cancer care, resulting in failure.
Another recent study used digital twins to predict patient outcomes for 1,800 patients with type 2 diabetes. Each patient had a digital twin that mirrored their dietary intake, blood glucose levels, metabolic status, and lifestyle habits. The platform suggested food choices to each patient based on current readings. At one year, patients showed significant improvement in hemoglobin A1c levels and required less anti-diabetic medication. Many experienced better weight reduction and less insulin resistance.
These two studies illustrate the technology’s value in improving patient care while highlighting the challenges that can limit its effectiveness. At present, the technology is being used in clinical trials and medical research to deliver twins that address the lack of real-world validation and data accuracy. As the volume of valid test data grows, physicians will be more likely to trust the technology in clinical settings.
How could digital twin technology change physicians’ daily lives?
If proven effective in clinical practice, digital human twins can enable physicians to shift care from reactive to predictive, simulate complex procedures, and manage chronic conditions. Ultimately, they will reduce guesswork, improve outcomes, and encourage lifelong learning.
Physicians will have greater confidence in the treatment plans, knowing the possible outcomes. They will provide more targeted medication to reduce trial-and-error-related office visits. They can experience innovative treatments for continual professional development. Digital twins have the potential to improve patient outcomes and help maintain physician wellness with time-saving technology.
A path forward
Digital twin technology holds significant promise for transforming healthcare, but its widespread clinical adoption will depend on addressing ethical and governance challenges first. Robust validation, bias mitigation, and strong protections for patient privacy and data security must be established before digital twins can be safely integrated into routine care. Meaningful dialogue around data quality, consent, transparency, and accountability is essential to ensure these tools support—rather than undermine—clinical decision-making.
Physicians play a critical role in this process. Your clinical expertise is needed to guide developers in defining appropriate use cases, identifying bias, and establishing rigorous validation standards. Without early and sustained physician involvement, digital twin applications risk falling short of their potential or eroding trust among patients and clinicians. Sermo offers physicians a confidential platform to discuss technology in healthcare. As a community, Sermo members can explore the key issues surrounding the technology and evaluate real-world implementations among global peers. It’s through these meaningful discussions that a path forward can be forged. Be part of the future by joining the Sermo community.
