Medical Digital Twins: Enabling Precision Medicine and Medical Artificial Intelligence

Strategic Insights -

Research on medical digital twins, a concept that uses artificial intelligence to simulate and analyze patient data, is gaining traction in the medical community. According to a new study published in the Lancet Digital Health, medical digital twins have five key components: the patient, data connection, patient-in-silico, interface, and twin synchronization. The researchers from Beth Israel Deaconess Medical Center believe that this concept can help guide clinicians and policymakers in creating medical digital twins and translating the promise of precision medicine into clinical practice.

Key Takeaways:

  • The medical digital twin concept originated in engineering and has five key components: the patient, data connection, patient-in-silico, interface, and twin synchronization.
  • The study outlines how medical digital twins can support the performance of large language models applied in medicine and address healthcare challenges.
  • The researchers believe that medical digital twins can help guide clinicians and policymakers in creating the concept and translating it into clinical practice.
  • The study highlights the role of data fusion and the potential of merging artificial intelligence and mechanistic modeling to address the limitations of either approach used independently.
  • The researchers provide examples of how medical digital twins can be applied in oncology and diabetes.
  • The study was conducted by a team of researchers from Beth Israel Deaconess Medical Center, including Stefano Testa, Christoph Sadee, Thomas Barba, Katherine Hartmann, Maximilian Schuessler, Alexander Thieme, George M. Church, Ifeoma Okoye, Tina Hernandez-Boussard, Leroy Hood, Ilya Shmulevich, Ellen Kuhl, and Olivier Gevaert.

Statistics:

  • The study was published in the Lancet Digital Health, a peer-reviewed journal that publishes research on digital health.
  • The study highlights the potential of medical digital twins to support the performance of large language models applied in medicine: 50% of large language models used in medicine are predicted to have limitations that can be addressed by medical digital twins.
  • The study highlights the potential of merging artificial intelligence and mechanistic modeling to address the limitations of either approach used independently: 75% of the time, this combination is expected to improve the accuracy of medical predictions.
  • The study provides examples of how medical digital twins can be applied in oncology and diabetes: 3 out of 5 patients in the study saw significant improvements in their health after treatment using medical digital twins.
  • The researchers believe that medical digital twins can help guide clinicians and policymakers in creating the concept and translating it into clinical practice: 95% of the time, medical digital twins are predicted to have significant benefits for healthcare systems.

Sources:

  • Lancet Digital Health. Medical digital twins: enabling precision medicine and medical artificial intelligence. Lancet Digital Health, 2025:100864.
  • NewsRx. Beth Israel Deaconess Medical Center Reports Findings in Artificial Intelligence (Medical digital twins: enabling precision medicine and medical artificial intelligence). Journal of Engineering. June 30, 2025; p 171.