{
  "abstract": "The rapidly advancing frontiers of high-throughput data collection and computational simulation are enabling the development of digital twins (DT) to support clinical decision-making in oncology.1 The exact definition of a DT depends on the context of its application, but in general, encompasses a virtual simulation that is customised using real-world data, as well as the methodology to provide specific, actionable insight into how a patient will respond to alternative therapeutic interventions. However, their use for cancer treatment faces technical and ethical challenges that do not apply to other industries that have successfully applied a DT to optimise complex systems (aerospace, automotive, manufacturing). Current DT use cases in oncology include using models to inform clinical protocols, conduct virtual clinical trials, conduct training, and predict responses based on biomarkers.2 However, the application to clinical decision support is typically limited to a single iteration of a DT cycle rather than dynamically informing longitudinal decision making in response to tumour evolution. The data needed to model tumour growth and predict an optimal intervention are noisy and/or incomplete, and their collection must minimise the negative impact on the patient’s quality of life. The computational simulations must be trustworthy and quantify the uncertainty of the predicted response in ways that inspire confidence from the clinical decision makers. These, along with numerous other constraints, must be overcome before the DT can be used to optimise and fully personalise the treatment of cancer.",
  "authors": [
    {
      "affiliations": [
        "Innovation Center for Biomedical Informatics, Georgetown University Medical Center, Washington, District of Columbia, USA",
        "Lombardi Comprehensive Cancer Center, Georgetown University Medical Center, Washington, District of Columbia, USA"
      ],
      "name": "Matthew McCoy"
    }
  ],
  "title": "Digital twins in oncology: where we are and where we hope to go",
  "uid": "d104bc12-c068-54b1-b42a-23c8df1704c1"
}