Large language models forecast patient health trajectories enabling digital twins - Takeaways - MDSpire

Large language models forecast patient health trajectories enabling digital twins

  • By

  • Nikita Makarov

  • Maria Bordukova

  • Papichaya Quengdaeng

  • Daniel Garger

  • Raul Rodriguez-Esteban

  • Fabian Schmich

  • Michael P. Menden

  • October 1, 2025

  • 0 min

Share

  • 1

    DT-GPT utilizes electronic health records to predict patient health trajectories without requiring data imputation or normalization.

  • 2

    The model outperformed state-of-the-art machine learning models, reducing mean absolute error in various clinical datasets.

  • 3

    DT-GPT maintains clinical variable distributions and demonstrates explainability through a human-interpretable interface.

  • 4

    The model's zero-shot forecasting capability highlights the potential of LLMs in clinical forecasting applications.

  • 5

    DT-GPT aims to enhance digital twin applications in clinical trials, treatment selection, and adverse event mitigation.

Original Source(s)

Large language models forecast patient health trajectories enabling digital twins

  • by Nikita Makarov, Maria Bordukova, Papichaya Quengdaeng, Daniel Garger, Raul Rodriguez-Esteban, Fabian Schmich, Michael P. Menden

  • October 1, 2025

Related Content

Navigating Healthcare Challenges for Family Caregivers of Individuals with Alzheimer’s Disease and Related Dementias

The Importance of Molecular Profiling in Delivering Guideline-Recommended Care—From Genetic Sequencing to Patient Outcomes

Worldwide Impact of Respiratory Cancers in 2022 and Projections for 2050: Incidence and Mortality Data from GLOBOCAN