Digital Twin Generators for Disease Modeling

• Main Authors: Alam, Nameyeh, Basilico, Jake, Bertolini, Daniele, Chetty, Satish Casie, D'Angelo, Heather, Douglas, Ryan, Fisher, Charles K, Fuller, Franklin, Gomes, Melissa, Gupta, Rishabh, Lang, Alex, Loukianov, Anton, Mak-McCully, Rachel, Murray, Cary, Pham, Hanalei, Qiao, Susanna, Ryapolova-Webb, Elena, Smith, Aaron, Theoharatos, Dimitri, Tolwani, Anil, Tramel, Eric W, Vidovszky, Anna, Viduya, Judy, Walsh, Jonathan R
• Format: Journal Article
• Language: English
• Published: 02-05-2024
• Subjects: Computer Science - Learning; Statistics - Machine Learning
• DOI: 10.48550/arxiv.2405.01488

A patient's digital twin is a computational model that describes the
evolution of their health over time. Digital twins have the potential to
revolutionize medicine by enabling individual-level computer simulations of
human health, which can be used to conduct more efficient clinical trials or to
recommend personalized treatment options. Due to the overwhelming complexity of
human biology, machine learning approaches that leverage large datasets of
historical patients' longitudinal health records to generate patients' digital
twins are more tractable than potential mechanistic models. In this manuscript,
we describe a neural network architecture that can learn conditional generative
models of clinical trajectories, which we call Digital Twin Generators (DTGs),
that can create digital twins of individual patients. We show that the same
neural network architecture can be trained to generate accurate digital twins
for patients across 13 different indications simply by changing the training
set and tuning hyperparameters. By introducing a general purpose architecture,
we aim to unlock the ability to scale machine learning approaches to larger
datasets and across more indications so that a digital twin could be created
for any patient in the world.