AI ‘Digital Twins’ Identify Drug Combination to Prevent Early Labour
Key Takeaways:
- A Stanford Medicine-led team has built SIMCO, an AI system that creates ‘digital twins’ of the immune systems of women who are pregnant to test drugs virtually.
- Aspirin plus lansoprazole was predicted to delay labour 15 times more than aspirin alone, and a human clinical trial has been approved.
- Folic acid may also help slow the immune changes that lead towards labour.
A promising new approach to preterm birth
Researchers have identified two common medicines that, taken together during pregnancy, may help prevent preterm birth.
The finding comes from a new Stanford Medicine-led, artificial intelligence (AI)-powered technique that uses a cell-by-cell atlas of the immune system to create ‘digital twins’ of the immune systems of individual women who are pregnant. These virtual replicas allow drugs to be tested without exposing real pregnancies to risk.
The study, published online on 21 October in Science Advances, builds on two decades of Stanford Medicine research into how the maternal immune system shifts during pregnancy. When this immune clock runs too quickly, preterm labour can result. Being born three or more weeks before the due date carries many risks for the baby, and preterm birth is now the leading cause of death in children under five worldwide.
“We’re on the brink of having a real impact for the treatment of preterm birth, a global health problem that affects 10% of pregnancies worldwide,” said senior study author Brice Gaudilliere, MD, PhD, professor of anaesthesiology, perioperative and pain medicine. He noted that it has been decades since a new pharmacological intervention for preventing preterm birth emerged.
The most promising result was the combination of aspirin and lansoprazole. Low-dose aspirin alone has a modest ability to prevent preterm birth, while lansoprazole is sometimes prescribed in pregnancy for acid reflux.
“In the digital twin system, these two drugs in combination had a large effect on the delay of labor, 15 times higher than aspirin alone,” Gaudilliere said.
How the digital twins were built
Stanford Medicine research has shown that immune signals play a key role in preparing the body for labour. Early in pregnancy, the immune system adapts to tolerate the foetus; later, a shift towards pro-inflammatory signals helps prepare for labour. If these signals switch on too soon, early labour can follow.
The new system, called SIMCO (simulated immune modelling of clinical outcomes), uses AI and machine learning to model how the body’s immune cells behave during pregnancy. The researchers collected blood samples from women who were pregnant and exposed the immune cells to many different drugs.
“We measured responses from a quarter of a billion cells, training the algorithm on a massive single-cell dataset,” Gaudilliere said.
“A digital twin has the prerequisite that you understand a biological system very well,” said co-lead author Jakob Einhaus, MD, a postdoctoral scholar at Stanford Medicine. “The digital twin represents the complexity and connections in the immune system.”
Einhaus shares lead authorship with graduate students Peter Neidlinger and Olivier Fondeur, and with Masaki Sato, MD, PhD, now at the National Center for Child Health and Development in Tokyo.
The team then added longitudinal data from 63 pregnancies, linked to clinical outcomes, so the model could connect patterns of cell behaviour to clinical states and account for differences between patients. With a small blood sample from a patient, the model can predict whether a given medication is likely to make her immune clock run faster or slower.
“You’re making the prediction for one individual based on knowledge of millions of single cells from dozens of patients,” Einhaus said.
For healthcare professionals keen to understand how AI tools like this are developing, The College of Contemporary Health’s AI Essentials for Primary Care course offers a practical introduction to AI in healthcare.
Common medications may delay delivery
After confirming that SIMCO accurately forecast time to labour, the researchers used it to evaluate nine drug regimens already known to be safe in pregnancy, chosen through a drug-repurposing pipeline developed by a group led by co-author Marina Sirota, PhD, professor of paediatrics at the University of California, San Francisco. Each digital twin was tested in treated and untreated scenarios.
The team was surprised by how strongly the model predicted that aspirin and lansoprazole would delay labour.
“Our findings suggest that salicylic acid [aspirin] and lansoprazole act synergistically … this synergy may enhance their immunomodulatory effect,” the investigators write in their paper.
The study also found that folic acid, already recommended to reduce the risk of spinal cord defects in the foetus, can delay immunological progression towards labour.
“We’re giving folic acid for spina bifida prevention, but maybe, unknowingly, we’re also helping the maternal immune system adapt to pregnancy,” Einhaus said. “We didn’t know it had such profound immunomodulatory properties.”
With aspirin and lansoprazole, immune processes that help sustain pregnancy were enhanced, while those that trigger labour were inhibited, Gaudilliere said. “So it all makes sense biologically. That was super exciting.”
From discovery to clinical trial
Because the model captures differences between patients, it can also predict which people are likely to respond to a drug. Limiting trial enrolment to likely responders could allow smaller trials to give clear answers about effectiveness.
The researchers showed that aspirin plus lansoprazole delayed labour in an animal model of inflammation in pregnancy, and then received US Food and Drug Administration approval for a human clinical trial.
“We’re going in a short time from discovery work – done the Stanford way, using technology, AI and really understanding the mechanism of a disease process – all the way to identifying a potentially impactful new treatment and testing in patients,” Gaudilliere said.
Researchers from UC San Diego and the AP-HP Sorbonne University Hospital Group in Paris also contributed. The work was funded by organisations including the Stanford Maternal and Child Health Research Institute, the March of Dimes Prematurity Research Centers at Stanford and UCSF, and the National Institute of Health.
CCH insight
From digital twins to drug-repurposing pipelines, AI is reshaping how new treatments are discovered and tested. Our AI Essentials for Primary Care CPD course helps healthcare professionals build confidence in understanding AI and its growing place in clinical practice.
Explore AI Essentials for Primary Care →
Source: Stanford Medicine
Read More
AI Tool Creates ‘Digital Twins’ of Patients to Forecast Future Health
Key Takeaways:
- New DT-GPT model creates virtual patient replicas to predict individual health trajectories with notable accuracy.
- The model outperformed 14 leading machine learning systems and demonstrated effective zero-shot predictions.
- Technology could accelerate drug development and shift healthcare towards more predictive and personalised practice.
Digital twins to forecast patient health
A new artificial intelligence model capable of generating virtual patient representations and forecasting future health outcomes has been described as a potential breakthrough for clinical research. The system, developed by researchers at the University of Melbourne, uses large language model (LLM) techniques to create personalised digital twins that mirror each individual’s clinical profile.
How the DT-GPT model was developed
The research team trained an existing large language model on three extensive datasets containing thousands of electronic health records. These datasets included information on people living with Alzheimer’s disease, people with non-small cell lung cancer, and people admitted to intensive care units. The aim was to equip the model with sufficient breadth of clinical data to enable it to generate detailed patient-level predictions.
The resulting tool, named DT-GPT, analysed each person’s medical history, such as laboratory values, diagnoses, and treatments. Using this information, it constructed a virtual counterpart for every individual and projected how their condition might evolve under ongoing clinical care.
Predictive performance and validation
Crucially, the model was not shown any actual health outcomes during training. This allowed researchers to rigorously assess the accuracy of its predictions once the model generated forecasts.
Associate Professor Michael Menden, lead researcher, explained the approach:
“For each patient, we created a virtual replica by initialising the model with their individual clinical profile.”
He added:
“For example, we created virtual twins of 35,131 intensive care unit (ICU) patients and accurately predicted what would happen to their magnesium levels, oxygen saturation and their respiratory rate over a 24 hour period, based on their laboratory results from the previous day.”
When benchmarked against 14 state-of-the-art machine learning models, DT-GPT consistently outperformed them in predictive accuracy.
Implications for clinical trials and personalised medicine
Researchers believe the tool has significant implications for the future of clinical trials. Because the model can simulate potential outcomes for large groups of virtual participants, it may help streamline drug development processes by reducing time and cost associated with early-stage testing.
Associate Professor Menden said:
“This technology paves the way for a shift from reactive to predictive and personalised medicine.”
He continued:
“It could enable doctors to anticipate if their patient’s health will deteriorate so they can intervene earlier.
“It could also be used to predict negative side effects of medications, allowing doctors to tailor treatment plans to suit each patient’s unique characteristics and medical history, ultimately increasing the chances of a positive health outcome.”
Conversational interface and handling of complex data
One of DT-GPT’s strengths is its ability to interpret large volumes of complex, unstructured clinical data. The system also includes a conversational interface that functions similarly to a chatbot, enabling clinicians and researchers to query the model directly and explore the reasoning behind specific predictions.
Zero-shot predictions: an advanced capability
Because DT-GPT is based on generative AI, it can also perform zero-shot predictions. These are informed estimates of clinical values that the model has not been explicitly trained to predict.
Associate Professor Menden illustrated this:
“To use an analogy, it’s like asking the model to predict how tall someone will grow without providing the person’s height records and only giving their previous weight and shoe sizes.”
He noted a key finding:
“Our model accurately predicted how lactate dehydrogenase (LDH) levels changed in non-small cell lung cancer patients 13 weeks after they started therapy, despite not training the model for this purpose.
“We compared it to traditional machine learning models, which were specifically trained for 69 clinical variables, including LDH, which we in comparison only educated guessed.
“Very surprisingly, the DT-GPT’s zero-shot predictions, its untrained guesses, were more accurate in 18 percent of cases.”
The study was recently published in NPJ Digital Medicine.
Next steps: expanding to other conditions
The team responsible for developing DT-GPT, in collaboration with the Royal Melbourne Women’s Hospital, have now established the foundation for a new company that will apply digital twin technology to support people living with endometriosis. This work highlights the potential wider applicability of the model across different medical conditions.
CCH insight
Digital twins are one of the more striking illustrations of where healthcare AI is heading: from reactive treatment towards prediction and personalisation. Whether or not tools like DT-GPT reach routine practice, the direction of travel makes AI literacy increasingly important for healthcare professionals – understanding what these systems can and cannot do, and how to interpret their outputs. CCH’s CPD-accredited digital health short courses are designed to help practitioners build that understanding and practical judgement, quickly and flexibly around a clinical schedule.
Explore CCH’s Digital Health courses →
Read More