Page 57 - 23_EETE_03

EE|Times EUROPE 57

Human Digital Twins Are Set to Revolutionize Medicine

Cardiac MRI intensity distributions show how the locations of cardiac fibrosis can vary between patients. Large image shows an
individual patient dataset that can be compared to the population. (Source: Caroline Roney)

“If you really want to simulate a response to different diets or Like Cedersund and many others in the digital twin community,
medications and also simulate immunological processes and the devel- Roney has been working with machine-learning methods and mecha-
opment of diabetes that could change over time, then you need our nistic simulations and is patently aware of the rising complexity of the
multi-organ digital twins, rather than machine learning and traditional ensuing models, which draw on data from increasingly diverse sources.
AI,” said Cedersund. “We’ve also been helping AstraZeneca with Beyond MRI scans, there are CT scans, time-series data including
diabetes-related diseases like inflammation and fatty liver disease.” electrocardiograms, pressure measurements and much more.
AstraZeneca isn’t the only pharmaceutical business pursuing “The idea of Edith is that anyone can contribute a case study …
digital-twin research. Denmark’s Novo Nordisk has been funding the and our vision is to build this big, accessible data repository so you
BioTwin project, which is generating personalized digital twins for can link to, say, different imaging or time-series datasets that could
users to monitor nutrition and weight. And U.K.-based startup Sanome relate to a disease, the age of a patient or a location in the body,”
is working toward a digital twin to detect health changes as early as Roney said.
possible.
As more players join the field, the need to integrate technologies STANDARDS AND ETHICS
and develop an open ecosystem for human digital twins in healthcare Data standardization is set to be an issue as human digital twins come
is becoming clear. The pan-European Ecosystem for Digital Twins to fruition, and Edith researchers are proceeding accordingly. A big
in Healthcare (Edith) project, launched last year, aims to produce a part of Roney’s research has been to understand how to use different
roadmap to virtual human twins. Members of the consortium will also datasets and combine population data with patient data to determine
develop a repository of digital twins and set the stage for a simulation the best therapy. The researchers are also ensuring that the growing
platform. body of data that digital twins can capture is used ethically and legally
Thus far, 19 organizations have joined Edith, including the and that any models under development will be useful to both patient
University of Amsterdam; the Belgium-based Virtual Physiological and clinician.
Human Institute; Italy’s AI institution, Pi School; the French National “We’ve all got different tools and resources but want to share them
Research Institute for Digital Science and Technology (Inria); and and make sure we’re not reinventing the wheel every time we [develop
Queen Mary University of London. a model],” Roney told EE Times Europe. “I’m hoping that one of the
Edith member Caroline Roney, lecturer in Computational Medicine things that will come out of Edith is that we’ll find a common way to do
at Queen Mary University, has been developing personalized models of this while moving the research forward.”
the heart’s atrial chambers and virtual cohorts to design patient- So exactly when does Roney expect an entire human digital twin
specific treatments for atrial fibrillation, a common abnormal heart to be ready to use? Not quite yet. “Right now, I’m really focusing on
rhythm. For example, she has been collaborating with researchers at getting these single organ models working really well,” she said. “We’re
Johns Hopkins University and the University of California San Diego working toward this vision of a digital twin of the whole body. This is
to create an atlas of atrial fibers using diffusion tensor MRI datasets. still a ways away, but we’re getting closer.”
Given her background, Roney is taking part in a cardiac case study as Cedersund, however, believes a human digital twin may be with us
part of Edith. That study will eventually be linked to case studies on soon. He was awarded €6 million in December to head up an EU-funded
additional organs and on related diseases such as cancer and osteopo- project, Stratif-AI, that will develop, evaluate and implement human
rosis, as well as to other types of data. For integrated digital twins, this digital twins to help prevent and treat strokes. As part of the project,
will be crucial. his multi-organ, body-wide models will be tested on hundreds of
“The world is waking up to how revolutionary personalized med- patients from eight hospitals in five European countries.
icine can be if made widely available, and we’re going to see how “We already have an app that presents data on a smartphone,” he
we can make digital twins in a way that can move research forward told EE Times Europe. “It has yet to be certified, but in about two years,
quickly,” Roney said of the Edith project. “We’ll be sharing and patients could start to use the early beta versions of these digital twins.
integrating research methods, datasets and models to really improve … This is a rapidly growing field that is gaining rapidly in popularity.” ■
therapies and learn from so many across our community on how best
to do this.” Rebecca Pool is a contributing writer for EE Times Europe.

www.eetimes.eu | MARCH 2023