In an increasingly digitalized medical industry, digital twins are among the hottest new technologies.
The term digital twin refers to a computational model that is an exact replica to a physical object, person, or mechanism. This technology can be applied to many areas to visualize outcomes and aid in the decision-making process. Digital twins can also implement machine-learning algorithms for future predictions.
Some digital twins can be connected to their real-life counterpart so the virtual simulation can receive live updates such as the altitude of a plane, or the heart rate of a person. Through this, many different simulations can be run in real-time.
In the past, digital twins have primarily been used in engineering or business sectors for machine testing and market analysis. With medicine becoming increasingly digitized, large amounts of data are available to be processed, and digital twins are one of the latest applications.
Currently, digital twins use patients’ personal data to simulate virtual models of organs and disease states. The intention is that digital twins will transform in silico—computer-based—trials where instead of testing possible treatments in a patient directly, a simulation can be run.
The use of machine-learning in pharmaceutical trials isn’t a foreign concept and has become more prevalent in assessing drug efficacy. ELEM BioTech, a European start-up company, has refined their simulated human heart models using machine-learning and offer testing to other companies.
Applying machine-learning algorithms to clinical trials enables scientists to predict drug-drug interactions, dosing, and treatment combinations for a patient. Rather than running repetitive trials, predictive algorithms can provide more accurate insight while saving time and money. This streamlined approach aims to provide patients with solutions faster than the traditional clinical trial route.
The virtualizing of medicine has expanded, with groups of scientists coming together to create a digital twin of the human immune system. This would work to model disease states, cancers, and infections, although scientists are still aiming to scale it up with more immune cells.
Similarly, to the Human Genome Project, a “roadmap” of their work has already been published in Nature Digital Medicine, supporting open science, and promoting collaborative research.
Although the public may fear digitization of medicine, it can be argued that current medical standards are less than scientific.
Most diagnoses and treatments are based on past patients with similar conditions, analyzing their disease course. Using a digital twin allows healthcare providers to select personalized treatments that are most likely to work for a specific patient.
The benefits are already being seen in cancer patients who can have digital replicas of their tumors—including genetic and molecular specifics—tested for drug efficacy. GlaxoSmithKline (GSK), a company focused on global healthcare has collaborated with cancer researchers at King’s College London to collect dynamic biomarkers, images, and samples from patients to form these models.
The demand for digital twin technology will grow and is expected to create a multi-billion-dollar market by 2026. The intersection between AI power and real-time data collection will have a profound influence on the development of medicine.
Tags
AI, innovation, machine-learning, medicine, Research
All final editorial decisions are made by the Editor(s)-in-Chief and/or the Managing Editor. Authors should not be contacted, targeted, or harassed under any circumstances. If you have any grievances with this article, please direct your comments to journal_editors@ams.queensu.ca.
