A digital twin is not an in sillico friend to entertain you or keep track of your tasks. A Digital Twin is an engineering concept, where individual physical devices are paired with digital models that dynamically reflect the operation and state of those devices.
The idea presupposes that it is possible to understand the operation of a device and its state from data emanating from sensors.
Digital twin models for mechanical devices, such as oil wells, or even complex ones like jet engines, have proven to be very useful, especially for remote monitoring.
That is the conceptual definition of digital twins. The physical definition is a combination of many technologies including a network of sensing devices, typically high speed communications, a gathering place for data and modeling, such as a cloud, hybrid cloud and other exotic architectures. Simulation models, a capacity for large-scale data and computing, especially edge computing, cloud computing and artificial intelligence are all necessary elements. And most of all, a model of how the “thing” works, preferably with pretty high fidelity.
In the example of a jet engine, every engine has its distinct characteristics, even across identical models. The historical pattern of its use, the aircraft to which it is attached, the conditions in which it operates, the quality and cadence of maintenance and any out-of-range events that occurred in its operation. Aircraft engine manufacturers are aware of these issues, and more, and endow the engines with many redundant sensors. The safety record of commercial aviation is a testament to how well the designers and manufacturers understand the variances and minutiae of operation.
There is a difference between a digital twin and simulation models, which have been used for decades, and may use the same type of sensor data, though not always. But simulations generate and manipulate data as part of the simulation. The whole point of a simulation is to project what CAN happen, not what is happening at the moment.
Designing simulations is as much art as science, and is always complicated. However, the characteristic of a simulation model is to start with the logic. In today's world, this is the reverse. Inside of a fully-formed model, we start with a hypothesis and gather as much data as we can find, and the machine looks for patterns. For example, classical propensity models try to understand what humans will do next, even though humans are unpredictable. The models can perform well in a short timeframe, but overall their predictive capability is quite low, but so are the stakes. That would be unacceptable in medicine. The ethical implications for therapy or preventative care are extreme.
This is my summary of the promise of Digital Twins for humans: to go beyond gathering and analyzing data, it has the potential to change medicine as we know it by designing a digital model to align the moving parts of a whole system. That applies to both individuals as well as the healthcare universe. The aim is to generate insights that lead to better decisions on a timely basis and, hopefully, better outcomes.
Digital twins for medicine - still immature
Digital Twins for medicine are an immature technology that propose to dynamically reflect the “-omics” (the addition of “omics” to a molecular term implies a comprehensive, or global, assessment of a set of molecule), such as genomics, biomics, proteomics, or metabolomics, as well as physical markers, demographic, and lifestyle data over time of an individual.
The aim of Digital Twins is to realize Personalized Medicine by identifying deviations from normal. It is questionable how feasible this is at our current level of knowledge. However, there are projects to use digital twin techniques for PARTS of the human body, such as the heart. If researchers choose to look at the heart as just a pump, this will not work. The heart is loaded with neurotransmitters, with sentient epithelial cells and even its own microbiome. In other words, it is much more complex than it appears at first blush. Nevertheless, some useful effects may emerge, but that’s the whole problem with human digital twins – researchers get far over their ski tips.
Some issues with digital twins are problematic:
- Normal: The digital model is used to depict what is normal for one person but compared to patterns in the population. Moral distinctions may be based on patterns found in the data, and the ethical and societal implications of Digital Twins are unclear.
- Inequality: The technology might be available to a limited and patterns identified across a population can lead to the discrimination we observe in poorly developed AI.
- Privacy: The implications of a healthcare organization, insurance company, or any other organization having a persistent, detailed picture of biological, genetic, physical, and lifestyle information of a person over time is a troubling ethical issue.
There are so many issues with twinning human beings that it is not ready for primetime. However, there is a reasonably feasible application of the technology in healthcare: A digital twin of a hospital with operational strategies, capacities, staffing, and care models is more aligned with the engineering model. GE Healthcare proposes uses such as assisting in bed shortages (meh, that's a simple model), spreading of germs, staff schedules, and operating rooms (that seems promising). The effect would be to improve patient care and performance (and also cost, apparently). Digital Twins can be a virtual test of alternatives without actual risks.
It begs the question, even if the data in the Digital Twin is pristine, are the models that determine “abnormal” and generate inferences good enough? The models and interpretation of the model are often wrong, catastrophically wrong. Lower salt or not for high blood pressure? Low-fat or high-protein?
For example, a 2013 study of a decade of medical journal articles found that of the 363 articles focused on standard of care practices, 146, or about 40%, led to reversals of the practice. A 2019 study of over 3,000 randomized controlled trials published in three prominent general medical journals concluded that 396 of these trials constituted medical reversals. The most common disease category among the reversals identified was cardiovascular disease.
Digital Twins also have the potential to impact a person's identity, since meaning can be assigned to the patterns in the data. The engineering paradigm inherent to Digital Twins based health care will raise novel ethical, legal, and social issues for therapy and enhancement. Digital Twins, for instance, can challenge equality…
The differences between persons can be sharply defined and made extremely transparent based on the differences in their compiled information, leading potentially to segmentation and discrimination. Personal Digital Twins are an asymptotically data-intense scenario that clarifies the importance of governance concerning the production and use of personal biological and lifestyle data.
This is the dim view. I ascribe to it. No data is secure from bad actors, who may learn what you bought from Amazon and the model of your iPhone and the path of your locations, but healthcare providers cannot prevent access to your most intimate information.