Destination Wellness
Opportunity for more ethical, effective, & pragmatic healthcare.
Imagine this Future
After the silent departure of the autonomous rideshare, Claire grasped her son’s hand reassuringly, then led him into the medical facility. Steve had grown to the age for his first bioscan. Easily finding the distinctive wing designed for children, the process began with preparing the boy for a safe and distracting-from-fear adventure.
“Mom, aren’t you a twin?” Steve asked. “Yes, but this is a different kind of twin. It’s not alive, like your uncle, it’s made of information and kept in computers, but it does grow alongside you and experience the same things in life as you,” she explained, “like when you get sick or if you break a bone.”
“But can’t they use my avatar from a video game?” he questions as he puts on the the biometrics suit inside the stage-1 kiosk. “No, it is important that this twin will be as exactly like you as will help us and our doctors keep you healthy,” Claire responds, adding, “Remember when grampa fell and broke his hip? Well, they made an exact copy of that bone from measurements in his twin.”
Steve is getting a baseline scan, capturing a multitude of “healthy normal” physiology measurements that will be updated again when he is fully grown. This baseline model is like the foundation and frame of a building, giving it structure. Parts will be updated periodically or following certain events. These events could be directly from Steve, such as wisdom teeth removal, or closely-tied to him, like proximity-measured environmental air quality and use of an asthma inhaler.
All medical data over the course of Steve’s life will be tied to this Human Digital Twin (HDT) [1]; most of it owned, controlled, (and possibly even monetized) by Steve. A modified HIPAA and elegantly implemented system for data governance enables compliant sharing and joint ownership, such as to integrate relevant family histories or to anonymize for inclusion into virtual clinical trials [2] and public health monitoring [3].
Wellness as Human Maintenance
The HHS website [4] frames the “wellness” concept in preventative measures such as nutrition & fitness, immunization, and medical screening — as well as considering mental health, environmental factors and lifestyle (tobacco, alcohol, safety). In envisioning this bright future, however, I would like to convey the wellness concept with a more inspirational impact.
NASA engineers originally developed the digital twin concept to improve performance and reliability of machinery (spacecraft) — which may seem a lower standard for care than we would wish human beings, but the reality is machinecare capabilities far surpass biomedicine [5]. (The physics of the mechanical models needed are better understood than biomechanical models, and compared to patients, the machines are more compliant, do not need privacy, and are easily monitored through connected sensors.)
In fact, machinecare has even adopted terminology like product lifecycle and prognostics and health management (PHM) [6]. Through condition-based maintenance, engineers can already achieve most of the outcomes in the vision for personalized/precision medicine. With machines, engineers seek to improve reliability (to minimize “down time”) while controlling costs of maintenance, repair and overhaul (MRO) services.
In healthcare, these concepts correspond to promoting wellness: by maintaining pragmatic quality of life — through prevention and avoidance of debilitating illness, medical conditions and injuries while minimizing interventions from drug therapies to medical procedures such as surgery (which impose costs — both financially and to quality of life).
Practicing wellness medicine would carefully monitor the slippery slopes of the diagram below, proactively steering the patient through life’s thicket of health risk factors to avoid descending into the low health red zone.
Data needed to inform such guidance may extend beyond our accustomed physiological and medical imaging examinations at a physician’s office, to include cellular, genomic, and microbiome (gut bacteria) assessments. Risk factors to consider would include physical activity and nutrition, as well as environmental exposure to pathogens, pollutants, elevated heat, and water and air quality (see: [8] High Definition Medicine, Tokamani, A., et. al., 2017).
Data quality is essential for this form of health management, so the twin model continues to accurately represent its corresponding patient and provides informative predictions — or clear and early alerts when measured health data diverge from what was expected based upon medical guidance.
One ambitious goal is the creation of biomechanical twins, which could correspond to a single patient, or a population — or (within the body of a patient) could range from tissue cells up to an entire organ or entire biological systems. For example, there have been several impressive twin examples advancing realism in cardiovascular simulations from Dassault, Siemens, Philips, government supercomputing facilities and universities.
But digital twins can benefit healthcare delivery in other ways as well: such as consolidating and integrating medical information, increasing the capacity of care delivery facilities, improving the resilience and agility of the supply chain, and reducing down time of essential medical devices. The FDA has engaged with industry to develop rigorous methods for model-informed product development and certification, even noting areas where “digital evidence” provides data superior to traditional clinical trials and animal models for regulatory evaluations of safety and efficacy.
While getting to this this state of practice for Wellness certainly constitutes a technical grand challenge [9], the benefits go beyond improving quality of life for individuals — the costs of conventional curative care are substantial and rapidly increasing, imposing financial burden on individuals and their families as well as society. 2022 U.S. government healthcare spending neared $2,000,000,000,000 (which is $6000 per citizen and 8% of GDP).
“Prolonging lifespan without prolonging health span is financially unsustainable for all nations.” — G. Tourassi
Back to the Future
Years later, Claire and family members virtually gather to consider her aging father’s health data, guided by a trusted health advocate. Intelligent machine agents help analyze and summarize [10] information to better understand the many factors influencing a difficult decision. Dad has accumulated the expected age-related maladies, which are put into an intuitive visualization that can be played back in time and compared against a population of similar patients.
But, importantly, the data can also be played forward in time — where his twin undergoes thousands of simulated therapy what-ifs on his behalf, to analyze sensitivities and help predict the most likely outcomes. These scenarios do not simply seek the longest lifespan as optimal, they soberly bring into focus the realistic paths to end of life.
Output from the models reduce uncertainty and reassure among emotional aspects of the options. One path may win a few years following a grueling and expensive treatment, another path is much shorter, but offers dignity and palliative options. A third option shows great promise for many years, but along the longer path gained from its success, dementia looms likely.
Steve looks up as Claire finishes the call, asking, “Is he still being stubborn about not getting the treatment?” She nods, pensive, but sniffs and smiles, “But I understand now. And he’ll be able to come visit us a few more times instead of being alone in a hospital and feeling nauseous all day. A thousand of his twins grew old and died a thousand different ways for him, and for us.”
“Just like yours helped you figure out how to stop getting those headaches,” the boy observes. His table-screen darkens, shutting down. Claire remembers a game played long ago with her father on such a screen, “The twin let us see a little into the uncertain darkness of the future and speak honestly — rather than our hearts blindly guiding us to roll the dice on hopes offered by the hospital.”
Steve tips his head, “And I guess his twin can still keep teaching us things that will help the rest of our family things make good choices.” Following a pause, he takes on the posture and voice of his grandfather, “I worked my whole life so we could all get out in the world — explore, feel wonder and make it little better.” Claire smiles slightly at her son, who continues, then as himself, “I know we still have a while, but he did OK, didn’t he? Lived a pretty long time.”
“What anybody gets —he got a lifetime,” she recalls, quoting, “no more, no less.”
© 2023 All Rights Reserved.
As I composed a submission into DARPA’s Polyplexus brainstorming “Dash to Accelerate Health Outcomes” exercise for ARPA-H, I could neither fully fit my own thoughts into the textboxes nor rely upon other citations to paint the picture I wished to convey. In this first part — a vision of what may be possible and why that is better, then a companion article will consider today’s obstacles and opportunities.
Next:
See also presentation given at BioITWorld 2023:
Biomedical Digital Twins, Perspective from Industry.
Sampled References
- The perioperative human digital twin, Lonsdale, H. et. al, Anesthesia & Analgesia 134(4): 885–892. (2022). DOI: 10.1213/ANE.0000000000005916
- Virtual Clinical Trials: Challenges and Opportunities, National Academies Press. (2019).
- Digital public health surveillance: a systematic scoping review., Shakeri Hossein Abad, Z., et al., npj Digit. Med. 4, 41. (2021). DOI: 10.1038/s41746–021–00407–6
- US Health & Human Services program: Prevention & Wellness
- Industrial Digital Twins: Leveraging Machine Healthcare, Irving, R., Multiscale Modeling in Physiology, NIH Interagency Modeling and Analysis Group. (2019). See also: webinar YouTube video/15 mins.
- NIST Report: ASME Standards Subcommittee on Advanced Monitoring, Diagnostics, and Prognostics for Manufacturing Operations, Weiss, B., et. al., NIST Advanced Manufacturing Series 100–31. (2020). DOI: 10.6028/NIST.AMS.100–31
- Dynamical systems approaches to personalized medicine, Davis, J., et. al., Current Opinion in Biotechnology, Vol 58, 168–174 (2019). DOI: 10.1016/j.copbio.2019.03.005
- High-Definition Medicine, Torkamani, A., et. al., Cell, vol. 170, iss. 5, 828–843. (2017). DOI 10.1016/j.cell.2017.08.007
- Computational medicine: Grand challenges and opportunities for revolutionizing personalized healthcare, Tourassi, G., Frontiers in Medical Engineering, vol 1. (2023). DOI: 10.3389/fmede.2023.1112763
- Digital twins for predictive oncology will be a paradigm shift for precision cancer care, Hernandez-Boussard, T., et. al., Nature Medicine, vol. 27, 2065–2069. (Dex 2021). DOI: 10.1038/s41591–021–01558–5
