Human Digital Twin: Sociotechnical Plan Highlights
Introduction
The digital twin (DT) concept can be traced back to the National Aeronautics and Space Administration (NASA) Apollo program in the 1970s, where two identical vehicles built for space efforts, and one was able to mirror the conditions of the other during a mission (Wei, 2021). This “mirroring” concept was called “Mirror Space Model,” and later “Information Mirror Model,” until NASA’s John Vickers introduced the new “digital twin” name in his book. Grievers, a professor at the University of Michigan who put forward the concept of “Virtual Digital Expression Equivalent to Physical Product” in a 2003 product lifecycle management (PLM) course, began using Vickers’ DT name in 2011 (Wei, 2021).
DT has been successfully applied mainly in the fields of manufacturing and PLM to perform simulations, optimization, make predictions and provide suggestions, performance improvement, and others (Wei, 2021). From these applications, the DT concept expanded to “augmented digital twin” to be applied in the healthcare field.
Scope
HDT technology features a model that analyzes a person’s health-related data from different sources to provide feedback to the patient and medical team on diagnoses, predictions of potential health complications or disease, and suggestions for treatment (Wei, 2021). The HDT is a replica of a real person that resides in cyber space; it is a digital version of a human being where the focus is synchronization of information (Miskinis, 2018).
As HDT technology is being further researched and developed, the most significant limitation it faces at this time is widespread adoption in clinical practice (Davenport & Kalakota, 2019). The technology seems capable of providing many benefits in the healthcare field, but widespread adoption may come after successfully conquering a multitude of challenges, such as artificial intelligence (AI) algorithm-related mistakes and biases making its way into the real person’s HDT model, impacting how the technology makes predictions and offers recommendations. There is also the actual integration of the technology into the field and existing infrastructures, teaching clinicians the use of the technology (Davenport & Kalakota, 2019).
Purpose
Research shows AI-related technologies like HDT have great potential in the transformation of patient care (Davenport & Kalakota, 2019). HDTs would be capable to predict treatment protocols based on attributes from the real human version, make predictions about patients acquiring a particular disease and other outcomes as HDTs would have all the necessary health-related information to carry out simulations (Davenport & Kalakota, 2019).
HDTs would enhance the patient-provider relationship with quick access to information, improved emergency response, reduced time spent on administrative tasks, and improved diagnostics (Bhagdev, 2021; Panetta, 2021).
Wei (2021) states medical treatment has evolved to be more customized, in essence, it is “smart medicine.”Supporting Forces
Challenging Forces
Several forces may pose challenges to HDT technology such as social, ethical, and cultural. In fact, the same economical and and technological supporting forces may also pose significant challenges to HDTs, such as AI-related technologies, for example, may have a negative impact on the cost of producing medical diagnoses and treatment recommendations (Davenport & Kalakota, 2019). The technology is advanced and costly to develop, and those added costs could potentially be passed on to end users.
Humans and their life cycles are extremely complex. It would be extremely difficult to integrate all of the complexities of humans into an HDT model, including the methodologies to ensure quality and reliable sources of data are being used (Wei, 2021), as it may negatively impact the accuracy of the HDT’s decision-making process (Davenport & Kalakota, 2019).
Methodology
The structured design process (SDP) is better suited for the construction of HDTs. SDP has its foundations on the nominal group technique (NGT), where each group member proposes a list of ideas and discussions take place as a group (Barnett, n.d.). SDP is used in software (SW)-heavy projects due to its approach of “divide and conquer,” by breaking down large processes into smaller, more manageable pieces to reduce complexities (Science Direct, 2022).
The HDT system architecture is to be deployed in the Cloud, where data from the real person and other diverse sources will be collected (data collection layer) from wearable smart devices, smart phones, treatment records at hospitals, etc. (Wei, 2021). Data received in different formats is to be cleaned and normalized prior to upload into the HDT model via encryption technology, from where analysis and/or simulations will then take place. The results would come as predictions, suggestions for treatment, and diagnoses (Wei, 2021).
Models
Figure 1. The HDT model. Adapted from “Is human digital twin possible?,” by Wei, 2021, Computer Methods and Programs in Biomedicine Update.
Figure 2. Layer architecture of the HDT application. Adapted from “Is human digital twin possible?,” by Wei, 2021, Computer Methods and Programs in Biomedicine Update.
Analytical Plan
The success of HDTs in healthcare can be evaluated by tracking the number of individuals that benefit from the technology, especially if those individuals are provided with timely feedback and visits to the hospital in emergency cases as a result of HDT intervention. End-user opinions and perceptions associated with level of technology understanding can become the first of a two-process evaluation, where the second would be an approach to measure the accuracy of the outputs.
The HDT system will undeniably face regulation, which at this time it is unknown the impact privacy laws, for example, would have on the technology. Once the technology reaches widespread adoption, the success of HTDs would be measured by the quality of the services it provides to their real human versions, such as treatment options, predictions, and recommendations.
Anticipated Results
Anticipated results of HTDs in healthcare include widespread adoption but not before facing several challenges that would potentially make the technology better and safer to use and interact with, including social and ethical issues that may arise by those opposed to the technology. The anticipated challenges would be highly intricate, such as the ability for developers to fully capture the complexities of the human body in a model, and how to accurately and effectively fuse the different sources of information being collected by the HDT system (Wei, 2021). There is also the issue of cost for both, technology development and end-users. Depending on results of potential regulation, government subsidies may be a possibility for willing participants that have also received a diagnosis where HDT can increase their chance of recovery or provide additional benefits to their overall healthcare program.
Conclusion
This sociotechnical plan dives into HDT technology and its potential to provide beneficial contributions to individuals and the healthcare industry. The Gartner 2021 Hype Cycle presents HDTs as an emergent technology in 10+ years (Panetta, 2021) that would offer digital replicas of real human beings in cyber space (Miskinis, 2018) to achieve information synchronization and be able to provide their real human versions with diagnoses, treatment options, recommendations, and predictions (Wei, 2021).
The HDT concept derives from an augmented version of the existing digital twin that has been successfully applied in manufacturing and PLM to carry out simulations, optimization, predictions, and make suggestions (Wei, 2021). The potential application of the DT concept on humans to create “smart healthcare” would attempt to predict illness or disease, offer treatment suggestions, diagnoses, and predictions (Davenport & Kalakota; Wei, 2021).
Areas of Future Research
There is a need to further research digitalization in the healthcare industry. Innovative technologies are being continually deployed in the areas of enhanced predictive analytics, AI capabilities, and powerful ML algorithms that would be key in achieving “smart healthcare,” including solidifying the HDT decision-making process as it processes larger and larger amounts of medical information that is constantly changing (Bhagdev, 2021; Davenport & Kalakota, 2019; Wei, 2021).
Comments
Post a Comment