A Paradigm Shift from Telemedicine to Autonomous Human Health and Performance for Long-Duration Space Missions

This paper discusses a Prognostics and Health Management [PHM]-based approach to implementing Human Health & Performance [HH&P] technologies. Targeted specifically are NASA’s “Autonomous Medical Decision” and “Integrated Biomedical Informatics” of “Human Health, Life Support, and Habitation Systems” in Technology Area 06 [TA 06] of NASA’s integrated technology roadmap [April 2012]. The proposed PHM-based implementation is to bridge PHM, an engineering discipline, to the HH&P technology domain to mitigate space travel risks by focusing on efforts to reduce countermeasure mass and volume, and drive down risks to an acceptable level. NASA’s Autonomous Medical Decision technology is based on wireless handheld devices and is a result of a necessary paradigm shift from telemedicine to HH&P autonomy. The Integrated Biomedical Informatics technology is based on Crew Electronic Health Records [CEHR], equipped with a predictive diagnostics capability developed for use by crew members rather than by healthcare professionals. This paper further explores the proposed PHM-based solutions for crew health maintenance in terms of predictive diagnostics to provide early and actionable real-time warnings to each crew member about health-related risks and impending health problems that otherwise might go undetected. The paper also discusses the paradigm’s hypothesis and its innovation methodology, as implemented with computed biomarkers. The suggested paradigm is to be validated on the International Space Station [ISS] to ensure that crew autonomy in terms of the inherent predictive capability and two-fault-tolerance of the methodology become the dominant design drivers in sustaining crew health and performance.


INTRODUCTION
For manned space exploration missions beyond Low Earth Orbit [LEO], regular resupply of consumables, delivery of new supplies and replacement components, as well as emergency quick-return options may not in general be easy, timely, or feasible.Success in such space missions requires solutions to difficult technical challenges, built on proven capabilities, which may require the development of new capabilities arising from the development of novel cuttingedge technologies.
The key to supporting the objectives of the "Global Exploration Roadmap" report (ISECG, 2013) lies in the development of technologies and capabilities that enable the testing of new and innovative concepts, approaches, countermeasures, and techniques to maintain crew health and performance.Specific requirements and recommendations (Williams, 2011 andVolkov, 2013) have been provided by experienced astronauts, including crew members, who have extensive first-hand experience aboard the International Space Station [ISS].These requirements and recommendations have triggered the development and validation of PHM-based technologies to enable autonomous health monitoring and tracking on space exploration missions requiring the paradigm shift from telemedicine to HH&P autonomy.

Definition of Prognostics and Health Management (PHM)
As defined in (Uckun, Goebel, and Lucas, 2008) Prognostics and Health Management (PHM) is an engineering discipline that focuses on the fundamental principles of system failures in an attempt to predict mean time between failures [MTBF], and links the principles to system life cycle management.PHM-based technology is a key enabling technology to provide early warning of failure, and assesses the potential for life extension.PHM concept implementation is now a required design feature for space systems (Uckun et al., 2008); as such, all new space systems must have built-in PHM elements such as failure tracking.
As a concept, PHM also enables systems to assess their own real-time performance (self-cognizant health management and diagnostics) under actual usage conditions and adaptively enhance life cycle sustainment with riskmitigation actions.

The "PHM for Astronauts" Paradigm
One particular PHM-based paradigm, "PHM for Astronauts" (Popov et al., 2013), with corresponding solutions could bring Human Health and Performance [HH&P] technologies to the required Technology Readiness Level [TRL] in order to mitigate the HH&P risks of manned space exploration missions.
To validate the "PHM for Astronauts" paradigm on the ISS, this paper discusses in detail particular PHM-based solutions for HH&P technology candidates, such as "Integrated Medical Equipment and Software Suite" (a.k.a."Integrated Biomedical Informatics" and "Autonomous Medical Decision" per (NASA, 2012)) and "Integrated Prevention and Treatment for Visual Changes and Non-Invasive Intracranial Pressure Measurement", as per NASA designation (NASA, 2015).Also, according to (NASA, 2016) and NASA's guidelines and procedures, actual ground testing is mandated of any technology/system before and in parallel with an end-to-end test utilizing NASA testbeds for the purpose of fulfilling NASA's requirements and certifications for space missions.
By default, such ground testing must always be part of the space technology/system development lifecycle.
Early self-diagnostic, prognostic, and autonomous identification of proper preventive responses to negative trends are critical in order to keep astronauts healthy with limited medical support.Personal health-tracking and health-management tools are required to predict future health conditions if no preventive measures are taken.

Verification and Validation of PHM-based Technology
Verification and Validation (V&V) activity is essential for TRL development and maturation to ensure proper maturity and performance of the PHM-based technology.Roychoudhury, Saxena, Celaya, and Goebel (2013) suggest the following definition of the V&V activity.Verification of a product/technology is the process in which stakeholders answer the query "are we building it right?",while validation of a product/technology is the process in which stakeholders answer the query "are we building the right thing?"In other words, verification is the quality control process of evaluating whether or not a product/technology complies with testable constraints imposed by requirements at the start of the development process.In contrast, validation is the quality assurance process of evaluating whether or not a product/technology accomplishes its intended function when fielded in the target application domain.with a predictive advanced diagnostics and prognostics capability incorporated into smart checklists and PHMbased algorithms could enhance the healthcare delivery on long-duration space missions.Data mining of historical and current biomedical and clinical data (beginning with the earliest space missions, up to and including current ISS missions) has been identified by a number of experts as a critical task in order to customize and further define and validate the PHM-based algorithms (Popov, Fink, McGregor, and Hess, 2016). Behavioral health and performance;

NASA TECHNOLOGY ROADMAP: HUMAN HEALTH, LIFE SUPPORT, AND HABITATION SYSTEMS
 Human factors and performance.
Additional focus areas include:  Tracking before and after missions;  Providing lessons-learned feedback as input for followon mission planning.

Medical Diagnosis/Prognosis
The objective of the Medical Diagnosis/Prognosis functional area is to provide advanced medical screening technologies for individuals selected to the astronaut corps, which is to be implemented prior to crew selection for specific missions; this is a primary and resourceeffective means to ensure crew health.

Long-Duration Health
The focus of the Long-Duration Health functional area is on providing validated technologies for medical practice to address the effects of the space environment on human systems.These capabilities can provide significant benefits for after-mission follow-up.

Behavioral Health and Performance
The objective of the Behavioral Health and Performance [BHP] functional area is to provide technologies to reduce the risks associated with extended space travel and Earth return.Technology advancements are needed for assessment, overall prevention, and treatment to preclude and/or manage deleterious outcomes as mission duration periods extend beyond six months (e.g., a trip to Mars).(2015).The papers (Fink, Popov, and Hess, 2014), (Popov, Fink, and Hess, 2013), (Popov, Fink, McGregor, and Hess, 2016), and (Kevorkova and Popov, 2016)  This discussion provides further explanation of the technologies as well as a summary table of the priority technologies and system functional areas of interest, the current SOTA, the major challenges for advancement, and the recommended milestones and activities to advance to a TRL-6 and higher, i.e., demonstration of the technology in a relevant mission environment or simulation thereof.

WHAT MAKES THE PROPOSED PARADIGM DIFFERENT
Manned space exploration objectives present significant new challenges to crew health, including the psychological and physiological effects of long-duration space missions.
The limited communication with ground-based personnel for diagnosis and consultation of medical events creates additional challenges during such missions.Providing healthcare capabilities for space exploration missions necessitates the definition of new requirements and development of technologies in order to ensure crew health, and thus mission success.As we go deeper into space and mission durations greatly increase, the dependence on telemedicine will need to decrease and the use of autonomous solutions, many enabled by PHM capabilities, will need to increase.Implementation of the proposed PHM-based healthcare paradigm enables identification of predictors and early detection of deterioration or impairment of astronaut health − before signs are detected or symptoms are manifested (see also "stressors" in Popov et al., 2013).The predictors could be implicit or explicit.For clarity, a discovered pattern or correlation as a result of data processing could serve as an example of an implicit predictor, while increase/decrease of a parameter measured by a sensor network could serve as an example of an explicit predictor or onset detector.
Imbalance in skin pH resulting from impending dehydration is an example of an explicit onset detector of the associated medical condition, while heart rate variability as an electrocardiogram [ECG] morphology parameter is an example of an implicit predictor.

Conventional Medicine Paradigm
Focus is on keeping astronauts healthy by predicting a deterioration or impairment of health before a sign is detected or a symptom is manifested  (Popov, A., Fink, W., McGregor, C., & Hess, A., 2016).
Similar to the approach taken to develop successful PHM systems for mechanical and electronics industries, a multidisciplinary system engineering approach to diagnostics and prognostics allows the development of a PHM system related to human health support on manned space exploration missions.This is part of the engineering element for Condition Based Management plus [CBM+] that provides a key element and focused approach to making this happen.Such an approach and criteria will need to be developed for all system hardware and software elements.The thoughtful and effectuated integration of this approach is key, as is an integrated organization team made up of the necessarily diverse domain member experts.Predictor identification, onset detection, and real-time monitoring followed by CBM+-based responses are key features making the proposed new paradigm different.Moreover, it is to be noted that most, if not all, of the wireless medical devices existing in today's market are limited to data acquisition, transmission, and conventional representation.As has been shown in Fink, Popov, and Hess (2014), Popov, Fink, and Hess (2013), and Popov, Fink, McGregor, and Hess (2016) the real-time health monitoring architecture should include all standard elements from data acquisition to processing to interpretation (and potentially intervention).
Inherent to the HeSS proposed for HH&P, a built-in diagnostics and encompassing self-test capability as part of CBM+ for the hardware of the PHM-based system as a whole and its components helps to ensure reliable and assured information.The properly implemented capability for space exploration missions in terms of HH&P autonomy should ensure data consistency and eliminate false indications, e.g., false-negative or false-positive test results that otherwise could cause crew stress and unnecessary responses to incorrect messages received from the system.

COMPUTATIONALLY GENERATED BIOMARKERS
A biomarker refers to a broad subcategory of medical signs that is, objective indications of the medical state observed from outside the patientwhich can be measured accurately and reproducibly (Popov, Fink, McGregor, and Hess, 2016).Medical signs stand in contrast to medical symptoms, which are limited to those indications of health or illness perceived by the patient.A biomarker is a health-related characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention.Electrocardiogram (ECG), sweat diagnostics, and other laboratory tests, such as cholesterol and blood sugar level assessment, are examples of biomarkers.
However, in addition to biomarkers defined by the medical community, there is the notion of derived "computational" biomarkers, invented by data analysts.By definition, these are biomarkers that are generated indirectly by applying computation to health-related data.For example, some computational biomarkers, such as ECG morphological variability (a.k.a.heart rate variability), are used in cardiovascular risk stratification to help match patients to therapies.Other examples, related to vision testing and ophthalmology pertaining to Vision Impairment and Intracranial Pressure [VIIP] syndrome, can be found in (Fink, Popov, and Hess, 2014).The heart is a muscle designed to pump blood through the body, and as such it can be viewed as an electrical device of sorts.The premise of morphological variability is that electrical instabilities in the human heart do not occur without cause.Rather, there are generally numerous smaller issues which are typically asymptomatic; these issues can collude in specific ways such that the set of smaller issues becomes a larger problem culminating, for example, in myocardial infarction.
Small issues may not be detectable with the human eye or with a single measurement or test instrument, leading a cardiologist looking at an ECG to determine that a patient's heart health is normal.But using ECG morphological variability as a computational biomarker would draw the cardiologist's attention to the cause of the variability.For example, while some ECGs would be considered subthreshold, others may be determined to be supra-threshold in terms of high risk, i.e., while some patients are at low risk, others are at high risk.
A key objective of the implementation of the suggested PHM-based paradigm and resulting technology solutions with predictive self-diagnostic capability is the identification of such computational biomarkers in order to stratify health-related risks and then match them to a corresponding health maintenance plan or regimen.

INTERFACE DESIGN
Crews on space exploration missions beyond LEO will be required to function without timely support from a mission control center or ground personnel.That is why a personcentered design of health monitoring, and a support system with an intuitive interface and effective information architecture, are critical for PHM-based solutions in terms of HH&P autonomy.As the paradigm shift from telemedicine to crew-based healthcare autonomy takes place, the interface design will assume a high-priority task.Built-in self-training modules of the suggested PHM-based technologies that support initial, recurrent, and just-in-time instructions on health support and healthcare delivery should ensure that crew members with potentially (and often likely) limited medical skills can apply the proper knowledge to both normal and abnormal health situations.
The current interface design and training methodologies on the ISS program reflect decades-old technology.For crewed space exploration missions with limited ground-based support (e.g., missions to and ultimately settlement on Mars), a more intuitive dashboard-based interface is required.The interface should be also customizable and equipped with a user-friendly language designed for astronauts, rather than for healthcare professionals.As shown in Popov, Fink, McGregor, and Hess, (2016) the interface should require minimal astronaut training and provide intuitive common (i.e., cross-cutting) operability between different systems, as well as increased capability, usability, and reliability.The design of such interfaces is another challenging but critical task.

CREW ELECTRONIC HEALTH RECORDS
The expected growth of both electronic medical records [EMR] and electronic health records [EHR] is spurred by the emergence of technologies and tools for self-diagnosis and autonomous preventive health maintenance based on predictive capabilities.Yet, such technologies combine advanced improvements with new risks.Although EMR and EHR have been in existence for over ten years, the selfdiagnostics and autonomous health maintenance technologies still have not been widely adopted (Popov et al., 2013).Aboard the ISS, the EHR-based healthcare that is provided and maintained by the Crew Medical Officer [CMO] is not an exception, since crew members should have the assurance that they properly interpret and understand the received recommendations, as well as know how to implement them in order to have the best chance possible of achieving the desired outcome.But most importantly, safety considerations should be always a part of any system design.Given that there is limited-to-CMO medical support on space exploration missions, health-tracking, self-diagnostic, and health management tools are required to predict future health conditions if no preventive measures are taken.
In addition to standard EHR items, such as vital signs and body mass measurement, at least the following types of information should be included in CEHR: NASA has extensive experience with electronic records systems used as research data repositories embedded into decision support systems (NASA, 2018).The life sciences data archive was created to retain human research data from both ground and flight experiments on astronauts and other test subjects.The Longitudinal Study of Astronaut Health [LSAH] in (NASA, 2018) is similar to those used to record clinical data collected during routine healthcare, from medical data acquired during a mission, and from occupational health surveillance data.A number of features have been incorporated to ensure data security while providing access to research data.
The increasing application of informatics in medicine has resulted in enhanced application of clinical decision support systems, i.e., information systems to enhance clinical decision making in healthcare.These are defined as "active knowledge systems which use two or more items of patient data to generate case-specific advice" based on the integration of a database of medical knowledge, patient data, and some form of artificial intelligence or inference engine (Fink, Hess, and Popov, 2014).The application of such systems is to enhance on-board clinical diagnosis and adherence to the condition-specific guidelines outlined in the respective health status checklists that are used by the ISS Integrated Medical Group [IMG].
Numerous experts have articulated a series of controversial statements regarding decision support systems, asserting that the field of biomedical informatics is inherently aimed at enhancing the quality of decisions made by health professionals rather than by the patients themselves.The focus on a health support system with predictive capability coupled with an adviceon-response feature would be an enabling factor to healthcare autonomy for long-duration spaceflight.As shown in Fink, Hess, and Popov, (2014) other controversial statements of relevance to healthcare in space include unproven assumptions that the crew will use knowledge-based systems and standalone decisionsupport tools.The current paradigm is to integrate decision support tools into electronic medical records that incorporate protocols, guidelines, and educational materials into the development of information-enabled, decision-supported health data and intervention management systems (NASA, 2015).
The feature of using medical monitoring technology to raise alerts in the case of deteriorating health conditions is widely used in current medicine practice.However, despite the presence of such systems to alert healthcare professionals, significant physiologic change in the crew has taken place often before these critical alert systems are activated, as shown in (ASEB/NRC, 2011).
An individual monitoring PHM-based technology has also an additional advantage that current health approaches do not provide: If a crew member receives an alert message, the technology could check related datasets directly for each crew member for common precursors or biomarkers before symptoms manifest.Unlike current tools, the technology provides an early protection method for exposure to unknown materials or environments which are detrimental to the space exploration mission.Such a powerful approach could help with the ontological risks to protect the crew as well.

CONCLUSION
Extending ISS operations into the following decade will facilitate sustained testing and associated technology advancements in preparation for deep-space, longduration missions.The development of realistic groundbased simulators and an in-space test bed are crucial to the development, verification, and validation of technologies.
To achieve HH&P autonomy paradigm goals relating to long-duration missions, a series of critical operational, programmatic, research, and development questions must be addressed in a timely manner.
The inclusion of autonomous healthcare into space exploration programs should be accomplished through the development of advanced technologies to assist the crew in their health management in terms of this paradigm shift.In particular, there is a requirement in NASA (2012), NASA (2015), and Williams (2011) to develop a suite of integrated advanced healthcare technologies that will assist in the real-time monitoring of the health status of the crew and the environment.Such a PHMbased system is tentatively called a Health Support System [HeSS].
A PHM-based modular decision support system with predictive capabilities would integrate digital physiologic sensors and data management technologies, as well as imaging technologies, into an integrated Webbased system.The system would have an intuitive and customizable interface with user-friendly language designed for use by an astronaut, rather than by a healthcare professional.This amounts to a far-reaching requirement for the development of an advanced technology with predictive capabilities that will assist in the monitoring of a crew's health status.This allows for prediction of impending health issues, keeping crew members cognizant of their health status, and provides for timely advisory countermeasures before such issues impact crew health and performance.
The ISS is an invaluable asset in the race to develop future healthcare requirements for long-duration missions.To maximize the benefits received, proper health data collection, analysis, reporting, and discussion amongst the space community, including engineering and medical communities, are required.Healthcare planning for the next decade of ISS utilization should be based on analysis of in-orbit experience and lessons learned in the first decade of ISS utilization.Building additional healthcare test objectives into the research priorities of the ISS program to develop a mature PHM-based predictive capability should help to further fulfill the role of the ISS as an exploration-enabling research platform.A health support system with predictive capabilities incorporated into smart checklists and PHM-based algorithms could enhance the delivery of healthcare on space exploration missions.This capability is particularly important as signal latency increases in proportion to the increased distance from Earth.The healthcare planning for such missions should begin immediately to prepare capabilities that will meet the unique requirements of providing the highest health support standards.Also, there is a need for a modular design of the proposed PHMbased system to allow for incremental adjustments to improve the system as it becomes available.The modularity that is implemented in terms of a two-faulttolerance solution could also provide for commonality in components reducing the logistic footprint and enabling both interchangeability and operations continuity.Such modular design solutions should be inherent to any modern space system and, particularly, for crewed space exploration missions, since mass and volume are crucial mission design elements.
Data mining biomedical and clinical data from the earliest to current space flights has been identified by a number of experts as a critical task.Along with conventional data analytics the "Big Data" analytics [BDA], another essential PHM component, could help to overcome some of the challenges involved.BDA is defined as an information management approach and a set of capabilities for uncovering additional value from health information."Big data" provides new opportunities to store and index previously unusable, siloed, and/or unstructured data for additional use by healthcare stakeholders.Applying BDA creates new business value by transforming these previously unusable data into new predictive insights and actionable knowledge.Getting the crew members engaged with their health maintenance at a proper level is necessary to improve the healthcare on manned space programs (Williams, 2011).
Being mindful of benefits of traditional PHM components, such as predictive data analytics, the authors suggest that to ensure success in HH&P autonomy the best strategy integrates the capabilities of both of the following: 1.A combination of the fundamental understanding of biological, physical, chemical, and electrical processes (a.k.a.model-driven approach); 2. Empirical methods (a.k.a.data-driven approach).
Many of the currently available commercial-off-the-shelf [COTS] biosensors and physiology-monitoring handheld devices with related software are more akin to beta versions than properly certified products.Nevertheless, the products could be an option to validate wireless PHMbased healthcare technologies with predictive capabilities on space exploration missions.A major barrier to greater adoption of COTS wearables is that they continue to be used as stand-alone devices instead of being integrated into an interoperable ecosystem, including "Big Data" applications, to provide healthcare at the required levels.
Both (NASA, 2012) and (NASA, 2015) provide technology candidates in the specific domains of Human Health, Life Support, and Habitation Systems [TA 06], necessary to achieve NASA's goals in human space exploration over the next few decades.The technology area breakdown structure is represented in Figure 1 and refers to the sub-technology areas [sub-TAs] included in the roadmap.

Figure 1 .
Figure 1.Technology Area Breakdown Structure from (NASA, 2015).Because the HH&P technologies are specifically oriented to help maintain the health of the crew and support optimal and sustained performance throughout the duration of a mission, the HH&P domain includes the following four functional focus areas as shown in (NASA, 2012) and (NASA, 2015):  Medical diagnosis/prognosis;  Long-duration health;


Nutritional and caloric intake;  Circadian actigraphy;  Sleep logs;  Cognitive performance measurement tools;  Physiologic audiometry;  Ocular tonometry;  Visual acuity;  Musculoskeletal assessment of muscle strength;  Bone mineral density.Traditional EHR could be integrated with all available information collected before, during, and after a mission, including non-traditional but relatable information.Some examples of non-traditional information include:  Environmental exposure histories;  Family and other psychological stressors;  Financial stressors.
volume, and drive down risk to an acceptable level.The technology solutions include wireless handheld devices and are supportive of the paradigm shift from telemedicine to health support autonomy.A key component of the technology solutions is a Crew Electronic Health Records [CEHR]-based system with predictive diagnostics capability, developed for crew members rather than for healthcare professionals.
The PHM-based paradigm and corresponding technology solutions bridge PHM to the HH&P domain to mitigate space travel risks by focusing on efforts to reduce countermeasure mass and To elevate HH&P technology to a TRL-6 level, this paper also discusses employing PHM principles such as Condition Based Maintenance [CBM], as well as techniques with data fusion and data mining capabilities.The purpose is to assess the value of CEHR augmented with real-time data monitoring for accurate predictive diagnostics on manned space exploration programs.The primary benefit of the development and deployment of these technologies for the HH&P domain, bringing them up to a TRL-6 level, is the ability to successfully achieve and sustain affordable human space missions to Low Earth Orbit and beyond.Continued space missions on the International Space Station Program [ISSP] directly contribute to the knowledge base and advancements in HH&P, as the ISS is currently the only test platform for crewed space missions in an actual space environment.Ground-and space-based test beds such as long-term Mars Space Habitat simulators and the ISS are crucial to the development and validation of the technologies needed for long-term, manned space exploration missions.
heavily dependent upon the incorporation of new technologies integrated into Personal Area Networks [PANs].A PAN should interface with wireless Local Area Networks [LANs] to incorporate health-related data in electronic health records.A Health Support System [HeSS] suggested in the roadmap.Particular technology candidates, namely the Integrated Biomedical Informatics and Autonomous Medical Decision technologies of the roadmap, are discussed in the following sections.
Some subject matter experts authoring NASA(2015)believe that each activity or milestone represented in the roadmap does indeed have a technology solution to pursue at the present time, or will have one within the time frame