Around the world, armies are recognizing the importance of maximizing the effectiveness of Soldiers physically, perceptually, and cognitively. Militaries are therefore studying effects of frustration, mental workload, stress, fear and fatigue on both cognitive and physical performance.
Early detection of illness and injury often yields a better prognosis. For example, if left undetected, infectious diseases can spread quickly through a population, endangering Warfighters and their missions, as well as the general population. Earlier diagnoses of injury (e.g., traumatic brain injury) can prevent inappropriate return-to-duty of Warfighters who may put themselves and others at risk.
Currently, understanding and assessing the readiness of the warfighter involves medical intervention with the help of advanced equipment, such as electrocardiographs (EKGs) and other specialized medical devices, that are too expensive and cumbersome to employ continuously or without supervision in non-controlled environments. On the other hand, currently 92 percent of adults in the United States own a cell phone, which could be used as the basis for continuous, passive health, and readiness assessment.
Embedded sensors in mobile devices (e.g., personal smartphones and smartwatches), such as accelerometer, GPS sensor, and Bluetooth sensor, have been applied to monitor human behaviors and track daily activities. The resulted data from these embedded sensors can be used to infer human health status, monitor mental health states, and deliver medical interventions
Traditionally, handcrafted features are extracted from mobile sensing data to analyze human behavior patterns. For example, motion features such as magnitude of acceleration, can be extracted from accelerometer to study their correlations with different user contexts (e.g., location, activity, social context). Instead of using handcrafted features, which are usually based on heuristics and domain knowledge, high level complex features can be automatically extracted from mobile sensing data, and leveraged to improve generalization of predictive modeling using deep learning algorithms.
DARPA launched the Warfighter Analytics using Smartphones for Health (WASH) program in 2018, with aim “to develop algorithms that enable continuous and real-time assessment of the Warfighter by leveraging data that is passively and unobtrusively captured by cellphone sensors.”
Warfighter Analytics using Smartphones for Health (WASH) program
The objective of WASH is to extract physiological signals, which may be weak and noisy, that are embedded in the data obtained through existing mobile device sensors (e.g., accelerometer, screen, and microphone). Such extraction and analysis, done on a continuous basis, may help determine current health status and identify latent or developing health disorders.
WASH research will explore the development of algorithms and techniques for identifying both known indicators of physiological problems (such as disease, illness, and/or injury) and deviations from the warfighter’s micro-behaviors that could indicate such problems. It is also expected that additional “digital biomarkers” of physiological problems may be identified during the research through the combination of big data analytics and medical ground truth provided to performers. Digital biomarkers are consumer-generated physiological and behavioral measures collected through connected digital tools, in this case a smartphone.
A prerequisite for the extraction and interpretation of the raw sensor data and any identified digital biomarkers is determining the context of such data collection and analysis, which may affect the relevance of any given sensor and permit “denoising,” or elimination of irrelevant or misleading readings. For example, relying on cellphone accelerometer data while the warfighter is in a moving vehicle would likely negatively influence the utility of such data for WASH-type analysis unless the auxiliary motion is identified and cancelled.
Thus, key focus areas will be the extraction of the signal context and the identification of complicated actions and environmental variables, and the association of user state with symptoms of illness conditions in order to identify potential illnesses and conditions before conventional symptomatic display. It is the union of personal behavior/characteristics, smartphone sensor collection, context of smartphone use, and disease biomarkers that will define the preclinical health determination of the WASH program.
The program goal is to enable the creation of a mobile application that passively assesses a warfighter’s readiness immediately and over time. This application seeks to provide:
- Clinicians with plausible health conditions supported by the analysis, determination, and fusion of digital biomarkers for disease correlation against ground truth;
- Commanders with unit readiness information, both at the current time and in the near future; and
- Users of the device with information about their current health status and early indicators of medical conditions
The Defense Advanced Research Projects Agency awarded a $5.1 million contract to Kryptowire, for technology that can gather and aggregate smartphone sensor data. The Department of Defense hopes to deploy predictive analytics to track the health of U.S. service members with their phones.
Kryptowire, a Department of Homeland Security-funded startup that specializes in mobile app security, plans to develop a secure and context-aware tool to gather data collected from users’ smartphones – during both everyday use and during clinical trials – to help improve health outcomes.
The company aims to enable data collection across both iOS and Android smartphones, enabling ease of deployment at scale, data anonymization, secure access control to device data, and transparent data collection, officials said
Beyond just U.S. service members, the aim is to eventually develop this as a consumer-facing technology, offering a new way to detect health problems as early as possible.
“Our strategy is to leverage the full power of mobile to collect health metrics in all patient settings, for continuous monitoring, from clinic to home, and to build the ground truth from all available data, including smartphone sensors, clinical studies, medical examinations etc. for a better-informed, real-time approach to disease detection and biomarker identification,” said Kryptowire CEO Angelos Stavrou.
There are also privacy concerns arising due to continuous monitoring via smartphones’ sensors, cameras and microphones. “If you’re activating a microphone on someone’s phone, that is going to raise a lot of alarms,” Jay Stanley, senior policy analyst with the American Civil Liberties Union, told the Washington Post. “People don’t want to feel like someone is listening in on their private life. That’s going to have to be subject to tight controls.”
DARPA Awards Charles River Analytics Contract to Build Smartphone App that Detects Warfighter Illness and Injury, in May 2018
Charles River Analytics Inc., developer of intelligent systems solutions, has been selected as a prime contractor under DARPA’s Warfighter Analytics using Smartphones for Health (WASH) program, performing under both technical areas.
As part of the WASH program, Charles River is leading a team that includes Assured Information Security; Tozny, LLC; and the University of Washington to develop a Health and Injury Prediction and Prevention using Complex Reasoning and Analytic Techniques Integrated on a Cellphone App (HIPPOCRATIC App). With all options exercised, the four-year contract for HIPPOCRATIC App is valued over $15.5 million.
The HIPPOCRATIC App system is designed to detect indicators of an injury. For example, if it senses a large jolt that may indicate a fall, the system begins monitoring for associated symptoms, such as a change in gait, using integrated smartphone sensors. “The HIPPOCRATIC App can lead to a revolutionary breakthrough in healthcare delivery for both Warfighters and civilians,” said Dr. Bethany Bracken, Senior Scientist at Charles River and Principal Investigator for the HIPPOCRATIC App.
“Earlier diagnoses of injury can prevent return-to-duty of impaired Warfighters; earlier treatment means Warfighters can perform their duties longer, saving the Department of Defense millions of dollars in training costs. We are proud that we are working with such a talented team to build this app. While our team is excited about the potential medical outcomes, we are also excited to demonstrate the right way to protect end-user information and privacy when applying advanced analytics. We will protect the privacy of user data using InnoVault, an end-to-end encryption toolkit developed by Tozny.”
InnoVault is a commercial toolkit for end-to-end control of stored structured data using strong encryption. The toolkit can be embedded in an app, web page, or server to ensure that tight control is maintained over the data from the point of creation, through transmission, storage, analysis, and finally expiration and deletion.
“The HIPPOCRATIC App will rely on sensitive user data whose privacy must be preserved for the entire data lifecycle,” said Isaac Potoczny-Jones, Tozny CEO. “Using encryption from start to finish gives us fine-grained control and strictly limits the parties who even touch the data. This protects both the end-users and the research partners whose analysis engines will access the data.”