The health of Australia’s military could soon be monitored using data collected by smartwatches and wearable devices and analyzed using a machine learning algorithm. A new project plans to use such data to detect biological and chemical threats that defense personnel face in the line of duty.
Wearable devices like fitness trackers and smartwatches collect various health data, including physical activity, heart rate, blood oxygen levels, blood pressure, sleep, and skin temperature. Now scientists are looking to harness this data to ensure the health of members of Australia’s military.
Researchers from the University of South Australia (Uni SA) are leading a project to determine whether health data collected by smartwatches and wearable devices can give troops an edge in protecting them against biological and chemical warfare threats. They’ve teamed up with Australia’s Department of Defence, Insight Via Artificial Intelligence and the University of Adelaide.
The ability of these devices to continuously monitor health is what drew the researchers to consider them as a potential way to ensure the health of military personnel. It also came down to an issue of cost, as traditional methods of checking for infections are expensive and time-consuming.
“Most diagnostic methods involve sampling blood or nasal fluid to detect pathogens responsible for infections,” said Siobhan Banks, lead researcher from Uni SA. “This approach is costly, time-consuming and requires a laboratory for analysis. Consumer wearable devices continuously measure vital signs, including heart rate, skin temperature and sleep, creating huge sets of data for each person. Changes in these parameters occur very quickly after infection as part of the immune response."
Because vital sign changes indicating infection occur before symptoms do, the researchers say they can use the data to treat the infection more quickly.
“What if we could make use of this passively recorded data to detect the earliest molecular and cellular physiological events, caused by pathogen exposure, even prior to active infection?” Banks said.
The researchers plan to develop a machine learning algorithm to detect early signs of infection, ‘taught’ by the data collected by the devices. The project will use a cloud service that links to the wearable devices and a smartphone app. The algorithm will recognize irregular readings and cluster people based on their health profiles.
In addition to ensuring that military personnel are fit for duty, the researchers say the information gathered would be particularly useful in detecting exposure to biological or chemical weapons early.
Source: University of South Australia