Approximately one in every three people aged over 65 years will fall every year; this ratio rises to about one in every two people for those over the age of 80 years. There are also large socioeconomic burdens associated with falls. The Australian Institute of Health and Welfare analysed a total of 9,924 injury-related deaths occurring in the community setting in Australia, for the years of 2003-04, showing that unintentional falls were the most common cause (30%) of injury-related death, followed by suicide (22%) and transport (17%). In Australia for 2005-06, falls represented 2.6% of all hospital admissions for persons aged 65 years and older, and the estimated number of injury-related events due to falls is rising year on year. The estimated cost of treating fall-related injuries in Australia is approximate $1b per annum.
Our research group is developing wearable ambulatory and environmental monitoring technologies to reduce the burden of falls among older individuals. The group’s inertial sensors and signal processing algorithms are developed in-house. They have developed algorithms to detect falls when they occur, or to predict future falls by identifying instability during normal movement, triggering the administration of a preventative rehabilitation program. They are also attempting to detect falls which occur at night time in the homes of older people living alone, using unobtrusive motion sensors and furniture load sensors; they have also investigated the use of 3D cameras systems, such as the Microsoft Kinect, for the same purpose.
Estimating daily activities and energy expenditure
In addition to falls, another serious health issue facing older people is a rising prevalence of cardiovascular disease, diabetes and obesity. Many management plans for these abovementioned disorders incorporate a recommendation for increased physical activity to improve health. However, without the capability to accurately monitor activity levels, it is difficult to discriminate whether the individual being managed is failing to adhere to the prescribed program, or is simply not responding to the management strategy. Therefore, unobtrusive monitoring of energy expenditure and activity is an area of research and development which has recently attracted significant attention.
Our group has been using our inertial sensors, described above in relation to the detection and prediction of falls, to estimate energy expended by people during their daily lives with better accuracy. While their device uses triaxial accelerometry and gyroscopy, one important improvement has been the addition of a barometric air pressure sensor, which allows the determination of when people are climbing stairs/hills. Furthermore, they have more recently demonstrated an implementation of these algorithms on a Samsung Galaxy mobile phone running the Android operating system.
Prof Nigel Lovell, Dr Stephen Redmond, Dr Michael Narayanan, Dr David Chang.