Detecting physical activity within lifelogs towards preventing obesity and aiding ambient assisted living

Chelsea Dobbins, Reza Rawassizadeh, Elaheh Momeni
2017 Neurocomputing  
Dobbins, C, Rawassizadeh, R and Momeni, E Detecting Physical Activity within Lifelogs towards Preventing Obesity and Aid Ambient Assisted Living http://researchonline.ljmu.ac.uk/id/eprint/5044/ Article LJMU has developed LJMU Research Online for users to access the research output of the University more effectively. Abstract. Obesity is a global health issue that affects 2.1 billion people worldwide and has an economic impact of approximately $2 trillion. It is a disease that can make the aging
more » ... process worse by impairing physical function, which can lead to people becoming more frail and immobile. Nevertheless, it is envisioned that technology can be used to aid in motivating behavioural changes to combat this preventable condition. The ubiquitous presence of wearable and mobile devices has enabled a continual stream of quantifiable data (e.g. physiological signals) to be collected about ourselves. This data can then be used to monitor physical activity to aid in self-reflection and motivation to alter behaviour. However, such information is susceptible to noise interference, which makes processing and extracting knowledge from such data challenging. This paper posits our approach that collects and processes physiological data that has been collected from tri-axial accelerometers and a heart-rate monitor, to detect physical activity. Furthermore, an end-user use case application has also been proposed that integrates these findings into a smartwatch visualisation. This provides a method of visualising the results to the user so that they are able to gain an overview of their activity. The goal of the paper has been to evaluate the performance of supervised machine learning in distinguishing physical activity. This has been achieved by (i) focusing on wearable sensors to collect data and using our methodology to process this raw lifelogging data so that features can be extracted/selected. (ii) Undertaking an evaluation between ten supervised learning classifiers to determine their accuracy in detecting human activity. To demonstrate the effectiveness of our method, this evaluation has been performed across a baseline method and two other methods. (iii) Undertaking an evaluation of the processing time of the approach and the smartwatch battery and network cost analysis between transferring data from the smartwatch to the phone. The results of the classifier evaluations indicate that our approach shows an improvement on existing studies, with accuracies of up to 99% and sensitivities of 100%. There is undisputable evidence that indicates that engaging in regular physical activity is essential for healthy ageing and plays a key role in preventing premature death and several chronic non-communicable diseases (NCDs), including cardiovascular disease, coronary heart disease, stroke, diabetes, cancer, hypertension, obesity, depression and osteoporosis [1] [2] . Globally, the number of older people is increasing and by 2050 is expected to double from 841 million to 2 billion [3] . As a result, addressing the costs of such demographic changes is vital for any nation. Being physically inactive is a global issue that affects 1 in 4 adults and is the fourth leading risk factor for mortality around the world [2], [4] [5] . This has serious consequences not only for our health but also on the economy. Currently, 2.1 billion people worldwide are either overweight or obese, with the economic impact of this affliction costing approximately $2.0 trillion [6] . The challenge is therefore to ensure that healthy life expectancy (HLE), i.e. the average amount of years that we live without disease/injury, increases at the same rate as life expectancy, and allows people to work for longer [7] . Obesity is a preventable disease and with proper interventions can be tackled. Most importantly, education, behaviour change and personal responsibility are critical elements of any program to reduce the onset of this disease; however, these are complex, difficult, resource-intensive and time-consuming processes for individuals to achieve [6], [8]. Factors, including the availability and affordability of food, changes in diet, psychological triggers (e.g. stress), increasingly sedentary nature of many jobs (e.g. office work), transportation, and increasing urbanization all contribute to the rise of this epidemic [6], [9]. As we age, our mean body weight and body mass index (BMI) increases, and the effects of years of reduced physical activity and a poor diet becomes more apparent and dangerous in later life, which can lead to premature physical deterioration and cognitive decline [10-11]. As a result, research into the effects of physical inactivity is growing rapidly, with initial results indicating that there are important negative health outcomes for various markers of this type of behaviour [12] . Therefore, in order for us to age healthily and without debilitating illnesses, awareness and modifications towards our lifestyle choices are essential. In today's society, computing devices are now capable of capturing and storing a phenomenal amount of personal information and are increasingly being used to support lifestyle choices. For instance, smartphones and wearable technologies have enabled us to collect a wide range of personalized data that can be used for selfreflection and thus influence behavioural change. A consequence of such technology innovation, improved connectivity and low-cost sensors is the new era of the Internet of Everything (IoE) [13] , which builds on the Internet of Things (IoT) paradigm to a landscape where internet-enabled devices permit people, processes, data, and things to make networked connections that are more relevant and valuable than ever before. This shift is being driven by smaller and more powerful wearable devices that allow items such as smart watches, glasses, health and fitness trackers, etc. to be worn on the body to collect data and transmit this information, over the network, to provide real-time sensing [13] . This market has seen a surge in such items and by 2018 it is predicated that globally, there will be 177 million wearable devices [13] . Furthermore, by 2020 there will be 50 billion internetconnected devices, which will surpass the projected world population of 7.6 billion; thus equating to 6.58 connected devices per person [14] . Additionally, advances in the areas of wireless communication, home automation and medical monitoring systems are also revolutionising the healthcare industry so that healthcare can be transferred from the hospital into the home [15] . One area where this surge in information capture, storage, retrieval and distribution is prevalent is within the area of lifelogging [16] . Lifelogging is a subset of pervasive computing and refers to the unified and continual digital recording of an individual's experiences that have been captured using a variety of sensors and that are stored permanently in a personal multimedia archive [17] . In other words, it is a platform that can be used to gather a continuous flow of personalized information, over a sustained period of time. Such records can then be used for a variety of activities and studies, including self-reflection, health monitoring and other social and healthrelated studies. For instance, people can examine patterns of their behaviours to reflect on their levels of activity and use this information to improve the quality of life [18] . One important type of information that is required for this activity is physiological data. The prevalent use of wearable devices is a popular method to capture such bodily information as these devices offer low battery utilization and often house a multitude of sensors that are capable of detecting physiological signals. As a result, the reflection of such personalized information provides precise and clear feedback of the user's psychological state in real-time, which may reinforce or contradict the users' self-appraisal [19] . For example, we may think that we are quite active but reflecting on our lifelogging Chelsea Dobbins, Reza Rawassizadeh, and Elaheh Momeni, "Detecting Physical Activity within Lifelogs towards Preventing Obesity and Aiding Ambient Assisted Living," Neurocomputing, Dec. 2016Dec. , http://dx.doi.org/10.1016Dec. /j.neucom.2016.088 data can either confirm or deny this belief and may result in altering behaviour, such as by walking more. Consequently, collecting data over a sustained period of time yields a phenomenal amount of information [20] . Lifelogs are complex and are composed of a variety of media and information. As such, this heterogeneous nature means that uses simple queries and ranking search results is unlikely to support many of the user's information retrieval tasks in this domain
doi:10.1016/j.neucom.2016.02.088 fatcat:7qm322p465d5xc472bhn2ixyla