Dem@Care: Dementia Ambient Care - Multi-Sensing Monitoring for Intelligent Remote Management and Decision Support (Dem@Care)

Published on: 13/09/2013

The increase in average lifespan across the world has been accompanied by an unprecedented upsurge in the occurrence of dementia with high socio-economic costs. The development of personal health systems provides a means of dealing with such problems in a meaningful and sustainable manner, enabling people with dementia to maintain independence and inclusion in society, while improving their quality of life and the effectiveness of their caregivers. Multi-parametric monitoring of daily activities, lifestyle, behaviour, in combination with medical data, provides clinicians a comprehensive image of the person’s condition and its progression, as well as immediate notification of emergencies, without their being physically present, allowing for the remote care of their condition through the provision of personalised adaptive feedback.

The objective of Dem@Care is the development of a complete system providing personal health services to people with dementia, as well as medical professionals and caregivers, by using a multitude of sensors, for context-aware, multi-parametric monitoring of lifestyle, ambient environment, and health parameters. Multi-sensor data analysis, combined with intelligent decision making mechanisms, will allow an accurate representation of the person's current status and will provide the appropriate feedback, both to the person and the associated caregivers, enhancing the standard clinical workflow. Many research challenges will arise, ranging from data collection and analysis to integration, interpretation and feedback. Aggregation of information from complementary sources will be a critical aspect of multi-sensor processing that will be addressed, along with the advance of knowledge and data management methodologies, for scalable, adaptive and meaningful interpretation of the person's condition. The data will be analysed and interpreted in conjunction with established or newly created medical knowledge, for the production of decision support systems with personalised feedback tailored to two different user groups: the people with dementia and their caregivers (health professionals, family). Appropriate user-friendly interfaces will be developed, tailored to the needs of each user group, facilitating the rapid incorporation and seamless integration of the proposed system in the users’ lives.

In Dem@Care, people with dementia will be monitored by various sensors in activities of daily living, such as cooking or watching TV, cognitive activities and social interactions. At the first stage, this will take place in controlled environments, such as labs, and later on in nursing homes, or in the homes of people with dementia. These observations will be automatically analysed and interpreted for the correlation of behavioural patterns with the progress of their condition and the enablement of direct, personalised feedback, which is expected to significantly enhance their quality of life, increase their independence and improve their health.

The deployment and integration of a solution like that proposed by Dem@Care in peoples’ daily lives is expected to have significant positive effects both on them and their carers. In particular, people with dementia and associated health problems will enjoy increased independence and have a higher sense of safety and security, thanks to the continuous multisensory monitoring, which will assist them in their daily activities and ensure feedback where it is needed.

Policy Context

Dementia, a leading cause of disability in the elderly, currently affects nearly 10 million people in Europe and over 35 million worldwide. Rising at unprecedented rates, these figures are projected to increase to 14 and 65.7 million respectively by 2030. The socioeconomic repercussions are equally staggering. In Europe alone, the total costs of dementia amount to over €180 billion in 2010 and are estimated to exceed €250 billion by 2030. The aforementioned inflict a significant burden on healthcare systems, society and the economy, necessitating effective treatment means, while preserving quality of life for the people affected and for their carers. A significant goal in this direction is to support people to continue independently at home for a longer period of time while also remaining active members of society.

Current policies are now turning towards the prolongation of people’s lives at home, so they can remain active members of society and even the workforce. This aims to alleviate the burden on healthcare systems, care homes, hospitals and also families who often need to abandon work in order to take care of their loved ones. The European Commission has been making concerted efforts towards the support of Ambient Assisted Living and various assisted living solutions in this regard. Policies for ageing well are being developed[i], taking advantage of ICT, while funding is being made available for the development of ICT technologies that will address the problem of ageing. The FP7 framework programme and its successor, Horizon 2020, fund applied research in these directions.

The ageing challenge is also being addressed through European participation in the Ambient Assisted Living Joint Programme (AAL JP) 2008-2013[ii], which involves a high ratio of SMEs so as to develop applied research solutions to the ageing problem. Efforts to expand AAL JP[iii] are being organised, with a follow up programme planned for 2014-2020 which will involve a host of stakeholders.

Within this context of ever increasing use of technology to address health and ageing problems, Dem@Care is proposing and developing a very relevant and timely solution of multi-sensor monitoring, to allow for remote monitoring and care. Even more importantly, Dem@Care’s focus is on personalized and adaptive multisensory monitoring, which will provide a comprehensive rich picture of a person’s situation and the progression of their condition. Finally, the carefully and intelligently tailored feedback to the patient and their carers is expected to greatly help both the person with dementia in their daily life, and even ameliorate their condition, while the effectiveness and quality of care provided by healthcare professionals is expected to improve considerably.


Description of target users and groups

The problem of dementia affects an ever-increasing part of the world population, from individuals and their families to professional caregivers like nurses and doctors, and even the workforce. Patients are target users whose autonomy is to be increased through the installation and use of multiple sensors for discreet daily life monitoring that will lead to personalised adaptive feedback provided by appropriate user-friendly interfaces and devices. Their families and loved ones, part of the general public, are also involved since they will have the option of remotely monitoring the person while professional personalised feedback is administered to them. Family members will not be forced to leave the workforce in order to help their loved ones, since the patients’ independence will be supported by the proposed solution, while carers will receive feedback to monitor the person’s condition and its progression via appropriate devices. Health professionals will obtain continuous and comprehensive information about the person’s condition in an efficient manner via the multi-sensor monitoring, leading to more effective feedback and care. The proposed solution is expected to allow health professionals to work with a larger number of patients, since they will spend less time on appointments and discussions with patients and carers to learn about their condition. A better and more objective picture will be provided by the proposed system, giving health professionals the freedom to spend more time on treatment and support with the patient, rather than trying to obtain measurements to evaluate or diagnose the condition: a large part of this data will be provided by the system.

Thus, the initial group of target users will be people with mild dementia or mild cognitive impairment who are capable of using simple technologies, such as a wristband physiological sensor or a wearable life-logging camera. Their needs focus around their daily life, mostly inside the home. Dem@Care will help them carry out daily activities by reminding them – but only when and if needed, by providing life-logging information and feedback from formal and informal carers. Dem@Care will also help their physiological, societal and psychological needs, by making sure they are leading a healthy lifestyle, they are active enough and have sufficient social interactions. The people with dementia being targeted are generally aged over 65, of both genders, suffering suffer from MCI to other forms of dementia. At this stage, full-blown Alzheimer’s patients have a very significant loss of independence, so they are not capable of using such a system with full independence. However, with the assistance of a carer, certain parts of the unobtrusive monitoring can take place. It should be noted that the comprehensive monitoring provided by Dem@Care can facilitate the improved monitoring of other conditions as well, although this is not its ultimate goal. All people using Dem@Care should be capable of understanding the implications of the installation of such a system and of their participation in similar experiments and monitoring environments. For this reason, they are provided with consent forms, which they sign if they agree to participate, only after full understanding and consent is obtained.

Informal carers can be family members, to whom appropriate training and information about the system is made available so they can effectively use it. Professional carers are mostly experts in dementia who are interested in supplementing and improving their care options through the use in ICT. They are already expressing great interest in such solutions and are expected to greatly benefit from the monitoring options provided, as well as the accurate and reliable feedback provided to them by the system.

Description of the way to implement the initiative

Dem@Care aspires to contribute to the timely diagnosis, assessment, maintenance and promotion of self-independence of people with dementia, by deepening the understanding of how the disease affects their everyday life and behaviour. It implements a multi-parametric closed-loop remote management solution that affords adaptive feedback to the person with dementia, while at the same time including clinicians into the remote follow-up, enabling them to maintain a comprehensive view of the health status and progress of the affected person.

Dem@Care’s overall goal is to provide an integrated remote care and management solution for people with dementia by bringing together leading experts in dementia, video and audio analysis, physiological sensor data monitoring, life-logging, lifestyle analysis, data mining and fusion, knowledge modelling and semantic inference. The core of Dem@Care lies in the multi-parametric monitoring, analysis and interpretation of the recorded sensor data, and in the effective coupling with personal and relevant clinical (diagnostic & treatment) knowledge that realise the support and delivery of tailored and adaptive feedback to the person with dementia. It is the multi-parametric monitoring and behaviour interpretation that enables the extraction of objective and meaningful information about the condition of the person with dementia, and allows for pertinent and timely diagnosis and medical follow-up, along with the delivery of appropriate supportive feedback and care plans.

Current ‘clinical loops’ for monitoring, diagnosis and feedback for treatment will be significantly enhanced through the use of ICT as it takes place in Dem@Care. The resulting rich information about the person’s condition and its evolution will be made available to both the person with dementia and their carers in the form of appropriate feedback through two closed loops. In particular, the system includes:

  • a loop for people with dementia and their informal caregivers to monitor and assess their cognitive and behavioural status by integrating a multiplicity of wearable and in-situ sensors, enable time evolving context-sensitive profiling to support reactive and proactive care, and afford personalised and adaptive feedback.
  • a loop for dementia clinicians to provide objective observations regarding the health progression of the person with dementia and medication effectiveness, warn about trends closely related to dementia (e.g. apathy), and support preventive care decision making and adjustment of treatment recommendations.

In order to succeed in its goals, Dem@Care has several research/development challenges, for each of which there is a measurable success criterion:

  1. Protocol design is a challenge with the multisensing measurements needed. Dem@Care deals with this by developing a protocol which will be straightforward to incorporate in people’s lives.
  2. Physiological sensing will have to be as discreet as possible. This is addressed through the use of environmental, unobtrusive sensors. Wearable sensors are to be used only certain times of the day, while the system has enough redundancies to allow for gaps in measurements.
  3. Innovative audiovisual algorithms are being developed to deal with the challenges of analysing such recordings, from a person’s daily life.
  4. New medical knowledge is to be obtained through the correlation of the multisensor measurements with the person’s condition and its evolution, which can be put to use for more effective treatment options, further enhanced by the fact that the feedback obtained is personalised and adaptive.
  5. Real life testing of the system is to take place, both in controlled environments like hospitals, nursing homes and homes, which will allow the tackling of practical challenges that arise, ranging from the end user’s acceptance and use of the system, to the effectiveness/reliability of the algorithms used and the effectiveness of the feedback provided to both the people with dementia and their carers.

The management structure and procedures for Dem@Care are drawn from best practices in EU projects and are based on principles of product-based planning, delegation of responsibility and exception-based reporting to ensure coherent scientific, administrative and financial coordination and provide the support and tools required for the achievement of the project objectives. They are designed to ensure the successful collaboration and involvement of all contractors in the direction of the project, while encouraging efficient decision making tailored to the goals, size and consortium synthesis. Due to the complexity of Dem@Care, a Project Management Board and a Project Advisory Board will help with their expertise in guiding the project.

Technology solution

Dem@Care will capitalise upon continuous advances in pervasive computing and sensor technologies to power a multi-parametric monitoring framework that will sustain context-aware, personalised and adaptive feedback mechanisms for the remote management of people with dementia. To allow for the comprehensive assessment of a person’s condition and enable the early identification of alarming and potentially hazardous behaviour patterns, as well as prolong independent living, Dem@Care will take full advantage of the gamut of sensors available for collecting information relevant to dementia. These will include, among others, sensors for vital signs monitoring, location sensors, lifestyle sensors, such as accelerometers, light and door sensors, as well as audiovisual sensors, namely wearable and static cameras, and wearable microphones. The complementary nature of several sensor inputs is expected to increase the robustness of their processing results, through the fusion of their outputs.

Following a time-evolving, dynamic approach to user profiling, in combination with semantic representations of medical knowledge (including care management plans, determination of indicators for behavioural and psychological symptoms of dementia and unusual, or perhaps even alarming behaviours, etc.), Dem@Care will enable adaptive monitoring and feedback services tailored to each person with dementia through closed loops that will enhance the current clinical workflow, while enriching the body-of-knowledge in dementia care. To maximise effectiveness, Dem@Care will combine state of the art machine learning and logic-inspired principles for knowledge discovery and reasoning, so as to ensure the reliable, accurate and robust interpretation and assessment of the monitored person’s behaviour and health status. Key features underpinning the proposed intelligent decision support services are the definition of flexible inference methodologies that can scale to the volumes of medical knowledge and sensor data involved, and cope effectively with uncertainty (probabilistic, fuzziness, etc.) inherent to the analysis and high-level interpretation of sensory input.

Interoperability is central in Dem@Care, as it aims to provide a uniquely comprehensive and personalised solution for dealing with issues related to dementia outside a hospital through the combination of the multisensorial processing technologies and advanced knowledge and reasoning methodologies. As a central aim of the Dem@Care consortium is to provide benefit to the European community, many of the project partners either use Open Source code in their deliverables or contribute their deliverables to the Open Source communities. Alternatively, some of the partners may be contributing to Standards, be they open standards or other. Details concerning open source code use and standard contributions will be addressed in the Consortium Agreement. In no case will a Partner be able to contribute the Foreground of another Partner, to either a Standard or open source without the approval of the owning Party.

Technology choice: Proprietary technology, Standards-based technology

Main results, benefits and impacts

Dem@Care expects to enhance current clinical practises and afford new knowledge related to the diagnosis and management of dementia. More specifically, Dem@Care is anticipated to result in:

  • Clinical protocols that will effectively correlate sensorial inputs related to behavioural and cognitive patterns with dementia-specific parameters.
  • Innovative and highly effective signal, image, audio and video processing algorithms for the recognition and interpretation of a person’s condition from multi-sensing data recorded in real, uncontrolled conditions.
  • Elicitation and validation of new clinical knowledge for improved diagnostic precision and effectiveness.

These results will allow Dem@Care to have a strong impact on the quality of life of people with dementia, with direct effects on their informal carers as well, by providing:

  • Improved quality of care, through objective, comprehensive diagnosis, and customised reactive and proactive support.
  • Increased safety and sense of security, through real-time adaptive feedback and alarms.
  • Timely treatment of disease symptoms, afforded through early detection of health status degradation and through personalised cognitive support.
  • Sustained independence and autonomy, enabling to better cope with daily life functions, while reducing the need for constant monitoring.
  • Prolonged stay at home as a result of the sustained independence.
  • Possibility for the elderly to remain active members of society, in some cases even of the workforce.
  • Possibility for family members to remain active members of the workforce for a longer period of time.
  • Great alleviation of financial and psychological burdens, which result from the need to constantly care for a person with dementia.

In parallel, Dem@Care expects to have significant socioeconomic benefits, including:

  • reduced cost of clinical care;
  • reduced cost of monitoring;
  • delayed admittance to nursing facilities;
  • sustainable at-home care solutions.

Dem@Care’s scientific benefits are also significant, due to the complexity of the system, its innovative character and the challenges to be faced by its integration in real life situations. In particular, Dem@Care will contribute to:

  • The development of multi-sensor processing algorithms, from physiological sensing to audiovisual monitoring, in real environments. This is a great challenge due to the difficult, uncontrolled nature of real environments, the different setups in homes and the considerable anthropometric variations to be faced. The degree to which the people being monitored are incapacitated adds an additional degree of difficulty to the development of such algorithms, as the recognition of erroneously performed actions will have to take place.
  • The unobtrusive installation of such systems in homes, nursing homes or hospital environments will incur the development of innovative solutions to maximize its discreet nature and allow users to “forget it is there”.
  • The correlation of comprehensive long term multi-sensor measurements with the evolution of a person’s condition. The availability of such vast amounts of data from so many different sources is unprecedented and is expected to lead to new insights and knowledge about dementia and ways to live with it.
  • The integration and fusion of multiple sensing measurements and results with a person’s condition and its evolution, with and without the incorporation of feedback and care solutions. This will also allow the close study of various treatments’ effects on the person.
  • The development of truly useful feedback for incapacitated individuals. Elderly with dementia are a particularly sensitive age group and should be dealt with very carefully, but should still receive necessary feedback and support to increase their independence and sense of security and safety.
  • The provision of feedback to carers, which does not overwhelm them, but still achieves the provision of the most comprehensive, important and relevant information extracted from the data.

Track record of sharing

Dem@Care faces numerous challenges due to the complexity of the solution being proposed, which is at the same time one of its most important advantages and innovations. The multitude of sensing modalities being used is a challenge to synchronise and integrate in a seamless manner. Nevertheless, it is doable and necessary, due to the complementarity of the information provided and the new insights it provides, which cannot be obtained through traditional routes like questionnaires. Dem@Care’s modular nature allows the use of certain parts of the system, e.g. only a part of the sensing modalities. This increases its flexibility and allows its use in a wide range of environments and situations, where the use of the entire system is not always possible. It also greatly facilitates the potential for transfer and exchange, as parts of it can be replicated in other contexts, depending on the needs of various applications. It is noteworthy that Dem@Care’s monitoring results are not limited to dementia, but can help with overall care and wellbeing of a person, especially as dementia is often accompanied by other co-morbidities. Similarly, carers (informal and formal) have a lot to gain by studying the correlations of the large amounts of data accumulated with the person’s condition.

The project aims and results have been shared with the Greek Association of Alzheimer's Disease and Related Disorders[i], where participants tested parts of the system and took part in recordings for the production of benchmark datasets of Activities of Daily Living (ADLs).

Dem@Care will also be presented in the REACTION consortium’s clustering event[ii] where similar projects will be presented and will exchange information.

Under Dem@Care, two workshops have been organised:

1) “Ambient Telecare”[iii] at the International Conference on Innovation in Medicine and Healthcare[iv].

2) Workshop on ‘Multimedia Indexing and Information Retrieval’ (ACM MM MIIRH)[v] at ACM MM 2013.

Finally, Dem@Care is organising a summer school, DemAAL[vi], with over 40 participants, where clinical and technical aspects of Ambient Assisted Living technologies for remote health management, ageing well and independent living will be presented both in theory and in practice.


Lessons learnt

Dem@Care is still underway, and going through its most interesting, yet challenging stage. Real life pilot experiments are being organised and setup in different countries and different environments, ranging from hospital labs to nursing homes and individuals’ private homes.

One important lesson it that recruitment is very important, especially as many people with dementia are not very familiar with technologies: it is necessary to maintain a fine balance when explaining to them how the use of such a system can benefit their lives, while respecting the vulnerability of their situation. Nevertheless, the unobtrusiveness of most sensors and the improvements that such a solution can bring upon people’s lives makes many people quite eager to try it, especially if it does not introduce big changes in their daily life.

Another important lesson learnt from our case is the usefulness multi-sensing data and the revelations it can provide about a person’s situation. Professional carers are already informing us on the benefits of Dem@Care: the person’s condition as reported in traditional doctors’ visits and questionnaires is in many cases quite different of what the recordings show. In this manner, professional carers obtain a more objective, clear and comprehensive picture of the person’s situation and can recommend the appropriate treatment options. Additionally, they are able to help larger numbers of people thanks to the automations introduced by the system.

Finally, the provision of appropriate feedback to both the patients and their carers can play a crucial role in their acceptance and eventually continuous use of such a solution. Currently, the feedback provided to the people with dementia is very simple and discreet, being made available only when absolutely necessary and tailored to their individual needs. The feedback for the carers, on the other hand, is significantly more complex and features numerous prioritisation, search, and visualisation options, which can greatly help them in obtaining a clearer picture of the person’s situation. Thus, we conclude that carefully designed and provided feedback can greatly help the end users and vastly increase the potential for use and integration of Dem@Care in informal and formal dementia care, in daily and professional lives.

Scope: Pan-European