Personalised ambient monitoring (PAM) of the mentally ill

James, C.J., Crowe, J., Magill, E., Brailsford, S.C., Amor, J., Prociow, P., Blum, J. and Mohiuddin, S. (2008) Personalised ambient monitoring (PAM) of the mentally ill. In 4th European Conference of the International Federation for Medical and Biological Engineering. Springer. pp. 1010-1013 . (doi:10.1007/978-3-540-89208-3_240).

Abstract

One in ten of the (UK) population will suffer a disabling mental disorder at some stage in their life. Bipolar disorder is one such illness and is characterized by periods of depression or manic activity interspersed with stretches of normality. Some patients are able to manage this condition via their self-awareness that enables them to detect the onset of debilitating episodes and so take effective action. Such self management can be achieved through a paper-based process, although more recently PDAs have been used with success. This presentation will introduce the Personalised Ambient Monitoring (PAM) concept that aims to augment such processes by automatically providing and merging environmental details and information relating to personal activity. Essentially the PAM project is investigating what may be loosely referred to as ‘electronic’ monitoring to automatically record ‘activity signatures’ and subsequently use this data to issue alerts. The types of data that we are considering using includes: location and activity (e.g. via GPS and accelerometers); and environment (e.g. temperature and light levels). Other types of sensor under consideration are passive IR sensors (within the home); and sound processing to log the audio ‘environment’. The use of such monitoring will be agreed between the patient and their health care team and it is anticipated that different patients will be comfortable with different sensor packages, thus personalizing the monitoring. Although such tele-monitoring is now generally common, its use in the treatment of the mentally ill is still in its infancy. This paper will consider the specific problems faced in applying it to this community along with the aims of this project. In addition, the use of modelling to predict the effects of the possible problems of sparse data that is expected, and to predict the effect on the overall patient pathway will be considered.

This record has no associated files available for download.

Contributors

Author: S.C. Brailsford

Download statistics