In this paper, we examine at-home activity rhythms and present a dozen of behavioral patterns obtained from an activity monitoring pilot study of 22 residents in an assisted living setting with four case studies. Established behavioral patterns have been captured using custom software based on a statistical predictive algorithm that models circadian activity rhythms (CARs) and their deviations. The CAR was statistically estimated based on the average amount of time a resident spent in each room within their assisted living apartment, and also on the activity level given by the average n.umber of motion events per room. A validated in-home monitoring system (IMS) recorded the monitored resident's movement data and established the occupancy period and activity level for each room. Using these data, residents' circadian behaviors were extracted, deviations indicating anomalies were detected, and the latter were correlated to activity reports generated by the IMS as well as notes of the facility's professional caregivers on the monitored residents. The system could be used to detect deviations in activity patterns and to warn caregivers of such deviations, which could reflect changes in health status, thus providing caregivers with the opportunity to apply standard of care diagnostics and to intervene in a timely manner.

The objective of this study was to assess the impact of passive health status monitoring on the cost of care, as well as the efficiencies of professional caregivers in assisted living. We performed a case-controlled study to assess economic impact of passive health status monitoring technology in an assisted-living facility. Passive monitoring systems were installed in the assisted-living units of 21 residents to track physiological parameters (heart rate and breathing rate), the activities of daily living (ADLs), and key alert conditions. Professional caregivers were provided with access to the wellness status of the monitored residents they serve. The monitored individuals' cost of medical care was compared to that of an age, gender, and health status matched cohort. Similarly, efficiency and workloads of professional caregivers providing care to the monitored individuals were compared to those of caregivers providing care to the control cohort in the control site. Over the 3-month period of the study, a comparison between the monitored and control cohorts showed reductions in billable interventions (47 vs. 73, p = 0.040), hospital days (7 vs. 33, p = 0.004), and estimated cost of care (21,187.02 dollars vs. 67,753.88 dollars with monitoring cost included, p = 0.034). A comparison between efficiency normalized workloads of monitoring and control sites' caregivers revealed significant differences both at the beginning (0.6 vs. 1.38, p = 0.041) and the end (0.84 vs. 1.94, p = 0.002) of the study. The results demonstrate that monitoring technologies have significantly reduced billable interventions, hospital days, and cost of care to payers, and had a positive impact on professional caregivers' efficiency.

Falls are very prevalent among the elderly. They are the second leading cause of unintentional-injury death for people of all ages and the leading cause of death for elders 79 years and older. Studies have shown that the medical outcome of a fall is largely dependent upon the response and rescue time. Hence, a highly accurate automatic fall detector is an important component of the living setting for older adult to expedite and improve the medical care provided to this population. Though there are several kinds of fall detectors currently available, they suffer from various drawbacks. Some of them are intrusive while others require the user to wear and activate the devices, and hence may fail in the event of user non-compliance. This paper describes the working principle and the design of a floor vibration-based fall detector that is completely passive and unobtrusive to the resident. The detector was designed to overcome some of the common drawbacks of the earlier fall detectors. The performance of the detector is evaluated by conducting controlled laboratory tests using anthropomorphic dummies. The results showed 100% fall detection rate with minimum potential for false alarms

This paper describes a non-contact imaging-based method to detect stage I pressure ulcers over a wide range of melanin levels. Two approaches were explored: the first used broad and narrow band visible spectrum imaging, and the second used near infrared (NIR) imaging. Preliminary results are presented together with results of numerical analysis of different erythema indices derived from the visible spectrum images. The results have shown that a low-cost imaging-based approach to detecting pressure ulcers is feasible and can yield promising results when applied to subjects with darker skin pigmentation

It has been observed in previous studies that the detection of stage I pressure ulcers becomes more difficult by unaided visual inspection and/or by using currently available techniques with darker skin subjects, due to increased melanin content. This difficulty is indicated by the elevated proportion of black and hispanic patients developing more serious stage III and IV pressure ulcers compared to white patients. The ultimate goal of this project, undertaken by MARC at the University of Virginia, is to develop a low-cost, non-contact imaging-based stage I pressure ulcer detection system for use by support staff in assisted living and skilled nursing facilities to increase the ulcer detection rate over a wide range of skin colors. This paper describes an image enhancement procedure that improves the detection of pressure ulcers when applied to the color images of ulcer sites. Preliminary results clearly indicate that the enhanced images exhibit higher contrast and make the pressure ulcer site more conspicuous to the examiner. The experiments show promising results even for subjects with black and dark brown skin colors

This paper describes a study designed to assess the impacts of passive health status monitoring technology in home health. Monitoring systems were installed in the homes of 13 home health clients to track physiological parameters (heart rate, breathing rate, and gait), the activities of daily living (ADLs) and key alert conditions of residents, such as falls. Activity reports and alert notifications were sent to professional caregivers in order to refine and target the care administered to clients participating in the study. Informal caregivers of participants were provided with access to the ongoing wellness status of their loved ones. The potential diagnostic utility of the monitoring data, the subjects' quality of life and health related quality of life, as well as the quality of life, strain and burden levels of the informal caregivers were assessed. Pre- and post-installation scores were compared. The results suggest that monitoring technologies could provide care coordination tools that have a positive impact on the perceived quality of life of monitored individuals, as well as a reduction in the strain levels of their informal caregivers, and may have a positive impact on the participants' health related quality of life

This paper describes a study designed to assess some psychosocial impacts of monitoring technology on seniors living in independent senior housing. monitoring systems were installed, in 25 independent living units in an apartment complex, to track the activities of daily living (ADLs) and key alert conditions of residents. Activity reports were sent to informal caregivers. Residents (N=25) were assessed using the satisfaction with life scales (SWLS) instrument, informal caregivers (N=26) were assessed using modified caregiver strain index (CSI) and caregiver burden interview (CBI) instruments, before and after the installation of the monitoring system. Paired t-test for means was applied to the pre- and post-monitoring scores of SWLS, CSI, and CBI. The Wilcoxon matched-pairs signed-ranks nonparametric test was applied to compare the number of informal care hours pre- and post-monitoring. No statistically significant increase was observed on SWLS results. No significant changes in CSI and CBI scores were detected. There was a statistically significant increase in the number of informal care hours provided by the informal caregivers of monitored individuals. The results indicate that monitoring technologies could have enabled informal caregivers to provide more care for their loved ones without increasing their burdens, strain levels or negatively affecting their quality of life

The development of technologies for monitoring the health status of older adults in their living settings is a long and expensive process that involves multiple stakeholders. If the technology is to be mass produced, widely deployed and utilized by different user groups, the technology has to meet certain feasibility criteria, including being acceptable, useful, and potentially beneficial to all the different users. This paper describes the development, content and content validity evaluation results of three custom survey instruments designed to assess the feasibility of using in-home health status, monitoring technologies. The instruments were designed to solicit input from three primary stakeholders in the care process: the monitoring candidate older adults, professional caregivers, and informal caregiver. The validity of the instruments content was evaluated by ten field experts who were asked to score each question in each of the three survey instruments on a 4-point Likert scale on relevance, clarity, and simplicity. All three survey instruments received significantly high overall mean content validity scores. The content validity evaluation results indicated that these instruments are ready to be used to solicit users' requirements from the user groups to guide the refinement of monitoring technologies

This paper explores the validity of a rule-based inference method of selected independent activities of daily living (ADLs). An inexpensive ADL monitoring system was installed in the community for 37 days to monitor a middle-aged, healthy individual living alone. The subject was given a personal digital assistant (PDA), running custom activity diary software, and asked to record activities in real-time. Rule-based activity inference algorithms were refined on data from 17 days, and data from the remaining 20 days were used for validation. The chisquare statistic was computed for 2 x 2 contingency tables comparing activities detected by the algorithms to user-logged activities. The phi (r()) and Cohen's kappa (kappa) coefficients were computed as measures of correlation. After correcting for subject noncompliance in logging activities, the kappa correlation between the meal detection algorithm and the PDA record was 0.84, with 91% sensitivity, and 100% specificity. Similarly, the kappa correlation between the shower detection algorithm and the PDA record is 0.69, with 67% sensitivity and 100% specificity. The detection algorithms and the sensory data did not miss any main meals or showering activities recorded on the PDA. The results suggest that rule-based algorithms can successfully detect meal preparation and showering activities using simple low-cost detectors. The sensors and detection algorithms reported events not recorded by the occupant on the PDA attributed to reporting noncompliance. Overall, the PDA activity journal was a compromise between paper diaries, which are more time consuming to keep, and may result in higher noncompliance errors, and video recording, which is considered intrusive.

In this paper, we present a rule-based approach to the inference of elders’ activity in two primary application areas: detecting Independent Activities of Daily Living (IADLs) for the detection of anomalies in activity data patterns consistent with arising health issues over a period of time, and the detection of possible emergency conditions passively and unobtrusively. We discuss our efforts using classification techniques leading to the rule-based inference approach, and compare results between the two approaches. The results have shown the viability and validity of knowledge-engineered rules, which outperformed automatically generated rules using random forest supervised learning; the κ correlation coefficient between the classification results of the random forest model and the PDA record was 0.79 , with 85% sensitivity and 93% specificity , compared to κ= 0.84 , with 91% sensitivity and 100% specificity for the knowledge engineered rule aimed at the detection of main meal preparation. The paper also presents experimental field trial results of the rule-based approach demonstrating the utility of the method and future directions for our research.

INTRODUCTION This abstract describes a passive unobtrusive gait-monitoring device, based on a highly sensitive optic fiber floor vibration sensor. Long-term in-home gait monitoring provides a measure of a person's functional ability and activity levels, which can help 'evaluate' a person's health over a long period of time. Such a passive gait-monitoring device can be useful for assessing healing/deterioration following therapeutic interventions. It can also contribute to the detection of general health problems early. The ability to distinguish between normal walking and limping or shuffling, which may be precursors to a fall, as well as detecting falls, is of utmost value to elder populations. Elders, who represent 12% of the population, account for 7 5% of deaths from falls [1]. The considerable cost involved in the treatment and hospitalization of fall injuries and even death due to falls could be greatly reduced if falls could be predicted and avoided through appropriate intervention. BACKGROUND Current gait analysis techniques broadly fall under three categories depending upon the type of device used: wearable devices, walk on devices and visual gait analysis tools and techniques. These gait laboratory equipment and analysis techniques yield excellent and detailed gait characteristics and enable clinicians to prescribe an appropriate intervention. However, the equipment required is extremely expensive, in the range of tens of thousands to a few hundred thousand dollars. The computational power required for the image based analysis make longitudinal in-home gait monitoring using these technologies impractical.

To provide biological specimens for scientific studies, the Medical Automation Research Center (MARC) designed and constructed a large-scale device that emphasizes the use of robotics and automation to integrate many associated laboratory operations. These included analysis, dilution, archival storage, and retrieval of purified human-derived specimens. Designers of automated biological repositories are challenged by complex engineering problems. In this paper, we present an overview of the biological repository (biorepository) and give details of the software architecture.

INTRODUCTION To date, there are few systems that provide a low-cost, passive way of acquiring important sleep monitoring data that requires no additional action from the subject outside of their normal daily routine [1]. There is, however, a great need for research in this area because of the large number of people affected by sleep related conditions who could benefit from knowing more about their sleep habits. About 40% of all American adults suffer from some kind of sleep disorder while about 70 million Americans are chronically sleep deprived [2]. Many feel that little substantial improvement can be made to correct their problems since 70% of sleep sufferers don't discuss the problem with their physician [2]. Objective sleep research has existed since 1922 when Szymansky ran the first such study [3]. The current gold standard for sleep research is polysomnography (PSG), which involves at least the recording of an electroencephalogram (EEG), a measurement of brain waves, an electrooculogram (EOG), a measurement of muscle activity in the eye area, and an electromyogram (EMG), a measurement of muscle activity in specific areas such as the arm or leg [4]. These electrode hookups prove valuable to assess sleep quality, but their attachment to the patient's body affects sleep. In an effort to provide a less obtrusive way to study sleep on a longer-term basis, actigraphs have been developed. These devices can be attached to any of the limbs to provide movement data based on the same principles behind accelerometers. They are also used in activity studies [5] and can provide 24 hour monitoring of the subject. This type of sensor, however, has its limitations in acquiring data that can be interpreted definitively to provide a good assessment of sleep quality. Researchers are dependent on patient journals to help correlate the data recorded on the actigraph and it is hard to distinguish different events that can occur throughout the night [3]. In addition, problems researchers have interpreting results from actigraphs are a direct result of the one-dimensional nature of the data recorded [4]. It is believed that valid sleep assessments can be made through the analysis of physiological characteristics such as body temperature, sleeping position and movement, breathing rate and heart rate. This