Project Description

This research creates a meaningful, manageable information resource that enables more complete and accurate assessment, diagnosis, treatment, and evaluation of behavioral problems for the elderly. Through activity and environmental monitoring in a skilled nursing facility, a continuous, voluminous audio and video record is captured. Through work in information extraction, behavior analysis and synthesis, this record is transformed into an information asset whose efficient, secure presentation empowers geriatric care specialists with greater insights into problems, effectiveness of treatments, and determination of environmental and social influences on patient behavior

The shortage of geriatric care professionals, the growth of the elderly population, and the societal benefits of improving quality of life and care in skilled nursing facilities underscore the need for CareMedia: automated video and sensor analysis for geriatric care. This research is validated with prototypes and trials in nursing homes and clinical settings, through project affiliate Western Psychiatric Institute and Clinic, guided by the direction of the medical professionals participating in the project and serving on the Advisory Board. The work focuses on senile dementia patients in a specialized care facility, where close monitoring is an objective toward providing better care.

Studies indicate that nearly 90% of patients with dementia may exhibit measurable agitation, behavior that can be classified as disturbed (a psychiatric or medical condition requiring pharmacological intervention) or disturbing (socially inappropriate behavior that may just be a means of expressing a need). In the absence of close monitoring, the caregiver cannot interpret behavior appropriately or assess why particular behaviors occur. The impact of the technologies developed by this research is to enable the geriatric care specialist to more accurately address situations and intervene appropriately, by creating continuous measurement instruments that extract the required data and highlight behaviors of interest. The research begins with disruptive vocalization, a particular behavior noted across senile dementia assessment scales. The coverage is then broadened ambitiously to integrate sensor and visual data for behavioral analysis and summarization in support of OBRA regulations requiring behavior management strategies that are not just chemical restraints. Specifically, this effort:

• Develops automatic techniques to identify and classify disruptive vocalizations according to standardized measures.
  
• Automatically analyzes aural, visual, and sensor information to recognize more complex behavioral occurrences, such as falls, wandering, physical aggression, and circadian patterns of activity.
  
• Provides medical staff with filtered audiovisual evidence from which more accurate behavioral logs can be produced, logs which are required by law and used in diagnosis, treatment, and evaluation.
  
• Establishes the reliability of the Pittsburgh Agitation Scale, an instrument for rating agitation in dementia patients, in skilled nursing facilities.
  
• Focuses on the efficient use of a detailed multimedia record, supporting intelligent browsing and filtering by collapsing redundancy and dealing with automatically produced content descriptors.
  

Data for diagnosis and clinical studies are now typically gathered by hand during a few brief periods per week. More detailed, exhaustive behavioral assessment scales have been developed, but have the drawback of requiring many observations and hence being too time-consuming for regular use by the clinical staff. The information technology developed in this work provides geriatric care specialists with a better window into the lives of senile dementia patients. Their behavior can then be more accurately interpreted, leading to treatment that reduces agitation while allowing awareness and responsiveness. This research builds on key Carnegie Mellon research efforts in the areas of automated capture and analysis of digital video and sensor data, wearable mobile computers, computer-based vision systems, and information retrieval systems for multimedia metadata.