4. Conclusion
In this study, in an effort to facilitate the monitoring of patient bedside activities and prevent accidents, we report
on the development of a system that can not only measure a patient’s head location in a bedside environment, but
also his or her location within a nursing home.
The bedside location system is implemented by measuring the patient’s head location using a Kinect sensor using
two sub-functions. One sub-function monitors the space in the vicinity of the bed using a RANSAC-based
recognition algorithm. The other function detects a patient’s head using an AdaBoost-based head recognition
algorithm. To confirm the effectiveness of this system, we created an experimental bedside environment within our
laboratory.
The indoor location sensing system determines a patient’s indoor location within the nursing home using RFID
communication. To grasp a patient’s real-time indoor location seamlessly, low-cost multiple RFID readers are
installed at various locations in the nursing home and RFID tags are embedding in the patient’s shoes. To confirm
the effectiveness of the developed system, it was installed in an actual Tokyo nursing home and the activities of
patient noted for wandering behavior were monitored over a one-week period. Using the data extracted during our
test, we could compare the patient’s behavior spatial movement patterns to determine if they fit those expected by a
wandering patient.