Image processing algorithms for patient mobility monitoring using pressure sensitive bed and floor arrays

This thesis focuses on the problem of unobtrusive patient monitoring using pressure sensor arrays to help assess the mobility of older adults. Using under mattress pressure sensor arrays data is captured continuously from occupants. The raw data is processed using specialized algorithms and important clinical information is extracted. Key clinical features including bed entry time, sleep duration, number of exits, and bed exit time (lie-to-sit and sit-to-stand times) can be extracted. These features can be used by clinicians to assist in assessing the patient’s mobility. A floor tile based pressure sensor array is used to capture data while the patient stands still on it and shifts their weight. The data is processed and clinical information is extracted related to the patient’s static and dynamic balance. Key clinical features include weight balancing, center of pressure under the feet, and amount of sway, assist in assessing the patient’s standing stability and static balance. Average and maximum movement velocity measures can assist in assessing the patient’s dynamic balance. The thesis shows that pressure sensor arrays can be used to collect data unobtrusively and using specialized image processing algorithms, clinically relevant information can be extracted which can assist in assessing the mobility and stability of older adults.