With the increasing share of elderly population worldwide, the need for assistive technologies to support clinicians in monitoring their health conditions is becoming more and more relevant. As a quantitative tool, geriatricians recently proposed the notion of frail elderly, which rapidly became a key element of clinical practices for the estimation of well-being in aging population. The evaluation of frailty is commonly based on self-reported outcomes and occasional physicians evaluations, and may therefore contain biased results. Another important aspect in the elderly population is hospitalization as a risk factor for patient’s well being and public costs. Hospitalization is the main cause of functional decline, especially in older adults. The reduction of hospitalization time may allow an improvement of elderly health conditions and a reduction of hospital costs. Furthermore, a gradual transition from a hospital environment to a home-like one, can contribute to the weaning of the patient from a condition of hospitalization to a condition of discharge to his home. The advent of new technologies allows for the design and implementation of smart environments to monitor elderly health status and activities, fulfilling all the requirements of health and safety of the patients. From these starting points, in this thesis I present data-driven methodologies to automatically evaluate one of the main aspects contributing to the frailty estimation, i.e., the motility of the subject. First I will describe a model of protected discharge facility, realized in collaboration and within the E.O. Ospedali Galliera (Genoa, Italy), where patients can be monitored by a system of sensors while physicians and nurses have the opportunity to monitor them remotely. This sensorised facility is being developed to assist elderly users after they have been dismissed from the hospital and before they are ready to go back home, with the perspective of coaching them towards a healthy lifestyle. The facility is equipped with a variety of sensors (vision, depth, ambient and wearable sensors and medical devices), but in my thesis I primarily focus on RGB-D sensors and present visual computing tools to automatically estimate motility features. I provide an extensive system assessment I carried out onthree different experimental sessions with help of young as well as healthy aging volunteers. The results I present are in agreement with the assessment manually performed by physicians, showing the potential capability of my approach to complement current protocols of evaluation.
|Titolo della tesi:||A protected discharge facility for the elderly: design and validation of a working proof-of-concept|
|Data di discussione:||2-mag-2019|
|Appare nelle tipologie:||Tesi di dottorato|