Multisensor data fusion algorithm for sleep quality estimation using multiple measurements
Résumé
Objectives/Introduction: Sleep monitoring represents an interest field in medical word. We spend a part of our life in bed, then an analyzing of our sleep represents a part of our medical consultation. Therefore, the supervision of the evolution of our sleep quality is primordial thing. A lot of systems have been implemented, tested and commercialized. Each system has a specified properties and characteristics. In our living lab, We study the correlation between the sleep quality and balance stability measured at wake up. In this study we use the withing system which is based on movement counted when the person was in the bed at night in order to measure sleep quality and for the balance stability is measured by scale which uses the movement counted when the person is on the scale. The balance stability is used to enhance the sleep quality score therefore the data fusion algorithms are tested on recorded database. The database is recorded by students in internship and phd students. In this work, the data base is described, the correlation study is presented and the fusion data algorithms are detailed. Estimation methods of the weight for each modality (scale and withing system) are prospected.