6533b86cfe1ef96bd12c8852

RESEARCH PRODUCT

Nouvelle approche pour l'estimation du rythme respiratoire basée sur la photopléthysmographie sans contact

subject

description

Respiratory rhythm is important information in medical context.Its assessment allows to predict some medical complications that could lead to death.However, it is often neglected by the medical staff due to a bad comprehension of its importance, or a lack of time.Automated measurement methods allow to improve this by continuously giving respiratory rate.Most of these methods needs a contact with the patient to efficiently measure the breathing rate.Unfortunately it leads to some issues which could forbid measurement or make it unconfortable for continuous monitoring.The continuous, every-day monitoring especially needs to be as discrete as possible to be forgotten by the patient.To deal with these drawbacks, several non contact respiratory rate assessment methods are currently developped.In these methods, the remote photoplethysmography uses the color variation of the skin due to blood volume in capillaries to obtain a cardiac and respiratory correlated signal.In this thesis, we focused on improving the respiratory rate measurement using remote photoplethysmography.To do this, we developped some methods whose goals are to efficiently combine color signals that were extracted from the video to obtain a single signal that maximizes the breathing information.In a second part, we developped a processing pipeline to assess the breathing rate from all informations that could be extracted from remote photoplethysmography signals.The processing pipeline was tested with state of the art combination methods and the methods that were developped during the thesis.