Video-Based Analysis of Heart Rate Applied to Falls

Falls are the number one cause of unintentional injuries and among the top-ten causes of deaths for older adults. An improved understanding of the causes of falls may lead to better approaches to prevention. Changes in blood pressure and heart rate can contribute to falls. Our goal is to develop tec...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2018 IEEE International Symposium on Medical Measurements and Applications (MeMeA) S. 1 - 5
Hauptverfasser: He, Xiaochuan, Goubran, Rafik, Bennett, Stephanie, Robinovitch, Stephen, Symes, Bobbi, Lo, Bryan, Ejupi, Andreas, Knoefel, Frank
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.06.2018
Schlagworte:
Cameras
fall analysis
Heart rate
Object tracking
Pulse measurements
remote patient monitoring
Sociology
Statistics
Streaming media
video analysis
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Falls are the number one cause of unintentional injuries and among the top-ten causes of deaths for older adults. An improved understanding of the causes of falls may lead to better approaches to prevention. Changes in blood pressure and heart rate can contribute to falls. Our goal is to develop techniques for measuring these parameters during real-life falls. In this paper, we apply advanced video processing techniques to footage of real-life falls in long-term care facilities to measure the heart rate of individuals before, during and after falls. We use an object-tracking algorithm to track a region of interest (ROI), such as the face, neck, hands or legs. We then estimate the heart rate based on subtle color changes in the ROI, which correlate with arterial blood flow. To verify the performance of the videobased method an experiment was conducted where the subject was wearing a pulse oximeter to measure the heart rate while being videotaped. Our results demonstrate that the proposed video-based processing method can be used to accurately capture heart rate variations during a fall. We also evaluate how the accuracy of the heart rate estimation depends on the size of the selected ROI. The proposed system could be used for remote patient monitoring and health informatics for improved detection and identification of the causes of falls.
DOI:10.1109/MeMeA.2018.8438773