VitaFALL: NXTGeUH System for Well-Being Monitoring with Fall Recognition and Real-Time Vital Sign Monitoring

Abstract

The blend of computerized data processing with the existing engineering and medic techniques has enabled explorers in the betterment of controlling of patients concerning the two at homes along with at clinics. In this work, numerous fall assessment for fall prediction and detection with vital signs monitoring techniques and methods particularly to establish a research gap and its allied research problems has been reviewed and incorporated using a triple-axis accelerometer and Vital Signs Parameters (Heartrate, Heartbeat, and Temperature monitoring) for the ancient people with a Internet of Medical Things based Vital Signs and Fall Detection (VitaFALL) is proposed which is well-timed and gives an effective judgment of the fall. The four layers comprise sensing, network, data processing and application layer. A caretaker and doctor can be notified by sending alert using a GSM and GPRS module in order that elder can be helped on time, however, a delay in the time is noticed when comparing the gradient and minimum value to predetermine the state of the old person. From a few decades, vital signs have been important parameters to find out the patient’s health level. Vital signs estimation has always been the initial step for the evaluation of the patient and this is also possible by checking the pulse rate or checking the palpation of their forehead for high temperature. ADXL335 Three-Axis Accelerometer Module, tri-axial 14-bit ± 8g accelerometer collects motion information in the VitaFall device. The basic idea is to avoid falls and not to detect them after the loss is done. Walking, stumbling, sitting, falling (right, forward, backward and left) and all other normal motion data patters in the daily life of an older adult (who did no longer have any records or walking issues) are collected. The proposed VitaFall Fall detection model has achieved 85% accuracy, specificity of 100%, and sensitivity of 96% when detecting directional falls. The model uses motion data, real-time vital signs values, falls history to foresee the lows, medians and the highs of falls risks in hospitalized elderly people. When compared with the manual falls risk tools known as the Morse Falls scale, the system got an accuracy of 85%, predictability of 100%, and a sensitivity of 100% too.