A smart inertial system for fall detection

The high incidence of falls in elderly and people with neurological disease, represents a serious issue which can have tremendous consequences. To reduce the effect on the society of this phenomenon, scientist in every part of the world, have tried to develop solutions ranging from gait training to more technological approaches such as the assistive device. Although it is evident that the scientific community has developed different and valuable paradigms for the ADLs and fall detection issue, in selecting one of the many possible solution, one has to consider the different advantages and disadvantages in terms of performance, user acceptance and power requirements. In order to meet the necessity to develop solutions and paradigms that can be suitable for battery-powered operations, with specific focus to low power embedded systems, in this paper a Fall and ADLs classification paradigm exploiting an event correlated approach is proposed. It is based on the maximum value of the correlation between two signals implemented in an ad-hoc developed embedded system based on an STmicroelectronics microcontroller equipped with sensors and communication facilities. Results confirm the suitability of the presented methodology showing an average sensitivity (S-e) of 0.97 and an average specificity (S-p) of 0.97.

Automated summary

Scientists worldwide have been developing solutions to address the high incidence of falls in elderly and people with neurological diseases, ranging from traditional gait training to more technological approaches. When choosing among the different solutions available, it is important to consider factors such as performance, user acceptance, and power requirements.