A new method to estimate the real upper limit of the false alarm rate in a 3 accelerometry-based fall detector for the elderly

Abstract

Abstract-Falls are a major concern for the elderly and their ability to remain healthy. Fall detection systems may notify emergency responders when no one apart from the injured is present. However, their real-world application is limited by a number of factors such as high false positive rates, lowcompliance, poor-usability and short battery lifetime. In order to improve these aspects we have developed a miniaturized 3D accelerometer integrated in a belt buckle, the actibelt R , and a fall detection algorithm. We have used a new evaluation method to assess the upper limit of the false alarm rate of our algorithm using a large set of long term standardized acceleration measurements recorded under real life conditions. Our algorithm has a false alarm rate of seventeen false alarms per month and has the potential to be reduced down to at most three false alarms per month when activities which require the sensor to be removed are eliminated. In laboratory settings, the algorithm has a sensitivity of 100%. The algorithm was sucessfully validated using data from a real-world fall.

Methods for the Real-World Evaluation of Fall Detection Technology: A Scoping Review

“…The level of detail provided about participants varied considerably. All but three [ 31 , 38 , 40 ] of the articles stated whether participants were community dwelling, in long-term care or hospital patients. Five articles did not provide any additional descriptive information on the participants [ 23 , 24 , 35 , 37 , 40 ].…”

“…Lipsitz et al [ 34 ] provided the most in-depth description with a breakdown of the proportion of participants with a range of 21 comorbidities. Eight articles reported results of mobility assessments [ 21 , 27 , 28 , 29 , 30 , 31 , 33 , 38 ], three articles provided information on walking aid use [ 20 , 27 , 28 ] and three articles additionally reported results of cognitive assessments [ 29 , 30 , 33 ]. None of the other 15 articles reported standardised measures of cognitive or mobility status.…”

“…Fifteen studies used wearable technology and ten used non-wearable, Table 2 shows full details of the devices used in each study. Accelerometers are the most common choice of sensor and have been used in 15 of the studies [ 17 , 21 , 22 , 23 , 24 , 25 , 29 , 30 , 31 , 32 , 33 , 34 , 36 , 38 , 40 ]. Eight studies tested some form of optical sensor [ 20 , 22 , 26 , 27 , 28 , 30 , 37 , 39 ], making them the most common choice of non-wearable devices.…”

Chameleon: personalised and adaptive fall detection of elderly people in home-based environments

“…Acceleration magnitude and angle are analysed as two parameters to detect fall in Brown (2005), and Hansen et al (2005). In Soaz et al (2012), authors focus on large acceleration produced by the body movement when a fall occurs to determine a possible impact. Once such a large impact is observed, the orientation of the sensor is calculated to differentiate between standing/sitting and lying, while the sum of the windowed standard deviation is cited to further determine the final fall status.…”

Real-time energy-efficient fall detection based on SSR energy efficiency strategy

“…However, such local in-hardware solutions also consume certain energy (Hu et al, 2011), which is not studied in their research. Compared with the server side computing based algorithms, Soaz et al (2012), Brown (2005), Chen et al (2005) and Purwar et al (2007) investigate accelerations of three orientations by a tri-accelerometer, combining with various parameters such as angle, time interval or others to identify the fall at the sensor node, and they only transmit a fall alarm when the fall is detected. Such local computing based algorithms consume less power by reducing wireless transmission.…”