Evaluation of Accelerometer-Based Fall Detection Algorithms on Real-World Falls

Abstract

Despite extensive preventive efforts, falls continue to be a major source of morbidity and mortality among elderly. Real-time detection of falls and their urgent communication to a telecare center may enable rapid medical assistance, thus increasing the sense of security of the elderly and reducing some of the negative consequences of falls. Many different approaches have been explored to automatically detect a fall using inertial sensors. Although previously published algorithms report high sensitivity (SE) and high specificity (SP), they have usually been tested on simulated falls performed by healthy volunteers. We recently collected acceleration data during a number of real-world falls among a patient population with a high-fall-risk as part of the SensAction-AAL European project. The aim of the present study is to benchmark the performance of thirteen published fall-detection algorithms when they are applied to the database of 29 real-world falls. To the best of our knowledge, this is the first systematic comparison of fall detection algorithms tested on real-world falls. We found that the SP average of the thirteen algorithms, was (mean±std) 83.0%±30.3% (maximum value = 98%). The SE was considerably lower (SE = 57.0%±27.3%, maximum value = 82.8%), much lower than the values obtained on simulated falls. The number of false alarms generated by the algorithms during 1-day monitoring of three representative fallers ranged from 3 to 85. The factors that affect the performance of the published algorithms, when they are applied to the real-world falls, are also discussed. These findings indicate the importance of testing fall-detection algorithms in real-life conditions in order to produce more effective automated alarm systems with higher acceptance. Further, the present results support the idea that a large, shared real-world fall database could, potentially, provide an enhanced understanding of the fall process and the information needed to design and evaluate a high-performance fall detector.

A comparison of public datasets for acceleration-based fall detection

“…However, it could also be explained by the fact that tFall has a large variety of movements acquired in a real environment. Bagalà et al [19] also pointed out a decrease in performance when testing algorithms in real circumstances.…”

“…In this regard, several authors have identified the need for having public datasets [19,20]. Some efforts have been performed in this direction since several datasets were made publicly available in the recent years: DLR [21] published in 2011, MobiFall [22] available in 2013 and tFall [20] uploaded in 2014 (the study of Fudickar et al [23] cites another public dataset but it seems that it cannot be downloaded currently).…”

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

“…The lack of more and various types of real fallings restricts the ecologic validity of our result of 100% sensitivity. An open accelerometry database including real world-falls would be of considerable importance for a continuous and independent refinement and validation of fall detection and prevention algorithms -see also [8].…”

“…This unfortunate accident provided us with an extremely valuable data for two main reasons: 1) the difficulty that in itself entails to get a real fall. As an example, to capture 100 real-world falls it would be necessary to record approximately 100,000 days of physical activity (300 person years) [8]. And 2), the detailed documentation about the real-world fall and previous weeks; in addition to the selfreported information, we have 24-hour actibelt R recordings and clinical functional test measurements (Timed Up and Go test, 10 meter test and 6 minute walking test) from the two weeks prior to the accident.…”

Mobile activity recognition and fall detection system for elderly people using Ameva algorithm

“…This characterization is because the training of the system under these circumstances is not possible. Although there are studies that identify these events dynamically [45], the learning process is usually performed on a synthetic dataset, due to the great handicap of generating data of real falls. Therefore, in this paper the problem of falling is addressed in parallel, defining an accelerometric profile criteria and a period of no accelerations.…”