The identification of vertical velocity profiles using an inertial sensor to investigate pre-impact detection of falls

doi:10.1016/j.medengphy.2007.12.003

Medical Engineering & Physics

Volume 30, Issue 7, September 2008, Pages 937-946

https://doi.org/10.1016/j.medengphy.2007.12.003 Get rights and content

Abstract

This study investigates distinguishing falls from normal Activities of Daily Living (ADL) by thresholding of the vertical velocity of the trunk. Also presented is the design and evaluation of a wearable inertial sensor, capable of accurately measuring these vertical velocity profiles, thus providing an alternative to optical motion capture systems.

Five young healthy subjects performed a number of simulated falls and normal ADL and their trunk vertical velocities were measured by both the optical motion capture system and the inertial sensor.

Through vertical velocity thresholding (VVT) of the trunk, obtained from the optical motion capture system, at −1.3 m/s, falls can be distinguished from normal ADL, with 100% accuracy and with an average of 323 ms prior to trunk impact and 140 ms prior to knee impact, in this subject group.

The vertical velocity profiles obtained using the inertial sensor, were then compared to those obtained using the optical motion capture system. The signals from the inertial sensor were combined to produce vertical velocity profiles using rotational mathematics and integration.

Results show high mean correlation (0.941: Coefficient of Multiple Correlations) and low mean percentage error (6.74%) between the signals generated from the inertial sensor to those from the optical motion capture system. The proposed system enables vertical velocity profiles to be measured from elderly subjects in a home environment where as this has previously been impractical.

Introduction

Falls in older people and the injuries that they sustain, are a major problem for their welfare, confidence and happiness and greatly contribute to their health costs, morbidity and mortality [1], [2]. The injuries that may occur to an older person as part of a fall can result in long-term hospitalisation which may result in a loss of independence for a faller [2].

If an older person, living alone, experiences a fall, he/she may be unable to get to a phone due to injuries sustained and could thus remain on the ground undetected for a long period. An incident such as this could have a potentially fatal outcome [3]. At the very least the elderly faller will have an increased fear-of-falling [4].

A solution to this is the automatic detection of a fall. A number of subject worn fall-detection systems and algorithms currently exist [5], [6], [7], [8], [9], [10], [11]. They detect the impact received by the body to determine if a fall has occurred. Even though 100% fall-detection accuracies have been achieved in testing using both accelerometers [6] and gyroscopes [9], this method of fall-detection has the limitation that the fall-detection system must still be operable subsequent to the impact. Thus, a fall-detection device, capable of detecting a fall prior to impact, is highly desirable. In addition, a sensor of this type could be combined with automotive air-bag technology to reduce the number of hip fractures that occur by inflating a protective air-bag over the hip, providing trochanteric padding. A system such as this could form part of a complete fall-detection and fall injury prevention, health-monitoring home system [12] which would promote more independent living in the elderly community. Patents detailing inflatable hip protectors to cushion the fall prior to impact do exist [13], [14], [15], however they neglect to address how pre-impact detection of the falls is accomplished.

Research has been carried out into the impact dynamics of falls and has identified quantities such as trunk impact velocity, which may be used to anticipate impact from falls. These have been identified using mathematical linked models to simulate a falling person [16] and video analysis of subjects performing simulated falls onto crash mats [17], [18]. Wu [19] compared vertical and horizontal velocity profiles obtained from subjects performing normal activities and simulated forward and backward falls. Wu identified unique characteristics in the velocity profiles that enabled the detection of a fall 300–400 ms prior to impact. However, it has been argued that thresholding of the vertical velocity of the trunk alone, is sufficient for pre-impact fall-detection [20]. We thus propose implementing a subset of Wu's algorithm using data from a body worn inertial sensor, specifically, use of vertical velocity thresholding (VVT) of the trunk, using inertial sensor data to investigate pre-impact detection of falls. The context in which this work is being carried out is the real-time pre-impact detection of falls using an inertial sensor.

Conventionally, velocity characteristics of the body are obtained using laboratory based optical motion capture systems. The subjects’ movements are confined to the capture volume of the optical motion capture system and often these camera set-ups are permanently located in dedicated areas, thus limiting the activities and subjects that can be monitored. In order to allow greater investigation of fall/ADL distinction and pre-impact fall-detection, using velocity characteristics, this paper also details the development and test of a sensor and associated mathematical procedure to produce vertical velocity profiles from a subject worn sensor. This arrangement will allow the investigation of the characteristics of trunk vertical velocity profiles of the elderly while they perform their normal activities in their own homes, without the need for an optical motion capture system to be installed. We thus propose to present an algorithm which uses inertial sensor data, to obtain trunk vertical velocity profiles, which will allow us to investigate the potential, for the pre-impact detection of a fall using this sensor.

In order to determine if VVT of the trunk has the potential for pre-impact fall-detection and to examine how accurate the vertical velocity signals derived from the inertial sensor are, a number of simulated falls and ADL were performed, while the subjects’ movements were recorded using both the inertial sensor and captured using an optical motion capture system. The trunk vertical velocity profiles of subjects performing normal ADL were then compared to those produced during falls. Following this, the vertical velocity profiles obtained from a mathematical transformation of the signals from the inertial sensor were compared by amplitude and closeness in shape to the profiles from the optical motion capture system.

Section snippets

Sensor design

In this study, the radial, tangential and medio-lateral accelerations as well as pitch, yaw and roll angular velocities of the trunk, of each subject were recorded during each activity, using an inertial sensor unit consisting of a tri-axial accelerometer and tri-axial gyroscope.

Calibration

Calibration of the tri-axial accelerometer and gyroscope sensors was performed using the method outlined by Ferraris et al. [21]. To compensate for the non-negligible gyroscope offset drift which occurs (considered as

Results

The maximum and minimum downward vertical velocity, $v_{v, \max}$ , and $v_{v, \min}$ , as well as the mean and standard deviation of the peak values for each fall and ADL, for both the inertial sensor and optical motion capture system, were recorded (Table 1). The maximum downward vertical velocity recorded from all the ADL is used as the threshold when distinguishing falls from normal ADL. The largest $v_{v, \max}$ recorded from all the ADL performed was, −1.286 m/s for the optical motion capture system and, −1.298

Discussion

The authors have shown that thresholding of the vertical velocity of the trunk alone, is sufficient to distinguish falls from normal ADL with 100% sensitivity (and 100% specificity, by design) and that pre-impact detection of falls with an average lead-time of 323 ms (range 155–750 ms) prior to trunk and knee impact can be achieved.

This paper also describes how vertical velocity profiles can be produced using a single inertial sensor; these profiles achieved high CMC values and low RMS error

Conflict of interest

No conflicts of interest exist.

Acknowledgments

The authors wish to acknowledge the assistance of: Analog Devices, BV for providing the ADXL210 accelerometers and ADXRS300 gyroscopes and Dr. John Nelson and the CAALYX FP6 project, caalyx.eu [12], for their financial support during the write up phase of this study. Míle Buíochas.

References (25)

A.K. Bourke et al.
Evaluation of a threshold-based tri-axial accelerometer fall detection algorithm
Gait Posture
(2007)
M.N. Nyan et al.
Distinguishing fall activities from normal activities by angular rate characteristics and high-speed camera characterization
Med Eng Phys
(2006)
A.J. van den Kroonenberg et al.
Hip impact velocities and body configurations for voluntary falls from standing height
J Biomech
(1996)
G. Wu
Distinguishing fall activities from normal activities by velocity characteristics
J Biomech
(2000)
J.C. Lötters et al.
Procedure for in-use calibration of triaxial accelerometers in medical applications
Sens Actuators A: Phys
(1998)
C. Smeesters et al.
Disturbance type and gait speed affect fall direction and impact location
J Biomech
(2001)
C.J. Lord et al.
Falls in the elderly: detection and assessment
P. Haentjens et al.
The Economic cost of hip fractures among elderly women
J Bone Joint Surg Am
(2001)
R.J. Gurley et al.
Persons found in their homes helpless or dead.
N Engl J Med
(1996)
Doughty K. Fall prevention and Management strategies based on intelligent detection monitoring and assessment. New...

J.Y. Hwang et al.

Development of novel algorithm and real-time monitoring ambulatory system using Bluetooth module for fall detection in the elderly

Conf Proc IEEE Eng Med Biol Soc

(2004)

K. Doughty et al.

The design of a practical and reliable fall detector for community and institutional telecare

J Telemed Telecare

(2000)

Cited by (125)

Human-centred assembly and disassembly systems: a survey on technologies, ergonomic, productivity and optimisation
2022, IFAC-PapersOnLine
Despite the increasing use of automation in the current industry 4.0 context, manual assembly and disassembly tasks are still common and in some situations even unavoidable. The interaction between humans and other elements of the Assembly and Disassembly Systems (ADS) has been discussed in the scientific literature with the dual objective of optimising human well-being and system performance. Human Factor (HF) related studies focused on new technologies such as motion tracking systems. These technologies enable the collection of data to improve understanding the work environment and its impact on employee well-being and productivity. In the literature, different metrics were suggested to measure the ergonomic and productivity scores. These measurements influence the global system performance and may allow its optimisation. In this paper, we provide a review of the literature on human-centred ADS. We mainly focus on technologies (used to capture human motion), metrics (used to assess human ergonomic risks and productivity), and operational research models (used to optimise the performance system considering economic and cycle time objectives). Future directions are discussed in the perspectives of this paper.
The influence of collision avoidance strategies on human-robot collaborative systems
2022, IFAC-PapersOnLine
Collaborative robots improve the traditional production systems, both automatic and manual, by allowing the resources to share the workspace. However, the shared workspace is both the greatest advantage and limit of collaborative robots. Indeed, interference between the human operator and the cobot may result in an emergency stop, reducing the performance of the system. Hence, a collision avoidance strategy is required to avoid collision; however, this may affect the performance of the system. Therefore, it is necessary to investigate how the shared workspace affects the system performance, and the effects of the adoption of a collision avoidance strategy. To this regard, a collision avoidance strategy has been developed and it is presented in this work and an experimental validation proved the efficacy of the proposed strategy. Lastly, simulation tests were carried out to analyse the performance of the system for different sizes of the collaboration area with respect to the workspace. The results shows that, while the increase in the collaboration area reduces the performance of the system, collision avoidance strategies mitigate this effect.
A novel gateway-based solution for remote elderly monitoring
2020, Journal of Biomedical Informatics
Internet of Things (IoT) technologies have been applied to various fields such as manufacturing, automobile industry and healthcare. IoT-based healthcare has a significant impact on real-time remote monitoring of patients’ health and consequently improving treatments and reducing healthcare costs. In fact, IoT has made healthcare more reliable, efficient, and accessible. Two major drawbacks which IoT suffers from can be expressed as: first, the limited battery capacity of the sensors is quickly depleted due to the continuous stream of data; second, the dependence of the system on the cloud for computations and processing causes latency in data transmission which is not accepted in real-time monitoring applications.
This research is conducted to develop a real-time, secure, and energy-efficient platform which provides a solution for reducing computation load on the cloud and diminishing data transmission delay. In the proposed platform, the sensors utilize a state-of-the-art power saving technique known as Compressive Sensing (CS). CS allows sensors to retrieve the sensed data using fewer measurements by sending a compressed signal. In this framework, the signal reconstruction and processing are computed locally on a Heterogeneous Multicore Platform (HMP) device to decrease the dependency on the cloud. In addition, a framework has been implemented to control the system, set different parameters, display the data as well as send live notifications to medical experts through the cloud in order to alert them of any eventual hazardous event or abnormality and allow quick interventions. Finally, a case study of the system is presented demonstrating the acquisition and monitoring of the data for a given subject in real-time. The obtained results reveal that the proposed solution reduces 15.4% of energy consumption in sensors, that makes this prototype a good candidate for IoT employment in healthcare.
Motion Analysis System (MAS) for production and ergonomics assessment in the manufacturing processes
2020, Computers and Industrial Engineering
Nowadays the Smart Factories operating within the Industry 4.0 revolution, require more and more reliable, fast and automatic tools for production analysis and improvement. Manufacturing companies, in which the human labour has a crucial role, need instruments able to manage complex production systems in terms of resource utilization, product mix, component allocation and material handling optimization.
In this context, this work presents an original hardware/software architecture, Motion Analysis System (MAS), aimed at the human body digitalization and analysis during the execution of manufacturing/assembly tasks within the common industrial workstation. MAS is based on the integration of the Motion Capture (MOCAP) technology with an ad hoc software developed for productive and ergonomic analysis of the operator during his work. MAS hardware integrates a network of depth cameras initially developed for gaming (Microsoft Kinect v2™, conceived for markerless MOCAP) and now used for industrial analysis, while an original software infrastructure is programmed to automatically and quantitatively provide productive information (human task analysis in terms of time execution and used space within the workplace, movements of hands and locations visited by the operator) and ergonomic information (full body analysis implementing all the internationally adopted indexes OWAS, REBA, NIOSH and EAWS). This double perspective makes MAS a unique and valuable tool for industrial managers oriented to the workplace analysis and design (in terms of productivity) without neglecting the operator health. This proposed contribution ends with a real industrial application analysing a water pump assembly station: the system setup is discussed and the key results obtained adopting MAS are presented and analysed.
Practical fall detection based on IoT technologies: A survey
2019, Internet of Things (Netherlands)
Fall is the second incident that leads to death over the world. Fall event happens to numerous groups of people consist of elderly, babies and also younger people. Admittedly, fall-related research must be considered as one of the most important aspects of a healthy lifestyle. For this reason, Internet of Things(IoT) is the emerging technology and a powerful candidate to develop fall diagnosis system. In this paper, we have discussed the three stages of fall as Prediction, Prevention, and Detection. We have illustrated Edge, Fog, and Cloud layers as IoT layers to develop a fall diagnosis system. At the end of the paper, we have considered the challenges of fall diagnosis systems and suggested future aspects.
40th Anniversary Issue: Reflections on papers from the archive on “Biomechanics”
2019, Medical Engineering and Physics

View all citing articles on Scopus

View full text

The identification of vertical velocity profiles using an inertial sensor to investigate pre-impact detection of falls

Abstract

Introduction

Section snippets

Sensor design

Calibration

Results

Discussion

Conflict of interest

Acknowledgments

Gait Posture

Med Eng Phys

J Biomech

J Biomech

Sens Actuators A: Phys

J Biomech

Falls in the elderly: detection and assessment

The Economic cost of hip fractures among elderly women

J Bone Joint Surg Am

Persons found in their homes helpless or dead.

N Engl J Med

Development of novel algorithm and real-time monitoring ambulatory system using Bluetooth module for fall detection in the elderly

Conf Proc IEEE Eng Med Biol Soc

The design of a practical and reliable fall detector for community and institutional telecare

J Telemed Telecare