Detecting Falls as Novelties in Acceleration Patterns Acquired with Smartphones

Despite being a major public health problem, falls in the elderly cannot be detected efficiently yet. Many studies have used acceleration as the main input to discriminate between falls and activities of daily living (ADL). In recent years, there has been an increasing interest in using smartphones for fall detection. The most promising results have been obtained by supervised Machine Learning algorithms. However, a drawback of these approaches is that they rely on falls simulated by young or mature people, which might not represent every possible fall situation and might be different from older people''s falls. Thus, we propose to tackle the problem of fall detection by applying a kind of novelty detection methods which rely only on true ADL. In this way, a fall is any abnormal movement with respect to ADL. A system based on these methods could easily adapt itself to new situations since new ADL could be recorded continuously and the system could be re-trained on the fly. The goal of this work is to explore the use of such novelty detectors by selecting one of them and by comparing it with a state-of-the-art traditional supervised method under different conditions. The data sets we have collected were recorded with smartphones. Ten volunteers simulated eight type of falls, whereas ADL were recorded while they carried the phone in their real life. Even though we have not collected data from the elderly, the data sets were suitable to check the adaptability of novelty detectors. They have been made publicly available to improve the reproducibility of our results. We have studied several novelty detection methods, selecting the nearest neighbour-based technique (NN) as the most suitable. Then, we have compared NN with the Support Vector Machine (SVM). In most situations a generic SVM outperformed an adapted NN.     

Automatic individual calibration in fall detection – an integrative ambulatory measurement framework

The objective of the current study was to demonstrate the utility of a new integrative ambulatory measurement (IAM) framework by developing and evaluating an individual calibration function in fall detection application. Ten healthy elderly persons were involved in a laboratory study and tested in a protocol comprising various types of activities of daily living and slip-induced backward falls. Inertial measurement units attached to the trunk and thigh segments were used to measure trunk angular kinematics and thigh accelerations. The effect of individual calibration was evaluated with previously developed fall detection algorithm. The results indicated that with individual calibration, the fall detection performance achieved approximately the same level of sensitivity (100% vs. 100%) and specificity (95.25% vs. 95.65%); however, response time was significantly lower than without (249 ms vs. 255 ms). It was concluded that the automatic individual calibration using the IAM framework improves the performance of fall detection, which has a greater implication in preventing/minimising injuries associated with fall accidents.     

Evaluation of accelerometer-based fall detection algorithms on real-world falls

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 wearable system for pre-impact fall detection

Unique features of body segment kinematics in falls and activities of daily living (ADL) are applied to make automatic detection of a fall in its descending phase, prior to impact, possible. Fall-related injuries can thus be prevented or reduced by deploying fall impact reduction systems, such as an inflatable airbag for hip protection, before the impact. In this application, the authors propose the following hypothesis: “Thigh segments normally do not exceed a certain threshold angle to the side and forward directions in ADL, whereas this abnormal behavior occurs during a fall activity”. Torso and thigh wearable inertial sensors (3D accelerometer and 2D gyroscope) are used and the whole system is based on a body area network (BAN) for the comfort of the wearer during a long term application. The hypothesis was validated in an experiment with 21 young healthy volunteers performing both normal ADL and fall activities. Results show that falls could be detected with an average lead-time of 700 ms before the impact occurs, with no false alarms (100% specificity), a sensitivity of 95.2%. This is the longest lead-time achieved so far in pre-impact fall detection.     

Machine learning algorithms based on signals from a single wearable inertial sensor can detect surface- and age-related differences in walking

The aim of this study was to investigate if a machine learning algorithm utilizing triaxial accelerometer, gyroscope, and magnetometer data from an inertial motion unit (IMU) could detect surface- and age-related differences in walking. Seventeen older (71.5 ± 4.2 years) and eighteen young (27.0 ± 4.7 years) healthy adults walked over flat and uneven brick surfaces wearing an inertial measurement unit (IMU) over the L5 vertebra. IMU data were binned into smaller data segments using 4-s sliding windows with 1-s step lengths. Ninety percent of the data were used as training inputs and the remaining ten percent were saved for testing. A deep learning network with long short-term memory units was used for training (fully supervised), prediction, and implementation. Four models were trained using the following inputs: all nine channels from every sensor in the IMU (fully trained model), accelerometer signals alone, gyroscope signals alone, and magnetometer signals alone. The fully trained models for surface and age outperformed all other models (area under the receiver operator curve, AUC = 0.97 and 0.96, respectively; p ≤ .045). The fully trained models for surface and age had high accuracy (96.3, 94.7%), precision (96.4, 95.2%), recall (96.3, 94.7%), and f1-score (96.3, 94.6%). These results demonstrate that processing the signals of a single IMU device with machine-learning algorithms enables the detection of surface conditions and age-group status from an individual’s walking behavior which, with further learning, may be utilized to facilitate identifying and intervening on fall risk.     

Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 2: Controlled trial in multiple centres

ObjectivesTo assess the effectiveness of trained nurses based in general practices individually prescribing a home exercise programme to reduce falls and injuries in elderly people and to estimate the cost effectiveness of the programme.DesignControlled trial with one year''s follow up.Setting32 general practices in seven southern New Zealand centres.Participants450 women and men aged 80 years and older.Intervention330 participants received the exercise programme (exercise centres) and 120 received usual care (control centres); 87% (371 of 426) completed the trial.ResultsFalls were reduced by 30% in the exercise centres (incidence rate ratio 0.70, 95% confidence interval 0.59 to 0.84). The programme was equally effective in men and women. The programme cost $NZ418 (£121) (at 1998 prices) per person to deliver for one year or $NZ1519 (£441) per fall prevented. Fewer participants had falls resulting in injuries, but there was no difference in the number who had serious injuries and no difference in hospital costs resulting from falls in exercise centres compared with control centres.ConclusionsAn individually tailored exercise programme, delivered by trained nurses from within general practices, was effective in reducing falls in three different centres. This strategy should be combined with other successful interventions to form part of home programmes to prevent falls in elderly people.

What is already known on this topic

One half of those aged 80 years and older will fall in any one year, often with serious health and social consequencesAn exercise programme delivered by a physiotherapist or trained district nurse was successful in reducing falls and moderate injuries in elderly people

What this study adds

An exercise programme to prevent falls in elderly people can be delivered safely and effectively by trained nurses in general practicesThe nurses obtained results that were consistent with the physiotherapist in the research setting and the district nurse in the accompanying paper     

CNV-CH: A Convex Hull Based Segmentation Approach to Detect Copy Number Variations (CNV) Using Next-Generation Sequencing Data

Copy number variation (CNV) is a form of structural alteration in the mammalian DNA sequence, which are associated with many complex neurological diseases as well as cancer. The development of next generation sequencing (NGS) technology provides us a new dimension towards detection of genomic locations with copy number variations. Here we develop an algorithm for detecting CNVs, which is based on depth of coverage data generated by NGS technology. In this work, we have used a novel way to represent the read count data as a two dimensional geometrical point. A key aspect of detecting the regions with CNVs, is to devise a proper segmentation algorithm that will distinguish the genomic locations having a significant difference in read count data. We have designed a new segmentation approach in this context, using convex hull algorithm on the geometrical representation of read count data. To our knowledge, most algorithms have used a single distribution model of read count data, but here in our approach, we have considered the read count data to follow two different distribution models independently, which adds to the robustness of detection of CNVs. In addition, our algorithm calls CNVs based on the multiple sample analysis approach resulting in a low false discovery rate with high precision.     

Physical Activity and Different Concepts of Fall Risk Estimation in Older People–Results of the ActiFE-Ulm Study

Objectives

To investigate the relationship between physical activity and two measures of fall incidence in an elderly population using person-years as well as hours walked as denominators and to compare these two approaches.

Design

Prospective cohort study with one-year follow-up of falls using fall calendars. Physical activity was defined as walking duration and recorded at baseline over one week using a thigh-worn uni-axial accelerometer (activPAL; PAL Technologies, Glasgow, Scotland). Average daily physical activity was extracted from these data and categorized in low (0–59 min), medium (60–119 min) and high (120 min and more) activity.

Setting

The ActiFE Ulm study located in Ulm and adjacent regions in Southern Germany.

Participants

1,214 community-dwelling older people (≥65 years, 56.4% men).

Measurements

Negative-binomial regression models were used to calculate fall rates and incidence rate ratios for each activity category each with using (1) person-years and (2) hours walked as denominators stratified by gender, age group, fall history, and walking speed. All analyses were adjusted either for gender, age, or both.

Results

No statistically significant association was seen between falls per person-year and average daily physical activity. However, when looking at falls per 100 hours walked, those who were low active sustained more falls per hours walked. The highest incidence rates of falls were seen in low-active persons with slow walking speed (0.57 (95% confidence interval (95% CI): 0.33 to 0.98) falls per 100 hours walked) or history of falls (0.60 (95% CI: 0.36 to 0.99) falls per 100 hours walked).

Conclusion

Falls per hours walked is a relevant and sensitive outcome measure. It complements the concept of incidence per person years, and gives an additional perspective on falls in community-dwelling older people.     

Fall classification by machine learning using mobile phones

Fall prevention is a critical component of health care; falls are a common source of injury in the elderly and are associated with significant levels of mortality and morbidity. Automatically detecting falls can allow rapid response to potential emergencies; in addition, knowing the cause or manner of a fall can be beneficial for prevention studies or a more tailored emergency response. The purpose of this study is to demonstrate techniques to not only reliably detect a fall but also to automatically classify the type. We asked 15 subjects to simulate four different types of falls-left and right lateral, forward trips, and backward slips-while wearing mobile phones and previously validated, dedicated accelerometers. Nine subjects also wore the devices for ten days, to provide data for comparison with the simulated falls. We applied five machine learning classifiers to a large time-series feature set to detect falls. Support vector machines and regularized logistic regression were able to identify a fall with 98% accuracy and classify the type of fall with 99% accuracy. This work demonstrates how current machine learning approaches can simplify data collection for prevention in fall-related research as well as improve rapid response to potential injuries due to falls.     

Scale Invariance Properties of Intracerebral EEG Improve Seizure Prediction in Mesial Temporal Lobe Epilepsy

Although treatment for epilepsy is available and effective for nearly 70 percent of patients, many remain in need of new therapeutic approaches. Predicting the impending seizures in these patients could significantly enhance their quality of life if the prediction performance is clinically practical. In this study, we investigate the improvement of the performance of a seizure prediction algorithm in 17 patients with mesial temporal lobe epilepsy by means of a novel measure. Scale-free dynamics of the intracerebral EEG are quantified through robust estimates of the scaling exponents—the first cumulants—derived from a wavelet leader and bootstrap based multifractal analysis. The cumulants are investigated for the discriminability between preictal and interictal epochs. The performance of our recently published patient-specific seizure prediction algorithm is then out-of-sample tested on long-lasting data using combinations of cumulants and state similarity measures previously introduced. By using the first cumulant in combination with state similarity measures, up to 13 of 17 patients had seizures predicted above chance with clinically practical levels of sensitivity (80.5%) and specificity (25.1% of total time under warning) for prediction horizons above 25 min. These results indicate that the scale-free dynamics of the preictal state are different from those of the interictal state. Quantifiers of these dynamics may carry a predictive power that can be used to improve seizure prediction performance.     

Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 1: Randomised controlled trial

ObjectivesTo assess the effectiveness of a trained district nurse individually prescribing a home based exercise programme to reduce falls and injuries in elderly people and to estimate the cost effectiveness of the programme.DesignRandomised controlled trial with one year''s follow up.SettingCommunity health service at a New Zealand hospital.Participants240 women and men aged 75 years and older.Intervention121 participants received the exercise programme (exercise group) and 119 received usual care (control group); 90% (211 of 233) completed the trial.ResultsFalls were reduced by 46% (incidence rate ratio 0.54, 95% confidence interval 0.32 to 0.90). Five hospital admissions were due to injuries caused by falls in the control group and none in the exercise group. The programme cost $NZ1803 (£523) (at 1998 prices) per fall prevented for delivering the programme and $NZ155 per fall prevented when hospital costs averted were considered.ConclusionA home exercise programme, previously shown to be successful when delivered by a physiotherapist, was also effective in reducing falls when delivered by a trained nurse from within a home health service. Serious injuries and hospital admissions due to falls were also reduced. The programme was cost effective in participants aged 80 years and older compared with younger participants.

What is already known on this topic

Falls are the costliest type of injury among elderly people, and the healthcare costs increase with frequency of falls and severity of injuriesAn exercise programme delivered by a physiotherapist was successful in reducing falls and moderate injuries in elderly people

What this study adds

An exercise programme to prevent falls in elderly people worked well when delivered by a district nurse from a home health service in the suburbs of a large cityResearchers, public health administrators, and health practitioners can work together to benefit elderly people in the community     

Automatic Fall Detection System Based on the Combined Use of a Smartphone and a Smartwatch

Due to their widespread popularity, decreasing costs, built-in sensors, computing power and communication capabilities, Android-based personal devices are being seen as an appealing technology for the deployment of wearable fall detection systems. In contrast with previous solutions in the existing literature, which are based on the performance of a single element (a smartphone), this paper proposes and evaluates a fall detection system that benefits from the detection performed by two popular personal devices: a smartphone and a smartwatch (both provided with an embedded accelerometer and a gyroscope). In the proposed architecture, a specific application in each component permanently tracks and analyses the patient’s movements. Diverse fall detection algorithms (commonly employed in the literature) were implemented in the developed Android apps to discriminate falls from the conventional activities of daily living of the patient. As a novelty, a fall is only assumed to have occurred if it is simultaneously and independently detected by the two Android devices (which can interact via Bluetooth communication). The system was systematically evaluated in an experimental testbed with actual test subjects simulating a set of falls and conventional movements associated with activities of daily living. The tests were repeated by varying the detection algorithm as well as the pre-defined mobility patterns executed by the subjects (i.e., the typology of the falls and non-fall movements). The proposed system was compared with the cases where only one device (the smartphone or the smartwatch) is considered to recognize and discriminate the falls. The obtained results show that the joint use of the two detection devices clearly increases the system’s capability to avoid false alarms or ‘false positives’ (those conventional movements misidentified as falls) while maintaining the effectiveness of the detection decisions (that is to say, without increasing the ratio of ‘false negatives’ or actual falls that remain undetected).     

Falls in People with Multiple Sclerosis Compared with Falls in Healthy Controls

Objective

To compare the risk, circumstances, consequences and causes of prospectively recorded falls between people with multiple sclerosis (PwMS) and healthy controls of similar age and gender.

Methods

58 PwMS and 58 healthy controls, who are community-dwelling, were recruited in this 6-month prospective cohort study. 90% of PwMS and 84% of healthy controls completed the study. Participants counted falls prospectively using fall calendars and noted fall location, fall-related injuries, and the cause of the falls. Kaplan Meier survival analysis and log-rank tests were performed to compare the distributions of survival without falling between PwMS and healthy controls.

Results

40.8% of controls and 71.2% of PwMS fell at least once. 48.1% of PwMS and 18.4% of healthy controls fell at least twice. 42.3% of PwMS and 20.4% of health controls sustained a fall-related injury. After adjusting for age and gender, the time to first fall (HR: 1.87, p = 0.033) and the time to recurrent falls (HR: 2.87, p = 0.0082) were significantly different between PwMS and healthy controls. PwMS reported an almost equal number of falls inside and outside, 86% of the falls in healthy controls were outside. Healthy controls were more likely to fall due to slipping on a slippery surface (39.5% vs 10.4%). PwMS more often attributed falls to distraction (31% vs 7%) and uniquely attributed falls to fatigue or heat.

Conclusions

Fall risk, circumstances, consequences, and causes are different for PwMS than for healthy people of the same age and gender. PwMS fall more, are more likely to be injured by a fall, and often fall indoors. PwMS, but not healthy controls, frequently fall because they are distracted, fatigued or hot.     

Two types of slip-induced falls among community dwelling older adults

Little is known about the landing behavior of the trailing (recovery) foot and ensuing types of falls following a forward slip in walking. The purposes of this study were to (1) determine if community-dwelling older adults experienced bilateral slips at the same rate as had been previously observed for young adults during over-ground walking; (2) determine if fall rate in older adults was dependent on slip type (unilateral vs. bilateral); and (3) identify differences in spatiotemporal variables of the trailing leg step between unilateral and bilateral slips. One-hundred-seventy-four participants experienced an unannounced, unrehearsed slip while walking on a 7-m walkway. Each trial was monitored with a motion capture system and bilateral ground reaction force plates. Although the experimental design, developed with original data from a young adult population, favored bilateral slips, more older adults (35%) than anticipated (10% previously observed in young, p<0.001) displayed a unilateral slip. The probability of fall was equal in the two types of slips. Eighty-two people recovered from the slip, while the remaining 92 (53%) fell. These 92 were classified into two exclusive categories based on the heel distance at the time of fall arrest using cluster analysis: those which resembled a fall into a "splits" position (n=47) or a feet-forward fall (n=45). All (100%) unilateral slips led to splits falls, as expected. Yet, not all bilateral slips (only 83%) resulted in feet-forward falls. A longer forward recovery step with a prolonged step time led to both feet slipping, nearly together, hence a feet-forward fall.     

Intervenciones para reducir los riesgos multifactoriales de caídas

This article provides an overview of common intervention strategies used to prevent falls and fall-related injuries in older people. Fall incidence increases with age and causes a tremendous amount of morbidity, mortality, and use of healthcare services. Major risk factors for falling are diverse, and many of them —such as balance impairment, unsteady gait, muscle weakness, drug side effects, and environmental hazards— are potentially modifiable. Medical assessment of fall risks and provision of appropriate interventions is challenging due to the complex nature of falls. Consensus panels of experts have developed evidence-based practice guidelines for fall prevention and management. Numerous fall prevention strategies have been studied, and there is considerable evidence to suggest that certain types of fall prevention strategies are more effective than others. Findings from individual studies have been substantiated by careful meta-analysis of large numbers of controlled clinical trials. These meta-analyses have concluded that the most effective (and cost-effective) fall reduction programs involve systematic fall risk assessment and targeted interventions, exercise programs, and environmental inspection and hazard reduction programs. One large recent meta-analysis showed that interventions using multidimensional risk assessment and risk reduction lowered the risk of falling by 18%, while exercise reduced the risk of falls by 12%. Home modification was effective when incorporated into a multi-factorial intervention, targeted to fall-prone individuals. These findings indicate that the most promising approaches to fall prevention will involve interdisciplinary collaboration in assessment and interventions.     

Active responses decrease impact forces at the hip and shoulder in falls to the side.

Active responses, such as using the arm to break the fall, may be an effective means of decreasing likelihood of injury in a fall and may help explain why only a small percentage of falls result in a fracture. We quantified the impact force at the hip and shoulder in falls to the side from a kneeling position under three conditions: (1) attempting to break the fall by using an arm; (2) falling with the body relaxed; and (3) falling with the body tensed. Subjects fell from a kneeling position onto a force platform array covered with foam padding and impact force data were recorded. The ground reaction force-time curve was generally bimodal due to sequential impacts of the hip and shoulder. Impact forces at the hip and shoulder were 12 and 16% less for the slap condition (p < 0.05) than for the tensed condition. The impact forces for the relaxed and tensed conditions were not significantly different, although impact forces tended to be less in the relaxed condition. We concluded that active responses reduce the impact forces experienced at the hip and shoulder in falls to the side. Decreased effectiveness of protective responses, due to increases in reaction time and decreases in strength with age, may help explain why so many hip fractures occur in the elderly but so few occur in younger people.     

Intervenciones basadas en el ejercicio y el entorno para la prevención de caídas en personas con deterioro cognitivo que viven en centros de cuidado: revisión sistemática y metaanálisis

This systematic review aims to report the effectiveness of interventions based on exercise and/or physical environment for reducing falls in cognitively impaired older adults living in long-term care facilities.In July 2014, a literature search was conducted using main databases and specialised sources. Randomised controlled trials assessing the effectiveness of fall prevention interventions, which used exercise or physical environment among elderly people with cognitive impairment living in long-term care facilities, were selected. Two independent reviewers checked the eligibility of the studies, and evaluated their methodological quality. If it was adequate, data were gathered.Fourteen studies with 3,539 participants using exercise and/or physical environment by a single or combined approach were included. The data gathered from studies that used both interventions showed a significant reduction in fall rate.Further research is needed to demonstrate the effectiveness of those interventions for preventing falls in the elderly with cognitive impairment living in long-term care establishments.     

Force-Controlled Balance Perturbations Associated with Falls in Older People: A Prospective Cohort Study

Balance recovery from an unpredictable postural perturbation can be a challenging task for many older people and poor recovery could contribute to their risk of falls. This study examined associations between responses to unpredictable perturbations and fall risk in older people. 242 older adults (80.0±4.4 years) underwent assessments of stepping responses to multi-directional force-controlled waist-pull perturbations. Participants returned monthly falls calendars for the subsequent 12 months. Future falls were associated with lower force thresholds for stepping in the posterior and lateral but not anterior directions. Those with lower posterior force thresholds for stepping were 68% more likely to fall at home than those with higher force thresholds for stepping. These results suggest that amount of force that can be withstood following an unpredictable balance perturbation predicts future falls in community-dwelling older adults. Perturbations in the posterior direction best discriminated between future fallers and non-fallers.     

Detection of Heart Sounds in Children with and without Pulmonary Arterial Hypertension―Daubechies Wavelets Approach

Background

Automatic detection of the 1^st (S1) and 2^nd (S2) heart sounds is difficult, and existing algorithms are imprecise. We sought to develop a wavelet-based algorithm for the detection of S1 and S2 in children with and without pulmonary arterial hypertension (PAH).

Method

Heart sounds were recorded at the second left intercostal space and the cardiac apex with a digital stethoscope simultaneously with pulmonary arterial pressure (PAP). We developed a Daubechies wavelet algorithm for the automatic detection of S1 and S2 using the wavelet coefficient ‘D ₆’ based on power spectral analysis. We compared our algorithm with four other Daubechies wavelet-based algorithms published by Liang, Kumar, Wang, and Zhong. We annotated S1 and S2 from an audiovisual examination of the phonocardiographic tracing by two trained cardiologists and the observation that in all subjects systole was shorter than diastole.

Results

We studied 22 subjects (9 males and 13 females, median age 6 years, range 0.25–19). Eleven subjects had a mean PAP < 25 mmHg. Eleven subjects had PAH with a mean PAP ≥ 25 mmHg. All subjects had a pulmonary artery wedge pressure ≤ 15 mmHg. The sensitivity (SE) and positive predictivity (+P) of our algorithm were 70% and 68%, respectively. In comparison, the SE and +P of Liang were 59% and 42%, Kumar 19% and 12%, Wang 50% and 45%, and Zhong 43% and 53%, respectively. Our algorithm demonstrated robustness and outperformed the other methods up to a signal-to-noise ratio (SNR) of 10 dB. For all algorithms, detection errors arose from low-amplitude peaks, fast heart rates, low signal-to-noise ratio, and fixed thresholds.

Conclusion

Our algorithm for the detection of S1 and S2 improves the performance of existing Daubechies-based algorithms and justifies the use of the wavelet coefficient ‘D ₆’ through power spectral analysis. Also, the robustness despite ambient noise may improve real world clinical performance.