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.     

Martial arts fall techniques reduce hip impact forces in naive subjects after a brief period of training.

Hip fractures are among the most serious consequences of falls in the elderly. Martial arts (MA) fall techniques may reduce hip fracture risk, as they are known to reduce hip impact forces by approximately 30% in experienced fallers. The purpose of this study was to investigate whether hip impact forces and velocities in MA falls would be smaller than in a 'natural' fall arrest strategy (Block) in young adults (without any prior experience) after a 30-min training session in sideways MA fall techniques. Ten subjects fell sideways from kneeling height. In order to identify experience-related differences, additional EMG data of both fall types were collected in inexperienced (n=10) and experienced fallers (n=5). Compared to Block falls, MA falls had significantly smaller hip impact forces (-17%) and velocities (-7%). EMG results revealed experience-related differences in the execution of the MA fall, indicative of less pronounced trunk rotation in the inexperienced fallers. This may explain their smaller reduction of impact forces compared to experienced fallers. In conclusion, the finding that a substantial reduction in impact forces can be achieved after a short training in MA techniques is very promising with respect to their use in interventions to prevent fall injuries.     

The effectiveness of falls clinics in the Netherlands

Recently, two articles are published about the effectiveness of a multifactorial fall prevention programme performed by two falls clinics in the Netherlands. Both studies have shown negative results. The question is if those two studies are representative for the real situation in most falls clinics in the Netherlands. Two important differences are the selection of the patients and the completion of the fall prevention program. It is important to select the patients with the highest fall risk (> or = 2 falls/year and/or > or = 4 fall risk factors) and to provide an active and direct management of the identified fall risk factors to improve the effectiveness of falls clinics.     

Wrist impact velocities are smaller in forward falls than backward falls from standing

The wrist is a common fracture site for both young and older adults. The purpose of this study was to compare wrist kinematics in backward and forward falls with different fall protective responses. We carried out within-subject comparison of impact velocities and maximum velocities during descent of the distal radius among three different fall configurations: (a) backward falls with knees flexed, (b) backward falls with knees extended and (c) forward falls with knees flexed. We also examined the effect of fall configuration on fall durations, elbow flexion, trunk flexion and forearm angles at impact. Forward falls resulted in smaller impact velocities of the distal radius, longer fall duration, longer braking duration, greater elbow flexion and more horizontal landing position of the forearm compared to backward falls. The distal radius impact velocity during forward falls (1.33 m/s) was significantly lower than in backward falls, and among the backward falls the impact velocity of the flexed knee strategy (2.01 m/s) was significantly lower than the extended knee strategy (2.27 m/s). These impact velocities were significantly reduced from the maximum velocities observed during descent (forward falls=3.57 m/s, backward falls with knee flexed=3.16 m/s, backward falls with knees extended=3.52 m/s). We conclude that (1) smaller impact velocities of the wrists in forward falls could imply a lower fracture risk compared to backward falls, and (2) fall protective responses reduced wrist impact velocities in all fall directions.     

CBO guidelines to prevent accidental falls in the elderly: how can it be used in the institutionalized elderly?

Fall incidents occur frequently in the community dwelling elderly and even more in the institutionalised elderly. Fall-related research data indicate positive effects of a multifactorial intervention targeted on prevention of falls and fall-related injuries. In November 2004 the guideline "Prevention of fall incidents in the elderly" developed by The Dutch Institute for Healthcare Improvement (CBO) was published. This guideline pays attention to the risk factors for falling and the prevention of fall incidents in all settings. The highlights for nursing homes are: all nursing home patients are at risk; perform a fall risk assessment to direct fall preventive activities; together with specific fall prevention for the patient general fall prevention for the institute has to be undertaken; a multifactorial approach is indicated. In nursing homes it is possible to perform such approach multidisciplinary.     

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.     

Adjusted geometric-mean: a novel performance measure for imbalanced bioinformatics datasets learning

One common and challenging problem faced by many bioinformatics applications, such as promoter recognition, splice site prediction, RNA gene prediction, drug discovery and protein classification, is the imbalance of the available datasets. In most of these applications, the positive data examples are largely outnumbered by the negative data examples, which often leads to the development of sub-optimal prediction models having high negative recognition rate (Specificity = SP) and low positive recognition rate (Sensitivity = SE). When class imbalance learning methods are applied, usually, the SE is increased at the expense of reducing some amount of the SP. In this paper, we point out that in these data-imbalanced bioinformatics applications, the goal of applying class imbalance learning methods would be to increase the SE as high as possible by keeping the reduction of SP as low as possible. We explain that the existing performance measures used in class imbalance learning can still produce sub-optimal models with respect to this classification goal. In order to overcome these problems, we introduce a new performance measure called Adjusted Geometric-mean (AGm). The experimental results obtained on ten real-world imbalanced bioinformatics datasets demonstrates that the AGm metric can achieve a lower rate of reduction of SP than the existing performance metrics, when increasing the SE through class imbalance learning methods. This characteristic of AGm metric makes it more suitable for achieving the proposed classification goal in imbalanced bioinformatics datasets learning.     

Falls among community-dwelling elderly: a pilot study of prevalence, circumstances and consequences in Flanders

This pilot study was designed to examine prevalence, circumstances and consequences of falls among 131 community-dwelling elderly (mean age = 76.7 y.; SD = 5.4; response = 29.8%). Based on a retrospective analysis, 39.7% of our study population had experienced one (65.4%) or more (34.6%) falls in the preceding year. Falls appeared to occur predominantly while getting up from the bed, chair or bath (18%), cycling (14%), walking on an uneven surface (14%) and during household work (14%). Fall-related injuries were reported in 67.3% of the fallers, fractures in 17.4%. Additionally, a majority of fallers (63.5%) experienced a mild to serious form of anxiety, 48.1% consulted a general practitioner after the fall and 34.6% was hospitalised. In line with data reported in the literature, falls are a common problem in Flanders and are associated with a negative impact on the physical and psychosocial well being of the elderly population. We conclude that there is an urgent need to develop and implement preventive measures as well as multidisciplinary strategies to identify, assess and target high-risk persons for falling.     

Cluster randomised trial of a targeted multifactorial intervention to prevent falls among older people in hospital

Objective To determine the efficacy of a targeted multifactorial falls prevention programme in elderly care wards with relatively short lengths of stay.Design Cluster randomised trial.Setting 24 elderly care wards in 12 hospitals in Sydney, Australia.Participants 3999 patients, mean age 79 years, with a median hospital stay of seven days.Interventions A nurse and physiotherapist each worked for 25 hours a week for three months in all intervention wards. They provided a targeted multifactorial intervention that included a risk assessment of falls, staff and patient education, drug review, modification of bedside and ward environments, an exercise programme, and alarms for selected patients.Main outcome measure Falls during hospital stay.Results Intervention and control wards were similar at baseline for previous rates of falls and individual patient characteristics. Overall, 381 falls occurred during the study. No difference was found in fall rates during follow-up between intervention and control wards: respectively, 9.26 falls per 1000 bed days and 9.20 falls per 1000 bed days (P=0.96). The incidence rate ratio adjusted for individual lengths of stay and previous fall rates in the ward was 0.96 (95% confidence interval 0.72 to 1.28).Conclusion A targeted multifactorial falls prevention programme was not effective among older people in hospital wards with relatively short lengths of stay.Trial registration Australian New Zealand Clinical Trials Registry ACTRNO 12605000467639.     

Risk profiles and preventive measures of falls in elderly persons

Falls in elderly persons are an important health problem. The results of the Longitudinal Aging Study Amsterdam show that thirty percent of older adults over the age of 65 years who live in the community (n = 1285) fall at least once a year. Recurrent falls were reported by about 11% of the participants. In one-year of follow-up, 22 fractures were recorded. In the 'single fall' group 11 subjects (3.9%) suffered from a fracture and in the 'recurrent fall' group 9 subjects (6.1%). The strongest predictors identified in the risk profile for recurrent falls were previous falls, urinary incontinence, visual impairment and functional limitations (Area Under Curve, 0.71). The probability of recurrent falls for subsequent scores of the screening test ranged from 4.7% (95% CI, 4.0-5.4%) to 46.8% (95% CI, 43.0-50.6%). Risk profiles are needed to identify people at high risk. For matters of feasibility and efficiency, preventive measures of falls should preferably be focussed on those subgroups that have the highest risk of falls.     

Martial arts fall techniques decrease the impact forces at the hip during sideways falling

Falls to the side and those with impact on the hip are risky for hip fractures in the elderly. A previous study has indicated that martial arts (MA) fall techniques can reduce hip impact force, but the underlying mechanism is unknown. Furthermore, the high impact forces at the hand used to break the fall have raised concerns because of the risk for wrist fractures. The purpose of the study was to get insight into the role of hand impact, impact velocity, and trunk orientation in the reduction of hip impact force in MA techniques. Six experienced judokas performed sideways falls from kneeling height using three fall techniques: block with arm technique (control), MA technique with use of the arm to break the fall (MA-a), and MA technique without use of the arm (MA-na). The results showed that the MA-a and MA-na technique reduced the impact force by 27.5% and 30%, respectively. Impact velocity was significantly reduced in the MA falls. Trunk orientation was significantly less vertical in the MA-a falls. No significant differences were found between the MA techniques. It was concluded that the reduction in hip impact force was associated with a lower impact velocity and less vertical trunk orientation. Rolling after impact, which is characteristic for MA falls, is likely to contribute to the reduction of impact forces, as well. Using the arm to break the fall was not essential for the MA technique to reduce hip impact force. These findings provided support for the incorporation of MA fall techniques in fall prevention programs for elderly.     

Genetic Diversity of Algal Viruses Which Lyse the Photosynthetic Picoflagellate Micromonas pusilla (Prasinophyceae)

The genetic similarity among eight clones of Micromonas pusilla virus (MpV) isolated from five geographic locations was measured by DNA hybridization. Our objective was to explore the existence of genetically distinct populations of MpV by comparing the similarity among MpVs isolated from a single water sample to the similarity among viruses isolated from geographically distant locations. The highest and lowest similarities we observed were 70% (plusmn) 1.1% (mean (plusmn) standard error [SE], n = 3) for virus strains SP1 and SP2 isolated from a California coastal water sample and 13% (plusmn) 1.9% for strains SP2 and PB6; the latter was isolated from New York estuarine water. However, the similarity between MpV isolated from a single water sample was not always greater than the similarity between viruses isolated from different locations. Viruses PB7 and PB8 were isolated from a single New York estuarine sample but were only 16% (plusmn) 0.5% similar, whereas PB7 was quite similar (43% (plusmn) 2.9%) to PL1, a virus from Texas coastal water. Overall, the similarity among MpVs isolated from a single geographic location, 34% (plusmn) 12.6% (mean (plusmn) SE, n = 4), was not significantly different from the similarity among MpVs isolated from geographically distant locations, 26.6% (plusmn) 2.7% (mean (plusmn) SE, n = 24) (P = 0.92, Mann-Whitney U test). Clones of MpV were more similar to each other than they were to the related algal virus PBCV-1, and three groups of MpVs consisting of (i) PL1, SG1, PB6, and PB7, (ii) PB8, and (iii) GM1, SP1, and SP2 were resolved. The genetic variation among MpVs isolated from a single water sample was as large as the variation between viruses isolated from different oceans. If MpVs within a geographic location share genetic characteristics not shared with MpVs from geographically distant locations, this was not reflected in the overall similarity of their genomes.     

Reliability of impact forces, hip angles and velocities during simulated forward falls using a novel Propelled Upper Limb fall ARrest Impact System (PULARIS)

Previous forward fall simulation methods have provided good kinematic and kinetic data, but are limited in that they have started the falls from a stationary position and have primarily simulated uni-directional motion. Therefore, a novel Propelled Upper Limb fall ARest Impact System (PULARIS) was designed to address these issues during assessments of a variety of fall scenarios. The purpose of this study was to present PULARIS and evaluate its ability to impact the upper extremities of participants with repeatable velocities, hand forces and hip angles in postures and with vertical and horizontal motion consistent with forward fall arrest. PULARIS consists of four steel tubing crossbars in a scissor-like arrangement that ride on metal trolleys within c-channel tracks in the ceiling. Participants are suspended beneath PULARIS by the legs and torso in a prone position and propelled horizontally via a motor and chain drive until they are quick released, and then impact floor-mounted force platforms with both hands. PULARIS velocity, hip angles and velocities and impact hand forces of ten participants (five male, five female) were collected during three fall types (straight-arm, self-selected and bent-arm) and two fall heights (0.05 m and 0.10 m) to assess the reliability of the impact conditions provided by the system. PULARIS and participant hip velocities were found to be quite repeatable (mean ICC?=?0.81) with small between trial errors (mean?=?0.03 m/s). The ratio of horizontal to vertical hip velocity components (~0.75) agreed well with previously reported data (0.70-0.80). Peak vertical hand impact forces were also found to be relatively consistent between trials with a mean ICC of 0.73 and mean between trial error of 13.4 N. Up to 83% of the horizontal hand impact forces displayed good to excellent reliability (ICC?>?0.6) with small between trial differences. Finally, the ICCs for between trial hip angles were all classified as good to excellent. Overall, PULARIS is a reliable method and is appropriate for studying the response of the distal upper extremity to impact loading during non-stationary, multi-directional movements indicative of a forward fall. This system performed well at different fall heights, and allows for a variety of upper and lower extremity, and hip postures to be tested successfully in different landing scenarios consistent with elderly and sport-related falls.     

Can martial arts techniques reduce fall severity? An in vivo study of femoral loading configurations in sideways falls

Sideways falls onto the hip are a major cause of femoral fractures in the elderly. Martial arts (MA) fall techniques decrease hip impact forces in sideways falls. The femoral fracture risk, however, also depends on the femoral loading configuration (direction and point of application of the force). The purpose of this study was to determine the effect of fall techniques, landing surface and fall height on the impact force and the loading configuration in sideways falls. Twelve experienced judokas performed sideways MA and Block ('natural') falls on a force plate, both with and without a judo mat on top. Kinematic and force data were analysed to determine the hip impact force and the loading configuration. In falls from a kneeling position, the MA technique reduced the impact force by 27%, but did not change the loading configuration. The use of the mat did not change the loading configuration. Falling from a standing changed the force direction. In all conditions, the point of application was distal and posterior to the greater trochanter, but it was less distal and more posterior in falls from standing than from kneeling position. The present decrease in hip impact force with an unchanged loading configuration indicates the potential protective effect of the MA technique on the femoral fracture risk. The change in loading configuration with an increased fall height warrant further studies to examine the effect of MA techniques on fall severity under more natural fall circumstances.     

Lactate threshold and performance adaptations to 4 weeks of training in untrained swimmers: volume vs. intensity

The purpose of this study was to investigate the impact of 4 weeks of high-intensity vs. high-volume swim training on lactate threshold (LT) characteristics and performance. Thirteen untrained swimmers with a mean age of 19.0 ± 0.5 undertook an incremental swimming test before and after 4 weeks of training for the determination of LT. Performance was evaluated by a 50-m maximum freestyle test. The swimmers were assigned to 1 of each of 2 training groups. The high-intensity group (n = 6) focused on sprint training (SP) and swam a total of 1,808 ± 210 m. The high-volume group (n = 7) followed the same program as the SP group but swam an additional 1,100 m (38% more) of endurance swimming (SP + End). A training effect was evident in both groups as seen by the similar improvements in sprint performance of the 50-m maximum time (p < 0.01), peak velocity increases and the lower value of lactate at the individual LTs (p < 0.01). Lactate threshold velocity improved only in the SP + End group from 1.20 ± 0.12 m·s(-1) pretraining to 1.32 ± 0.12 m·s(-1) posttraining (p = 0.77, effect size = 1, p < 0.01), expressed by the rightward shifts of the individual lactate-velocity curves, indicating an improvement in the aerobic capacity. Peak lactate and lactate concentrations at LT did not significantly change. In conclusion, this study was able to demonstrate that 4 weeks of either high-intensity or high-volume training was able to demonstrate similar improvements in swimming performance. In the case of lack of significant changes in lactate profiling in response to high-intensity training, we could suggest a dissociation between the 2.     

Neuromuscular fatigue during repetitive stimulation in elderly and young adults.

The purpose of this investigation was to examine the integrity of neuromuscular transmission and impulse propagation during fatigue by examining the muscle compound action potential (M wave) in elderly and young adults. The tibialis anterior muscle of nine elderly [mean = 67.7 (SE 1.7) years] and nine young [mean = 26.7 (SE 1.2) years] adults was maximally stimulated repetitively at frequencies of 20, 30 or 40 Hz for 60 s on separate occasions. There was a significantly smaller resting M wave amplitude [7.9 (SE 0.4) mV versus 9.9 (SE 0.6) mV] and M wave area [0.038 (SE 0.005) mV s versus 0.06 (SE 0.004) mV.s] in the elderly versus the young adults respectively. Measurement of the evoked muscle contractile properties revealed significantly (P < 0.05) longer twitch durations and a significantly (P < 0.05) greater peak twitch torque [4.6 (SE 0.4) Nm versus 3.2 (SE 0.5) Nm] in the elderly versus the young adults, respectively. The elderly adults had a significantly greater torque decline during the 20-Hz trial; however, the decline in torque during the 30-Hz and 40-Hz trials was similar in the elderly and the young adults (30 Hz: 40%; 40 Hz: 56%). Throughout each of the stimulation trials, the decline in torque was accompanied by a significant reduction in M wave amplitude (20 Hz: 14%; 30 Hz: 53%; 40 Hz: 67%); M wave area also declined significantly during the 30-Hz (31%) and 40-Hz (53%) trials. There was no significant difference between the elderly and the young adults in the reduction in the M wave amplitude or area during each trial.(ABSTRACT TRUNCATED AT 250 WORDS)     

Conventional Pap smear and liquid-based cytology as screening tools in low-resource settings in Latin America: experience of the Latin American screening study

OBJECTIVE: To evaluate the performance of the conventional Pap test and liquid-based cytology (LBC) in an ongoing multicenter trial testing optional screening tools (cytology, screening colposcopy, visual inspection with acetic acid, visual inspection with Lugol's Iodine, cervicography and Hybrid Capture II [HCII] (Digene Brazil, S?o Paulo, Brazil) conventional and self-sampling), for cervical cancer in Brazil and Argentina. STUDY DESIGN: A cohort of 12,107 women attending four clinics (Campinas, S?o Paulo, Porto Alegre, Buenos Aires) were randomized into the 8 diagnostic arms. Women testing positive with any of the tests were referred for colposcopy, and cervical biopsies were used as the gold standard to assess performance characteristics of the diagnostic tests. Conventional Pap smears were sampled by all clinics (n = 10,240), and LBC (Autocyte PREP, [TriPath Imaging, Burlington, North Carolina, U.S.A.], n=320, and DNA-Citoliq [Digene Brazil], n =1,346) was performed by 1 of the clinics. RESULTS: Conventional Pap smears showed no squamous intraepithelial lesions (normal) in 8,946 (87.4%) and LBC in 1,373 (82.4%). Using high grade squamous intraepithelial lesions (HSIL) as the cutoff, Pap smears predicted high grade (cervical intraepithelial neoplasia [CIN] 3) with OR 63.0 (95% CI, 36.90-107.70), standard error (SE) 59%, SP 97.8%, positive predictive value (PPV) 68.1% and negative predictive value (NPV) 96.7%. The same figures for Autocyte PREP were: OR 9.0 (95% CI, 2.43-33.24), sensitivity (SE) 33.3%, specificity (SP) 100%, PPV 100% and negative PV (NPV) 88.8%. DNA-Citoliq detected CIN 3 as follows: OR 11.8 (95% CI 2.60-53.26), SE 40.0%, SP 94.6%, PPV 40.0% and NPV 94.6%. Lowering the cutoff to low grade squamous intraepithelial lesions increased SE and NPV but compromised SP and PPV. The detection rates for high grade lesions after an atypical squamous cells of undetermined significance diagnosis were similar with the 3 techniques. In our settings, the 3 methods of cervical cytology were slightly different in performance. The conventional Pap smear had the highest SE, while Autocyte PREP had 100% SP and PPV in detecting CIN3 with the HSIL cutoff. All 3 tests had lower SE but higher SP as compared to HCII.     

Possible Role of Nutrition in Prevention of Sarcopenia and Falls

Objective: In this review, we analyze the foundation of sarcopenia as a potentially modifiable risk factor for falls, and we try to formulate practical strategies for nutritional interventions aimed at reducing the risk for sarcopenia and falls in our elderly patients.Methods: An extensive literature search was performed using the PubMed and the Google Scholar databases.Results: Falls are a common and costly source of injury and death in elderly adults. A large proportion of injurious falls are due to a trip or slip, suggesting that muscular factors are major determinants of both fall risk and the risk for fall-related injury.Conclusion: An increasing body of evidence links sarcopenia, the loss of muscle strength and mass that occurs with advancing age, with an increased risk for falls. Nutritional factors, as well as exercise, can help with both prevention and treatment of sarcopenia and may reduce the risk of falls in the elderly.Abbreviations: 25-OHD = 25-hydroxyvitamin D; EAA = essential amino acid; IGF-1 = insulin-like growth factor 1; IU = international units; MPS = muscle protein synthesis; PUFA = polyunsaturated fatty acid     

Disturbance type and gait speed affect fall direction and impact location

Since falling to the side and impacting on or near the hip increase hip fracture risk, we examined the fall direction and pelvis impact location resulting from four disturbances (faint, slip, step down, trip) at three gait speeds (fast, normal, slow) in 14 young adults instructed not to attempt recovery. We hypothesized that certain disturbances such as faints and slips and slow walking speed were more likely to result in an impact on the hip. For each trial, the fall direction, impact location and pelvis impact velocity were measured. The results showed that both disturbance type and gait speed significantly affected fall direction and impact location (analysis of covariance with repeated measures, p< or =0.0001) with a significant interaction (p<0.05). Trips and steps down usually resulted in forward falls, with frontal impacts regardless of gait speed. At fast gait speed, slips and faints also usually resulted in forward falls, with frontal impacts. As gait speed decreased, however, slips usually resulted in sideways or backward falls, with impact on the hip or buttocks, and faints resulted in a greater number of sideways falls, with impact near the hip. Therefore, compared to other disturbances and gait speeds, slipping or fainting while walking slowly was more likely to result in an impact on the hip, suggesting a greater risk for hip fracture. Furthermore, 56% of the impact velocities generated were within one standard deviation of the estimate of the mean impact velocity needed to fracture an elderly femur.