Kinect-based choice reaching and stepping reaction time tests for clinical and in-home assessment of fall risk in older people: a prospective study

Stepping Beyond Assessment: Fall Risk Prediction Models Among Older Adults from Cumulative Change in Gait Parameter Estimates

Falls significantly affect the health of older adults. Injuries sustained through falls have long-term consequences on the ability to live independently and age in place, and are the leading cause of injury death in the United States for seniors. Early fall risk detection provides an important opportunity for prospective intervention by healthcare providers and home caregivers. In-home depth sensor technologies have been developed for real-time fall detection and gait parameter estimation including walking speed, the sixth vital sign, which has been shown to correlate with the risk of falling. This study evaluates the use of supervised classification for estimating fall risk from cumulative changes in gait parameter estimates as captured by 3D depth sensors placed within the homes of older adult participants. Using recall as the primary metric for model success rate due to the severity of fall injuries sustained by false negatives, we demonstrate an enhancement of assessing fall risk with univariate logistic regression using multivariate logistic regression, support vector, and hierarchical tree-based modeling techniques by an improvement of 18.80%, 31.78%, and 33.94%, respectively, in the 14 days preceding a fall event. Random forest and XGBoost models resulted in recall and precision scores of 0.805 compared to the best univariate regression model of Y-Entropy with a recall of 0.639 and precision of 0.527 for the 14-day window leading to a predicted fall event.

A multifactorial fall risk assessment system for older people utilizing a low-cost, markerless Microsoft Kinect

Falls among older people are a major health concern. This study aims to develop a multifactorial fall risk assessment system for older people using a low-cost, markerless Microsoft Kinect. A Kinect-based test battery was designed to comprehensively assess major fall risk factors. A follow-up experiment was conducted with 102 older participants to assess their fall risks. Participants were divided into high and low fall risk groups based on their prospective falls over a 6-month period. Results showed that the high fall risk group performed significantly worse on the Kinect-based test battery. The developed random forest classification model achieved an average classification accuracy of 84.7%. In addition, the individual's performance was computed as the percentile value of a normative database to visualise deficiencies and targets for intervention. These findings indicate that the developed system can not only screen out 'at risk' older individuals with good accuracy, but also identify potential fall risk factors for effective fall intervention.Practitioner summary: Falls are the leading cause of injuries in older people. We newly developed a multifactorial fall risk assessment system for older people utilising a low-cost, markerless Kinect. Results showed that the developed system can screen out 'at risk' individuals and identify potential risk factors for effective fall intervention.

The Effects of Exergames on Physical and Psychological Health in Young Adults

Objective: This study aims to examine the effects of Kinect-based exergaming on physical (balance and reaction time) and psychosocial health factors (mental well-being and subjective happiness level) in young adults. Materials and Methods: A total of 22 men aged between 18 and 40, who play sedentary videogames at least once a week, were randomly divided into 2 groups to participate in either a Kinect-based exergame program (experimental group, n = 11) or to continue playing sedentary videogames without making any changes (control group, n = 11). The Kinect-based exergame program, including three exergames (Kinect Adventure: Reflex Ridge, Dance Central 3, Kinect Sports: Volleyball), was implemented with the experimental group for 5 weeks, three times a week, and 45 minutes per session. Balance and reaction time were assessed using SportKAT® 4000 Kinesthetic Ability Trainer and FitLight Trainer™ devices, respectively. To measure mental well-being and subjective happiness level, scales with proven psychometric qualities, whose validity and reliability studies were conducted, were used. Results: Results of the study showed statistically significant improvement for the experimental group on the measures of reaction time [t(10) = -4.69, P = 0.001], subjective happiness [t(10) = 8.94, P = 0.000], and mental well-being [t(10) = 4.62, P = 0.001]. In addition, the difference between the groups in reaction time [t(20) = 1.66, P = 0.01], right leg static balance [t(20) = -2.96, P = 0.008], mental well-being [t(20) = 2.77, P = 0.012], and subjective happiness levels (U = 0.50, P = 0.000) was found to be statistically significant in favor of the experimental group. Conclusion: The findings suggest that Kinect-based exergaming positively supports adults' reaction time, mental well-being, and subjective happiness.

Kinect-based rapid movement training to improve balance recovery for stroke fall prevention: a randomized controlled trial

Background

Falls are more prevalent in stroke survivors than age-matched healthy older adults because of their functional impairment. Rapid balance recovery reaction with adequate range-of-motion and fast response and movement time are crucial to minimize fall risk and prevent serious injurious falls when postural disturbances occur. A Kinect-based Rapid Movement Training (RMT) program was developed to provide real-time feedback to promote faster and larger arm reaching and leg stepping distances toward targets in 22 different directions.

Objective

To evaluate the effectiveness of the interactive RMT and Conventional Balance Training (CBT) on chronic stroke survivors’ overall balance and balance recovery reaction.

Methods

In this assessor-blinded randomized controlled trial, chronic stroke survivors were randomized to receive twenty training sessions (60-min each) of either RMT or CBT. Pre- and post-training assessments included clinical tests, as well as kinematic measurements and electromyography during simulated forward fall through a “lean-and-release” perturbation system.

Results

Thirty participants were recruited (RMT = 16, CBT = 14). RMT led to significant improvement in balance control (Berg Balance Scale: pre = 49.13, post = 52.75; P = .001), gait control (Timed-Up-and-Go Test: pre = 14.66 s, post = 12.62 s; P = .011), and motor functions (Fugl-Meyer Assessment of Motor Recovery: pre = 60.63, post = 65.19; P = .015), which matched the effectiveness of CBT. Both groups preferred to use their non-paretic leg to take the initial step to restore stability, and their stepping leg’s rectus femoris reacted significantly faster post-training (P = .036).

Conclusion

The RMT was as effective as conventional balance training to provide beneficial effects on chronic stroke survivors’ overall balance, motor function and improving balance recovery with faster muscle response.

Trial registration: The study was registered at Clinicaltrials.gov (https://clinicaltrials.gov/ct2/show/NCT03183635, NCT03183635) on 12 June 2017.

Ankle motor strategy use in older and younger adults as assessed by a two-dimensional kinematic analysis smart device application

Background/aims

Motor strategies change with age, resulting in balance deficits. Clinical options for objectively measuring motor strategies are limited. The use of two-dimensional kinematic analysis through smart devices and applications may provide a cost-effective portable solution for measuring differences in motor strategy use between older and younger adults. The aims of this study were to investigate the concurrent validity of a two-dimensional software and two-dimensional application and to determine whether the application could capture the difference in ankle motor strategy use by older and younger adults (construct validity).

Methods

Video clips captured by Sony and iPad Air cameras were used to assess concurrent validity between the two-dimensional software and application. A total of 30 older (72.6 ± 4.0 years) and 30 younger (26.5±4.5 years) adults performed forward and backward stepping. A two-dimensional applicationmeasured the ankle position in degrees and time taken in seconds to initiate and complete a compensatory step.

Results

The two-dimensional software and app demonstrated excellent reliability (ICC2,1 ≥0.994). The ankle forward stop angle differed significantly between older and younger adults (P=0.012). No differences were observed in time taken to initiate and complete a compensatory step.

Conclusions

The two-dimensional application appears to be a valid alternative to two-dimensional software for measuring ankle motor strategies. Further investigation for clinical use is warranted.

Chronic Musculoskeletal Pain and Foot Reaction Time in Older Adults

This cross-sectional study examines the association between chronic musculoskeletal pain and foot reaction time (RT) among older community-living adults. Participants were 307 adults aged 71 years and older in the MOBILIZE Boston Study II. Pain severity, interference, and location were measured by the Brief Pain Inventory and a joint pain questionnaire. With participants seated, simple foot reaction time was measured as self-selected foot response time to an intermittent light, and choice foot reaction time was measured as response time to the light on the corresponding side of the sensored gait mat. We performed multivariable linear regression to determine associations of pain and foot RT, adjusted for sociodemographic and health characteristics, and serially adjusted for cognitive function (MMSE or Trail Making A). Pain severity and interference were associated with slower simple foot reaction time (P < .05). Pain severity and knee pain were associated with slower choice foot reaction time (P < .05). Adjustment for cognitive measures had little impact on the pain-RT relationship. This significant relationship was only observed among participants with less education. These results support the idea that chronic pain may lead to slower foot RT, thus could represent a fall hazard in older adults. Neuromotor mechanisms underlying the pain-fall relationship warrant further investigation. PERSPECTIVE: This study provides insights on the mechanisms underlying the pain-fall relationship. Chronic pain may contribute to slower foot RT thus increase fall risk in older adults. This may help inform interventions such as stepping training to reduce fall risk in older adults living with chronic pain.

The Feasibility and Acceptability of Virtual Therapy Environments for Early ICU Mobilization

BACKGROUND

Early mobilization in critically ill patients has been documented to reduce neuromuscular complications, but access to rehabilitation services is often limited in the intensive care unit (ICU). Virtual platforms that deliver therapy may increase access to rehabilitation services and improve patient motivation to participate. Xbox Kinect Jintronix software delivers interventions using games and activities designed for therapy.

OBJECTIVE

To determine the feasibility of the Jintronix virtual therapy system in an ICU setting by observing outcomes related to patient satisfaction, adverse events, and technical applicability.

DESIGN

Prospective observational feasibility study.

SETTING

Medical ICU at a Level 1 Trauma Center within an academic medical center.

PARTICIPANTS

Twenty adults admitted to ICU.

INTERVENTION

Participants engaged in one therapy session with a maximum of 14 Jintronix modules targeting arm, leg, and/or trunk strength, range of motion, and endurance.

OUTCOME MEASURES

A trained physician-observer recorded events related to adverse events and technical applicability. Participants completed a survey to determine patient satisfaction.

RESULTS

A total of 188 modules were completed. Participants completed a median of nine modules each. Five participants completed the maximum of 14 modules. Fatigue was the most common reason for cessation. Mean intervention time was 29 minutes. There were no falls, lines dislodged, or medical events. Participants required physical assistance or verbal cues in 36% of modules. Technical errors affected 25% of modules and led to activity cessation in one case. Nearly all participants reported the activity was enjoyable, comfortable, safe, easy to understand, would improve range of motion and strength, and would motivate them to continue.

CONCLUSIONS

Use of a virtual therapy environment in an intensive care setting is feasible.

Validity and Reliability of the Kinect for Assessment of Standardized Transitional Movements and Balance: Systematic Review and Translation into Practice

The validity and reliability of using the Kinect camera to measure standardized assessment of transitional movement, stepping, and balance was systematically reviewed and critically appraised for quality of the methods and results. The study made recommendations of specific tests for practice based on inclusion of both validity and reliability testing as well as quality of results. Authors' willingness to share their software was reported. Translation into practice is limited by lack of redundancy among studies and access to the software to implement the tests.

Automated Finger Chase (ballistic tracking) in the Assessment of Cerebellar Ataxia

A hallmark of cerebellar disease is impaired accuracy of intended movement which is often summarized as ataxia or incoordination. The diagnosis and assessment of cerebellar ataxia (CA) is primarily based on the expert clinician's visual and auditory observations of the performance of these tasks, and as such, a significant level of subjectivity is implied. In order to address the limitations of this subjectivity we designed a novel automated system, utilizing the Microsoft Kinect device, to capture the finger chase task (in the assessment of upper limb ataxia) which is a part of the assessment of cerebellar upper limb function. Capturing the movements of the marker attached on the subject's finger when following the target point generated by the program that mimics the finger movement of the clinician, we were able to capture the disability and provide a novel objective measure of the CA affecting upper limb function. In our approach, we essentially quantified the difference between the intended and achieved trajectories using Dynamic Time Warping (DTW) technique. Further, signal delay times and directional changes of the velocity of the marker were considered in characterizing the disability associated with patient's finger movements. Finally, Principal Component Analysis (PCA) was employed to combine all the relevant features, reduce feature dimension while enhancing the robustness. This analysis demonstrates a significant separation between normal subjects and CA patients, highlighting this approach as a potential diagnostic aid in the objective assessment of Cerebellar ataxia.

Development and Evaluation of a Kinect-based Rapid Movement Therapy Training Platform for Balance Rehabilitation

With the growing aging and overall population, the demand for healthcare professionals and their burden increases by time. Effective balance recovery reaction is required to prevent falls. The aim of this project is to provide low-cost portable balance training system that trains the two important components of effective balance recovery reaction: faster movement completion time (MT) and larger range of motion (ROM). This is done by a Kinect-based interactive rapid movement therapy training platform for reaching and stepping actions. The platform provides real-time feedback to the patient, generates a report for healthcare professionals to monitor the patient's progress, and can be utilized in patient's home or community centers. A pilot study to test the platform was conducted on seventeen stroke patients and it has shown significant improvement in both MT (faster) and ROM (larger).

Trunk motion visual feedback during walking improves dynamic balance in older adults: Assessor blinded randomized controlled trial

BACKGROUND

Virtual reality and augmented feedback have become more prevalent as training methods to improve balance. Few reports exist on the benefits of providing trunk motion visual feedback (VFB) during treadmill walking, and most of those reports only describe within session changes.

RESEARCH QUESTION

To determine whether trunk motion VFB treadmill walking would improve over-ground balance for older adults with self-reported balance problems.

METHODS

40 adults (75.8 years (SD 6.5)) with self-reported balance difficulties or a history of falling were randomized to a control or experimental group. Everyone walked on a treadmill at a comfortable speed 3×/week for 4 weeks in 2 min bouts separated by a seated rest. The control group was instructed to look at a stationary bulls-eye target while the experimental group also saw a moving cursor superimposed on the stationary bulls-eye that represented VFB of their walking trunk motion. The experimental group was instructed to keep the cursor in the center of the bulls-eye. Somatosensory (monofilaments and joint position testing) and vestibular function (canal specific clinical head impulses) was evaluated prior to intervention. Balance and mobility were tested before and after the intervention using Berg Balance Test, BESTest, mini-BESTest, and Six Minute Walk.

RESULTS

There were no significant differences between groups before the intervention. The experimental group significantly improved on the BESTest (p = 0.031) and the mini-BEST (p = 0.019). The control group did not improve significantly on any measure. Individuals with more profound sensory impairments had a larger improvement on dynamic balance subtests of the BESTest.

SIGNIFICANCE

Older adults with self-reported balance problems improve their dynamic balance after training using trunk motion VFB treadmill walking. Individuals with worse sensory function may benefit more from trunk motion VFB during walking than individuals with intact sensory function.

Novel sensing technology in fall risk assessment in older adults: a systematic review

Background

Falls are a major health problem for older adults with significant physical and psychological consequences. A first step of successful fall prevention is to identify those at risk of falling. Recent advancement in sensing technology offers the possibility of objective, low-cost and easy-to-implement fall risk assessment. The objective of this systematic review is to assess the current state of sensing technology on providing objective fall risk assessment in older adults.

Methods

A systematic review was conducted in accordance to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis statement (PRISMA).

Results

Twenty-two studies out of 855 articles were systematically identified and included in this review. Pertinent methodological features (sensing technique, assessment activities, outcome variables, and fall discrimination/prediction models) were extracted from each article. Four major sensing technologies (inertial sensors, video/depth camera, pressure sensing platform and laser sensing) were reported to provide accurate fall risk diagnostic in older adults. Steady state walking, static/dynamic balance, and functional mobility were used as the assessment activity. A diverse range of diagnostic accuracy across studies (47.9% - 100%) were reported, due to variation in measured kinematic/kinetic parameters and modelling techniques.

Conclusions

A wide range of sensor technologies have been utilized in fall risk assessment in older adults. Overall, these devices have the potential to provide an accurate, inexpensive, and easy-to-implement fall risk assessment. However, the variation in measured parameters, assessment tools, sensor sites, movement tasks, and modelling techniques, precludes a firm conclusion on their ability to predict future falls. Future work is needed to determine a clinical meaningful and easy to interpret fall risk diagnosis utilizing sensing technology. Additionally, the gap between functional evaluation and user experience to technology should be addressed.

Using Embedded Sensors in Independent Living to Predict Gait Changes and Falls

This study explored using Big Data, totaling 66 terabytes over 10 years, captured from sensor systems installed in independent living apartments to predict falls from pre-fall changes in residents' Kinect-recorded gait parameters. Over a period of 3 to 48 months, we analyzed gait parameters continuously collected for residents who actually fell ( n = 13) and those who did not fall ( n = 10). We analyzed associations between participants' fall events ( n = 69) and pre-fall changes in in-home gait speed and stride length ( n = 2,070). Preliminary results indicate that a cumulative change in speed over time is associated with the probability of a fall ( p < .0001). The odds of a resident falling within 3 weeks after a cumulative change of 2.54 cm/s is 4.22 times the odds of a resident falling within 3 weeks after no change in in-home gait speed. Results demonstrate using sensors to measure in-home gait parameters associated with the occurrence of future falls.