Essay:

Essay details:

  • Subject area(s): Engineering
  • Price: Free download
  • Published on: 7th September 2019
  • File format: Text
  • Number of pages: 2

Text preview of this essay:

This page is a preview - download the full version of this essay above.

Longitudinal assessment of falls in Parkinson’s disease using inertial sensors and the Timed Up and Go test

Barry R. Greene*1 PhD, Brian Caulfield2 PhD, Dronacharya Lamichhane3 MD, William Bond4,5 MD, Jessica Svendsen4 BA, Connie Zurski4 CNS, Dyveke Pratt3,6 MD

1. Kinesis Health Technologies, Dublin D04 V2N9, Ireland (e-mail: [email protected]).

2. Insight centre and Insight Centre for Data Analytics, University College Dublin, Dublin 4, Ireland

3. OSF Health Care, Illinois Neurological Institute, University of Illinois College of Medicine, Peoria, IL 61605, USA

4. Jump Simulation, OSF HealthCare, Peoria, IL 61605, USA

5. Dept. of Emergency Medicine, University of Illinois College of Medicine at Peoria, Peoria, IL 61605, USA

6. Saint Thomas Rutherford hospitalist services, Murfreesboro, TN 37129, USA

*Corresponding author

Abstract

Objective: To examine the predictive validity for falls of a using a TUG test, quantified using body-worn sensors (QTUG) in people with Parkinson’s Disease (PD). We also sought to examine the inter-session reliability of QTUG sensor measures and their association with the UPDRS.

Approach: 6 month longitudinal study of 15 PD patients. Participants were asked to complete a weekly diary recording any falls activity, for 6 months following baseline assessment. Participants were assessed monthly, using a TUG test, quantified using body-worn sensors.

Main Results: Results suggest that the QTUG FRE recorded at baseline is 73.33% [44.90, 92.21] accurate in predicting falls within 90 days, while the TUG time at baseline was 46.67% [21.27, 73.41] accurate. The TUG time and QTUG FRE were strongly correlated with UPDRS. 52 of 59 inertial sensor parameters exhibited excellent inter-session reliability, five exhibited moderate reliability, while two parameters exhibited poor reliability.

Significance: Results suggest that QTUG is a reliable tool for assessment of gait and mobility in PD and, furthermore that it may have utility in predicting falls in PD patients.

Keywords: Falls, Parkinson’s disease, sensors, reliability

Abbreviations: TUG – Timed Up and Go; QTUG – Quantitative Timed Up and Go; FRE – Falls Risk Estimate; FE - Frailty Estimate; UPDRS – Unified Parkinson’s Disease Rating Scale

Introduction

Parkinson’s disease (PD) is a progressive neurodegenerative disease with an estimated prevalence of 0.3% in industrialised countries (de Lau and Breteler, 2006). Prevalence increases with age to 1% in the over 60s and higher in the over 80s. The costs associated with Parkinson’s disease are significant with costs in the US alone estimated to be $23Bn per year (Huse et al., 2005, Kowal et al., 2013), with costs in the UK reported to be between £449 and £3.3Bn per year (Findley, 2007).

People with Parkinson’s disease are at much higher risk of falls than the general population (Bloem et al., 2004), they are also twice as likely to fall as patients with other neurological conditions (Stolze et al., 2004, Kalilani et al., 2016), with falls occurring more frequently especially when the disease becomes advanced. It has been estimated that a large proportion of PD patients will fall at some point during the course of their disease. Despite this high prevalence, clinicians do not currently have an accurate and reliable means to assess this risk. Current clinical practice suggests the best predictor of a fall in PD patients is the occurrence of a fall in the preceding year (Pickering et al., 2007). However, such an assessment, which could be considered inadequate, relies on participant recall which can be flawed and unreliable, particularly in populations prone to cognitive decline. Additionally, data on historical falls in PD do not provide any information on increased risk of a first falls brought about by disease progression or comorbidities (Nocera et al., 2013) in the intervening period. PD is usually assessed in a clinical environment using clinical scales, such as the Unified Parkinson\'s Disease Rating Scale (UPDRS) (Ramaker et al., 2002) which can be subjective, with significant variation in administration.

The Timed Up and Go test is a standard test of mobility, widely used to screen for gait and balance issues in older adults (Mathias et al., 1986, Podsiadlo and Richardson, 1991, Shumway-Cook et al., 2000). The TUG test is also used to assess balance, mobility and falls risk in Parkinson’s (Nocera et al., 2013). The time to complete the test (TUG time) has been shown to have moderate predictive ability for falls in community dwelling older adults (Barry E et al., 2014, Thrane et al., 2007) and has been shown to be modestly predictive of falls in patients with PD (Nocera et al., 2013). However, the TUG can also be subjective and varies widely in its implementation. The TUG time itself does not provide any indication of specific mobility impairments that can be associated with Parkinson’s.

Previous research has demonstrated that a TUG test quantified with inertial sensors (QTUG) is reliable in measurement of gait and mobility(Smith et al., 2016), as well as its accuracy in assessment of falls in community dwelling older adults (Greene et al., 2016). The utility of the tool in examining gait, mobility and falls risk in Parkinson’s disease has not yet been examined. Several previous studies have also examined the value of an instrumented TUG test in assessment of gait and mobility in Parkinson’s (Salarian et al., 2010, Mariani et al., 2013, Weiss et al., 2010). Similarly, a number of studies have used body-worn sensors to assess falls risk in PD. Weiss et al (Weiss et al., 2014) found that an accelerometer worn for three days on the lower back discriminated PD patients with a history of falls from PD patients with no history of falls. To our knowledge there has not yet been a prospective study on the validity of body-worn sensors for prediction of falls in PD patients. Additionally, to our knowledge the inter-session reliability of body-worn sensor measures obtained during a TUG test has not been examined in PD patients.

This study aimed to examine the utility of a quantified TUG in assessment of gait and mobility in patients with Parkinson’s disease. Specifically, we aimed to examine the association of the falls risk, frailty estimates and TUG time with UPDRS scores, we also aimed to examine the predictive validity of QTUG for falls in Parkinson’s patients, using prospective follow-up data collected using weekly fall diaries. We also aimed to determine the reliability, in this population, of the quantitative gait and mobility measures calculated for each QTUG test.

Data set

We report a single site longitudinal study of Parkinson’s disease patients. A total of 16 participants were recruited from the OSF HealthCare-Illinois Neurological Institute (Peoria, Illinois, USA), Sensor data were not available for one participant, leaving 15 participants for analysis (5 female, mean age 67.3±7.1) Data are summarised in Table 1. Patients were assessed over a 6 month period. QTUG assessments were conducted on a monthly basis, following an initial baseline assessment. A total of 94 QTUG recordings were available for the 15 participants. Participants were evaluated three times using the UPDRS part III; at baseline, 90 days and 180 days. The mean Unified Parkinson’s Disease Rating Scale-Part III (UPDRS) score at baseline was 14.3±9.5.

All patients received a cognitive assessment using the Mini Mental State Examination (MMSE) at baseline and at 180 days, mean MMSE at baseline was 28.42±1.92.

Inclusion criteria: able to provide written informed consent, aged 40 to 80, Idiopathic Parkinson’s disease (meeting UK Brain Bank criteria), responsive to Levodopa for at least four years. MMSE score greater than 22 and able to walk at least 3m independently.

Exclusion criteria: Atypical Parkinsonism, Hoehn and Yahr stage 5, MMSE 21 or less, use of assisted device for ambulation, co-morbidities affecting balance: severe neuropathy, weakness, bilateral hip replacement, syncopal episodes causing falls, diagnosed with lumbar radiculopathy, spinal stenosis or any other back conditions with the potential to affect fall behaviour. Drug abuse or alcoholism, patient unable to provide informed consent.

Ethical approval was received from the Peoria Institutional Review Board.

Methods

Fall data

Participants were asked to complete a weekly diary recording any falls activity, for 6 months following baseline assessment. The diary was collected on a weekly basis by a researcher and collated for later analysis. Each diary captured information about the frequency, timing, location and severity of each fall.

Gait and mobility assessment

The gait and mobility of each participant was assessed, during the Timed Up and Go (TUG) test, on a monthly basis, using an inertial sensor and software system (Kinesis QTUG™, Kinesis Health Technologies, Clonskeagh, Dublin, Ireland). Sensors were placed on each leg, below the knee, while participants completed the TUG test. Each sensor contained a tri-axial gyroscope and a tri-axial accelerometer. Sensor data were streamed via Bluetooth to a tablet computer, for subsequent analysis. The software measures 59 gait and mobility parameters during the TUG test, including the time to complete the test (TUG time), a statistical estimate of the patient’s risk of having a fall, known as the falls risk estimate (FRE) , as well as a statistical estimate of the patient’s frailty level (known as the frailty estimate (FE)) (Greene et al., 2014).

TUG test protocol

The TUG test requires the subject to get up from a chair, walk three metres, turn through 180° at a designated spot, return to the seat and sit back down. The time taken to complete the test was recorded by a researcher using a stopwatch. Subjects were asked to complete the TUG test, ‘as fast as safely possible’. A standard chair with armrests was used. The software timer was started the moment the clinician said ‘go’, and stopped the moment the subject’s back touched the back rest of the chair. Each subject was given time to become familiar with the test and the test was demonstrated to them beforehand.  

Statistical analysis

Linear mixed effects models were used to examine the association of UPDRS with FRE, FE and TUG time. Assessment session and patient were included as random factors. In addition, the correlation of each QTUG measure with UPDRS at baseline was examined using Pearson’s correlation coefficient.

The predictive validity of the QTUG measures for assessing falls risk was calculated using standard metrics. Accuracy (Acc) is defined as the proportion of participants correctly classified by the software as being a ‘faller’ or ‘non-faller’ (a faller is defined as having one or more falls in the follow-up period); sensitivity (Sens) is defined as the proportion of participants labelled as fallers correctly classified by the software as such; specificity (Spec) is defined as the proportion of the non-fallers correctly identified by the software. Positive predictive value (PPV) is defined as the proportion of participants the software classified as fallers, who are correctly classified; negative predictive value (NPV) is the proportion of those participants the classified by the software as non-fallers, who were classified correctly. A binomial proportion confidence interval was used to estimate confidence intervals. A 70% threshold was used for the QTUG FRE and FE, as the cut-off value to identify participants at high risk of falls.

The predictive validity of the TUG time was also calculated, in order to provide a comparator for the QTUG results. A cut-off time of 11.5s was chosen for high risk of falls, based on previously reported research on predicting falls in Parkinson’s Disease (Nocera et al., 2013).

Inter-session reliability across multiple weeks, was examined using ICC(2,k)(Shrout and Fleiss, 1979). An ICC value of 0.7 or greater was considered to demonstrate excellent reliability, while 0.4-0.7 was moderately reliable. ICC values less than 0.4 were considered poor. Reliability statistics were calculated using all available recordings for each participant.

Statistical analyses were performed using Matlab (version 9.1, Mathworks, Natick, VA).

Results

UPDRS results

Decrease of UPDRS total score of greater than or equal to 8 points is considered clinically significant after 6 months (Schrag et al., 2006). Zero of 13 patients demonstrated clinically significant decrease at 6 months. Schulman et al (Shulman et al., 2010) reported a minimal clinically important difference of 4.3 for total UPRDS score, with 9.1 for moderate and 17.1 for large CID. At 90 days three patients exhibited an increase in UPDRS total of more than 4.3, while only two patients exhibited a decrease in UPDRS total score.

At 180 days 7 patients showed an increase of more than 4.3 in total UPDRS score, while only 1 patient showed a decrease in UPRDS total score of more than 4.3.

At baseline, the TUG time was significantly correlated with the QTUG FRE (ρ=0.77, p<0.01), TUG time at baseline was also significantly correlated with QTUG frailty estimate (ρ= 0.68, p<0.01).

The UPDRS score at baseline was significantly correlated with QTUG FRE (ρ=0.60, p<0.05), the baseline UPDRS score was not significantly correlated with the QTUG frailty estimate (ρ=0.41, p=0.13).

A linear mixed effects model, with FRE as a fixed effect, and assessment session and patient ID as random factors demonstrated that FRE was significantly associated with UPDRS across patients and assessments. A linear mixed effects model, with FE as a fixed effect, and assessment session and patient ID as random factors showed that FE was significantly associated with UPDRS across patients and assessments. Similarly, a linear mixed effect model found a significant association between TUG time and UPDRS. In addition, the correlation of the UPDRS baseline score with TUG time at baseline was significant (ρ=0.62, p<0.05).

Falls risk assessment

Weekly falls diaries were available for 15 participants. Complete follow-up data at 180 days were available for 12 of 15 participants, while complete follow-up data at 90 days were available for all 15 participants. At 90 days, 4 of 15 participants had experienced a fall, while at 180 days, 8 of the 12 participants remaining had experienced a fall. A total of 181 falls were recorded, the distribution of falls per participant are detailed in Table 1. The baseline QTUG FRE was 73.33% [44.90, 92.21] accurate in predicting falls within 90 days, while the baseline frailty estimate and TUG time were 60.00% [32.29, 83.67] and 46.67% [21.27, 73.41] accurate respectively.

The baseline FRE was 58.33% [27.67, 84.83] accurate in predicting falls at 180 days, while the baseline frailty estimate and TUG time were 41.67% [15.17, 72.33] and 58.33% [27.67, 84.83] accurate respectively. Detailed performance results for the predictive validity for falls are detailed in Table 2.

Inter-session reliability

The reliability of 59 inertial sensor derived parameters calculated for each TUG test was examined across an average of 6 weekly sessions (see Table 3). The QTUG FRE and frailty estimates demonstrated excellent inter-session reliability (ICC>0.7). The TUG time and all temporal and spatial gait parameters also demonstrated excellent inter-session reliability. Six of eight gait variability parameters demonstrated excellent reliability while the remaining two demonstrated moderate reliability (0.4<ICC<0.7). Three of four gait symmetry parameters demonstrated excellent reliability, while the fourth demonstrated poor reliability (ICC<0.4). For the turn parameters, three of seven demonstrated excellent reliability, three moderate and one poor reliability. All angular velocity parameters demonstrated excellent inter-session reliability.

Discussion

We report a longitudinal study of body-worn sensor data obtained during a TUG test from participants with Parkinson’s disease. We have found that sensor derived measures of falls risk and frailty are strongly associated with UPDRS across multiple assessments. We also report the predictive accuracy of these measures in predicting falls in PD patients, using prospective follow-up data obtained from weekly fall diaries. In addition, we examine the inter-session reliability of sensor-derived measures of gait and mobility, obtained, during the TUG test in patients with PD.

Results suggest that the TUG time, FRE and frailty estimates are strongly correlated with the UPDRS score at baseline, in addition linear mixed effects models showed there is a strong association between TUG time, FRE and frailty estimates with UPDRS scores taken at baseline, 90 days and 180 days.

Analysis of the TUG time, FRE and frailty estimates at 90 days found that the QTUG FRE was markedly more accurate than the TUG time in predicting falls (73.33% compared to 46.67%), Similarly the frailty estimate was more accurate in predicting falls than the TUG time (60.00% compared to 46.67%). Examining follow-up data at 180 days, all metrics demonstrated lower accuracy in predicting falls; the QTUG FRE score and TUG time were equally accurate (58.33%), while the frailty estimate was less accurate (41.67%). The reduction in accuracy in all metrics between 90 and 180 days follow-up may be due to the lower number of patients with complete follow-up data available at 180 days (12 compared to 15 at 90 days) or may be a feature of this population, in that forecasting of falls in PD patients is less accurate over a longer time window. The reported results provide a statistically independent validation of the QTUG FRE and FE for use in prediction of falls in PD.

On average, each study participant was assessed using QTUG seven times; once at baseline and then once a month for 6 months. Examining the inter-session test-retest reliability of the gait and mobility parameters derived from the inertial sensor data from each TUG test, across all 7 sessions found that 52 of 59 parameters exhibited excellent reliability, while five exhibited moderate reliability, two parameters exhibited poor reliability. It is noteworthy that the TUG time, FRE and frailty estimates all exhibited excellent reliability, supporting their use in the longitudinal assessment of PD. Given the nature of the TUG test, the turn strategy (e.g. turn left or right, pivot turn or multiple step) a participant chooses while executing the turn portion of the TUG test, has a major bearing on the reliability of the turn parameters, i.e. if a participant chose to vary their turn strategy between assessments. Turn strategy can also affect the reliability of the calculated gait variability and gait symmetry parameters and could perhaps explain the lower reliability values obtained for some of these parameters. In general, the strong results reported for test-retest reliability suggest that QTUG could be a valuable tool for ongoing assessment of gait and mobility in Parkinson’s Disease.

Study limitations

A major limitation of the present study in the small size of the data set used, the reported associations with UPDRS and results for prediction of falls would need to be confirmed through a much larger population sample. While the sample is very small, the longitudinal study design and use of prospective weekly falls diaries are in line with international best practice.

Conclusions

Results suggest that QTUG may be a reliable tool for longitudinal assessment of gait and mobility in PD and furthermore, it may have utility in predicting falls in PD patients. Future work will seek to validate the present results on a larger data set.

Conflict of Interest and Acknowledgements

Author BRG is a director of Kinesis Health Technologies Ltd, a company with a license to commercialise this technology. All authors contributed to the preparation of the manuscript. Funding for this study was provided by Care Innovations LLC who are a shareholder in Kinesis Health Technologies.

All authors meet the criteria for authorship stated in the Uniform Requirements for Manuscripts Submitted to Biomedical Journals. Each of the authors has read and concurs with the content in the final manuscript. The material within has not been and will not be submitted for publication elsewhere except as an abstract.

We would like to acknowledge the help and support of the staff of the Illinois Neurological Institute, and the participants involved in this study.

References

BARRY E, GALVIN R, KEOGH C, HORGAN F & T., F. 2014. Is the Timed Up and Go test a useful predictor of risk of falls in community dwelling older adults: a systematic review and meta-analysis. BMC Geriatr., 1, 14.

BLOEM, B. R., HAUSDORFF, J. M., VISSER, J. E. & GILADI, N. 2004. Falls and freezing of gait in Parkinson\'s disease: A review of two interconnected, episodic phenomena. Movement Disorders, 19, 871-884.

DE LAU, L. M. L. & BRETELER, M. M. B. 2006. Epidemiology of Parkinson\'s disease. The Lancet Neurology, 5, 525-535.

FINDLEY, L. J. 2007. The economic impact of Parkinson\'s disease. Parkinsonism & Related Disorders, 13, Supplement, S8-S12.

GREENE, B. R., DOHENY, E. P., O’HALLORAN, A. & KENNY, R. A. 2014. Frailty status can be accurately assessed using inertial sensors and the TUG test. Age and Ageing, 43, 406-411.

GREENE, B. R., REDMOND, S. J. & CAULFIELD, B. 2016. Fall risk assessment through automatic combination of clinical fall risk factors and body-worn sensor data. IEEE Journal of Biomedical and Health Informatics, 21, 1-1.

HUSE, D. M., SCHULMAN, K., ORSINI, L., CASTELLI-HALEY, J., KENNEDY, S. & LENHART, G. 2005. Burden of illness in Parkinson\'s disease. Movement Disorders, 20, 1449-1454.

KALILANI, L., ASGHARNEJAD, M., PALOKANGAS, T. & DURGIN, T. 2016. Comparing the Incidence of Falls/Fractures in Parkinson’s Disease Patients in the US Population. PLoS ONE, 11, e0161689.

KOWAL, S. L., DALL, T. M., CHAKRABARTI, R., STORM, M. V. & JAIN, A. 2013. The current and projected economic burden of Parkinson\'s disease in the United States. Movement Disorders, 28, 311-318.

MARIANI, B., JIMÉNEZ, M. C., VINGERHOETS, F. J. G. & AMINIAN, K. 2013. On-Shoe Wearable Sensors for Gait and Turning Assessment of Patients With Parkinson\'s Disease. IEEE Transactions on Biomedical Engineering, 60, 155-158.

MATHIAS, S., NAYAK, U. & ISAACS, B. 1986. Balance in elderly patients: the \"get-up and go\" test. . Arch. Phys. Med. Rehabil., 67, 387-9.

NOCERA, J. R., STEGEMÖLLER, E. L., MALATY, I. A., OKUN, M. S., MARSISKE, M., HASS, C. J. & NATIONAL PARKINSON FOUNDATION QUALITY IMPROVEMENT INITIATIVE, I. 2013. Using the Timed Up & Go Test in a Clinical Setting to Predict Falling in Parkinson\'s Disease. Archives of physical medicine and rehabilitation, 94, 1300-1305.

PICKERING, R. M., GRIMBERGEN, Y. A. M., RIGNEY, U., ASHBURN, A., MAZIBRADA, G., WOOD, B., GRAY, P., KERR, G. & BLOEM, B. R. 2007. A meta-analysis of six prospective studies of falling in Parkinson\'s disease. Movement Disorders, 22, 1892-1900.

PODSIADLO, D. & RICHARDSON, S. 1991. The timed \"Up & Go\": a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc, 39, 142-148.

RAMAKER, C., MARINUS, J., STIGGELBOUT, A. M. & VAN HILTEN, B. J. 2002. Systematic evaluation of rating scales for impairment and disability in Parkinson\'s disease. Movement Disorders, 17, 867-876.

SALARIAN, A., HORAK, F. B., ZAMPIERI, C., CARLSON-KUHTA, P., NUTT, J. G. & AMINIAN, K. 2010. iTUG, a Sensitive and Reliable Measure of Mobility. Neural Systems and Rehabilitation Engineering, IEEE Transactions on, 18, 303-310.

SCHRAG, A., SAMPAIO, C., COUNSELL, N. & POEWE, W. 2006. Minimal clinically important change on the unified Parkinson\'s disease rating scale. Movement Disorders, 21, 1200-1207.

SHROUT, P. E. & FLEISS, J. L. 1979. Intraclass correlations: Uses in assessing rater reliability. Psychological Bulletin, 86, 420-428.

SHULMAN, L. M., GRUBER-BALDINI, A. L., ANDERSON, K. E., FISHMAN, P. S., REICH, S. G. & WEINER, W. J. 2010. The clinically important difference on the unified parkinson\'s disease rating scale. Archives of Neurology, 67, 64-70.

SHUMWAY-COOK, A., BRAUER, S. & WOOLLACOTT, M. 2000. Predicting the probability for falls in community-dwelling older adults using the Timed Up & Go Test. Phys Ther, 80, 896 - 903.

SMITH, E., WALSH, L., DOYLE, J., GREENE, B. & BLAKE, C. 2016. The reliability of the quantitative timed up and go test (QTUG) measured over five consecutive days under single and dual-task conditions in community dwelling older adults. Gait & Posture, 43, 239-244.

STOLZE, H., KLEBE, S., ZECHLIN, C., BAECKER, C., FRIEGE, L. & DEUSCHL, G. 2004. Falls in frequent neurological diseases. Journal of Neurology, 251, 79-84.

THRANE, G., JOAKIMSEN, R. & THORNQUIST, E. 2007. The association between timed up and go test and history of falls: The Tromso study. BMC Geriatrics, 7, 1.

WEISS, A., HERMAN, T., GILADI, N. & HAUSDORFF, J. M. 2014. Objective Assessment of Fall Risk in Parkinson\'s Disease Using a Body-Fixed Sensor Worn for 3 Days. PLoS ONE, 9, e96675.

WEISS, A., HERMAN, T., PLOTNIK, M., BROZGOL, M., MAIDAN, I., GILADI, N., GUREVICH, T. & HAUSDORFF, J. M. 2010. Can an accelerometer enhance the utility of the Timed Up & Go Test when evaluating patients with Parkinson\'s disease? Medical Engineering & Physics, 32, 119-125.

Figures

Figure 1

Figure legends

Figure 1: (a) Variation of FRE and Frailty at baseline with TUG time at baseline (b) variation of FRE and Frailty at baseline with UPDRS score at baseline.

Tables

Table 1: Clinical data for each participant at baseline, as well as number of falls recorded per participant and UPDRS scores at baseline, 90 days and 180 days.

ID Age (yrs) Gender Weight (kg) Height (cm) No. falls UPDRS baseline UPDRS day 90 UPDRS day 180

1 81 M 76.7 177.8 0 33 28

2 65 M 87.1 177.8 0 14 17 15

3 65 M 110.2 188.0 0 13 16 22

4 71 M 88.0 180.3 1 9 22 18

5 67 M 88.5 182.9 0 8 3 4

6 59 M 78.5 165.1 0 1 10 3

7 67 M 88.5 172.7 1 20 15 27

8 75 F 85.7 154.9 0 20 21 30

9 72 F 88.9 172.7 11 11 39

10 56 M 70.3 172.7 1 15 10 10

11 73 M 81.6 175.3 30 14 18 14

12 54 F 65.8 165.1 1 11 15 25

13 67 F 56.7 157.5 6 10 9 24

14 69 M 91.6 185.4 3 9 9 10

15 69 F 45.8 165.1 127 38 28

Table 2: Performance of TUG time, FRE and frailty estimate in predicting fall within 90 and 180 days in PD patients.

90-day follow-up 180-day follow-up

N (fallers/total) 4/15 8/12

FRE Frailty TUG time FRE Frailty - TUG time

Acc (%) 73.33 60.00 46.67 58.33 41.67 58.33

Sens (%) 50.00 25.00 50.00 37.50 25.00 62.50

Spec (%) 81.82 72.73 45.45 100.00 75.00 50.00

PPV (%) 50.00 25.00 25.00 100.00 66.67 71.43

NPV (%) 81.82 72.73 71.43 44.44 33.33 40.00

Table 3:Inter-session reliability measured using intraclass correlation coefficients (ICC(2,k)), with 95% confidence intervals for each inertial sensor derived parameter.

Variable name ICC 95% CI

Falls risk estimate (%) 0.86 (0.71-0.94)

Frailty estimate (%) 0.91 (0.82-0.97)

Temporal gait parameters

TUG test time (s) 0.77 (0.53-0.91)

Time to stand (s) 0.71 (0.41-0.89)

Time to Sit (s) 0.81 (0.61-0.93)

Mean stance time (s) 0.80 (0.59-0.92)

Mean swing time (s) 0.89 (0.77-0.96)

Mean stride time (s) 0.85 (0.70-0.94)

Mean step time (s) 0.85 (0.70-0.94)

Mean double support (%) 0.79 (0.58-0.92)

Mean single support (%) 0.84 (0.68-0.94)

Cadence (steps/min) 0.78 (0.55-0.91)

Number of gait cycles 0.78 (0.55-0.91)

Number of steps 0.75 (0.49-0.90)

Walk time (s) 0.81 (0.61-0.92)

Gait variability parameters

CV stride velocity (%) 0.89 (0.77-0.96)

CV stride length (%) 0.45 (0-0.79)

Swing time variability (%) 0.79 (0.56-0.92)

Double support variability (%) 0.80 (0.59-0.92)

Stance time variability (%) 0.78 (0.56-0.92)

Step time variability (%) 0.47 (0-0.79)

Stride time variability (%) 0.75 (0.49-0.90)

Single support variability (%) 0.74 (0.47-0.90)

Gait symmetry parameters

Step time asymmetry (%) 0.39 (0-0.76)

Swing time asymmetry (%) 0.79 (0.58-0.92)

Stride time asymmetry (%) 0.79 (0.57-0.92)

Stance time asymmetry (%) 0.78 (0.55-0.91)

Spatial gait parameters

Mean stride velocity (cm/s) 0.88 (0.76-0.95)

Mean stride length (cm/s) 0.89 0.78-0.96)

Turn parameters

Return from turn time (s) 0.76 (0.52-0.91)

Turn mid-point time (s) 0.76 (0.52-0.91)

Turning time (s) 0.58 (0.14-0.84)

Turn magnitude (deg/s) 0.51 (0-0.81)

Walk ratio 0.85 (0.69-0.94)

Number of strides in turn 0.30 (0-0.73)

Ratio strides/turning time 0.42 (0-0.77)

Angular velocity parameters

Magnitude range at mid-swing points (deg/s) 0.73 (0.45-0.89)

Min Z-axis ang. vel. x Height (deg.m/s) 0.93 (0.87-0.97)

Max X-axis ang. vel. × Height (deg.m/s) 0.93 (0.87-0.97)

Max X-axis ang. vel. (deg/s) 0.93 (0.85-0.97)

Min Z-axis ang. vel. (deg/s) 0.92 (0.84-0.97)

Mean Y-axis ang. vel. × Height (deg.m/s) 0.91 (0.83-0.97)

Mean Z-axis ang. vel. x Height (deg.m/s) 0.91 (0.83-0.97)

Mean X-axis ang. vel. × Height (deg.m/s) 0.90 (0.80-0.96)

Mean Y-axis ang. vel. (deg/s) 0.90 (0.80-0.96)

Mean Z-axis ang. vel. (deg/s) 0.90 (0.79-0.96)

Max Z-axis ang. vel. (deg/s) 0.84 (0.68-0.94)

Min X-axis ang. vel. × Height (deg.m/s) 0.83 (0.66-0.93)

Mean X-axis ang. vel. (deg/s) 0.89 (0.77-0.96)

Max Y-axis ang. vel. × Height (deg.m/s) 0.87 (0.75-0.95)

Magnitude mean at mid-swing points (deg/s) 0.87 (0.74-0.95)

Max Z-axis ang. vel. x Height (deg.m/s) 0.87 (0.74-0.95)

CV X-axis ang. vel. (%) 0.87 (0.73-0.95)

Max Y-axis ang. vel. (deg/s) 0.85 (0.70-0.94)

CV Y-axis ang. vel. (%) 0.79 (0.58-0.92)

Min Y-axis ang. vel. × Height (deg.m/s) 0.78 (0.56-0.92)

CV Z-axis ang. vel. (%) 0.76 (0.52-0.91)

Min Y-axis ang. vel. (deg/s) 0.76 (0.51-0.91)

Min X-axis ang. vel. (deg/s) 0.81 (0.61-0.92)

...(download the rest of the essay above)

About this essay:

This essay was submitted to us by a student in order to help you with your studies.

If you use part of this page in your own work, you need to provide a citation, as follows:

Essay Sauce, . Available from:< https://www.essaysauce.com/essays/engineering/2017-7-3-1499088454.php > [Accessed 14.10.19].