GEORGE ® : A Pilot Study of a Smartphone Application for Huntington’s Disease

Abstract

Background:

Current Huntington’s disease (HD) measures are limited to subjective, episodic assessments conducted in clinic. Smartphones can enable the collection of objective, real-world data but their use has not been extensively evaluated in HD.

Objective:

Develop and evaluate a smartphone application to assess feasibility of use and key features of HD in clinic and at home.

Methods:

We developed GEORGE^®, an Android smartphone application for HD which assesses voice, chorea, balance, gait, and finger tapping speed. We then conducted an observational pilot study of individuals with manifest HD, prodromal HD, and without a movement disorder. In clinic, participants performed standard clinical assessments and a battery of active tasks in GEORGE. At home, participants were instructed to complete the activities thrice daily for one month. Sensor data were used to measure chorea, tap rate, and step count. Audio data was not analyzed.

Results:

Twenty-three participants (8 manifest HD, 5 prodromal HD, 10 controls) enrolled, and all but one completed the study. On average, participants used the application 2.1 times daily. We observed a significant difference in chorea score (HD: 19.5; prodromal HD: 4.5, p = 0.007; controls: 4.3, p = 0.001) and tap rate (HD: 2.5 taps/s; prodromal HD: 8.9 taps/s, p = 0.001; controls: 8.1 taps/s, p = 0.001) between individuals with and without manifest HD. Tap rate correlated strongly with the traditional UHDRS finger tapping score (left hand: r = –0.82, p = 0.022; right hand: r = –0.79, p = 0.03).

Conclusion:

GEORGE is an acceptable and effective tool to differentiate individuals with and without manifest HD and measure key disease features. Refinement of the application’s interface and activities will improve its usability and sensitivity and, ideally, make it useful for clinical care and research.

Keywords

Huntington’s disease chorea motor function gait smartphone technology

INTRODUCTION

Huntington’s disease (HD) is an inherited neurodegenerative disorder which results in progressive motor, cognitive, and psychiatric symptoms [1, 2]. Chorea and gait dysfunction are prominent motor features of HD. Current measures of HD, including the Unified Huntington Disease Rating Scale (UHDRS) [3], are subjective, limited to in-clinic settings, and suffer from limitations such as inter-rater variability [4]. These rating scales may therefore fail to accurately capture the severity, intra- and inter-day variability, and early signs of motor symptoms.

Sensor technologies, such as video [5], motion capture systems [6], wearables [7], and smartphones [8], have emerged as suitable alternatives to assess motor function. These devices may be employed in the home setting to more frequently and accurately evaluate disease impact on daily life. Smartphones have become increasingly inexpensive, ubiquitous, and advanced and thus present a unique opportunity to characterize disease features. The touchscreen, microphone, and embedded sensors may be used in combination to collect multimodal longitudinal data. Smartphones enable both active data collection, where individuals are directed to perform regularly timed tasks, and continuous, passive monitoring, which occurs in the background. Over the last five years, open-source platforms such as ResearchKit [9] and ResearchStack [10] have enabled medical researchers to design, test, and adopt smartphone applications in clinical studies.

In contrast to Parkinson’s disease [8, 11], few studies have used smartphones to evaluate HD. One study employed smartphones to characterize patient compliance in remote digital trials of movement disorders [12]. Specifically in HD, motion sensors embedded in iPhones have been used for in-clinic gait assessments [13], and a smartphone-based cognitive assessment tool was developed for remote measurement of tapping speed and visual memory performance [14]. To our knowledge, smartphone-embedded sensors have not yet been used to study chorea.

Inspired by the Arora et al. study [8], we developed a smartphone application called GEORGE^® to objectively and continuously monitor HD symptoms both in clinic and at home (Fig. 1). The name of the application is a tribute to Dr. George Huntington, who first described “Huntington’s chorea” in 1872 and recognized it as an inherited disorder [15, 16]. GEORGE uses the smartphone’s sensors to assess chorea, balance and gait, the touchscreen to assess tapping speed, and the microphone to assess voice. In this study, participants used GEORGE for one month, during which time they were asked to complete specified activities daily. Our goal was to assess consistency of use of a smartphone application in individuals with HD as well as their adherence to instructions and feedback on design. We also sought to observe whether GEORGE could differentiate individuals with and without manifest HD, quantify chorea, tap rate, and gait, and shed additional light on the real-world impact of the disease.

Fig. 1

Introductory screen and instructions for finger tapping activity in the GEORGE^® smartphone application.

MATERIALS AND METHODS

Study overview and design

We conducted a 1-month observational study using an Android smartphone application GEORGE, developed at the University of Rochester, of individuals with manifest HD or prodromal HD and in controls without a movement disorder. All procedures were approved by the University of Rochester’s institutional review board. The prodromal and manifest HD study population was recruited from the University of Rochester Movement Disorders Clinic. Most control participants were unaffected family members; some were recruited from local study interest registries. Participants with manifest HD had self-reported clinical diagnoses of HD with a positive family history and clinical features characteristic of the disease. Participants with prodromal HD had undergone confirmatory genetic testing (CAG trinucleotide expansion of at least 36 repeats) but did not have sufficient motor signs to warrant a diagnosis of HD as determined by study investigators.

At the baseline in-clinic visit, participants completed the Montreal Cognitive Assessment (MoCA) [17] and signed the written consent form and a video waiver form. Participants who scored than less than 22 on the MoCA were further screened for comprehension of study activities by an investigator before providing consent. Then, participants were asked about basic demographic data, medical history, including characteristics of their HD, if applicable, concomitant medications, and information regarding their technology use. Participants were shown how to download and register in the GEORGE application, either on a personal smartphone equipped with Android OS and Internet access or on a loaner smartphone meeting these requirements, supplied by the study team. Upon completion of the installation process, participants were video-recorded while performing one set of activities in the application and during traditional clinical assessments, namely the UHDRS assessments [3], Timed Up and Go [18], and Ten-Meter Walk [19].

Following the baseline visit, participants used the GEORGE application daily for one month. They were instructed to complete a set of activities, which took approximately 3–5 minutes, at least three times daily, preferably in the morning, around noontime, and at night. Once daily, participants were asked to answer “Daily Check-in” questions regarding potential symptoms (e.g., fatigue, feelings of hopelessness, and depression). In addition, those who were taking HD medications were asked to log the time and dose of each administration.

After one month, participants returned to the clinic to repeat the UHDRS assessments [3], Timed Up and Go [18], and Ten-Meter Walk [19] and perform a final set of activities in the GEORGE application. These assessments were also video-recorded for time-stamping purposes. At this time, participants either returned the loaner smartphone or were given the option to remove GEORGE from their personal device. To conclude study activities, participants completed a survey which afforded the opportunity to share experiences with the GEORGE application, perceived benefits and limitations of its design, and interest in future similar research. Some participants also completed an optional 15-min user experience interview with a study team member. Survey results, interview responses, and in-clinic video recordings were all used in a user experience evaluation of the accessibility and functionality of GEORGE.

Smartphone application

The GEORGE application included a set of four activities for which the following instructions were provided: 1) chorea activity—hold the phone in the hand with its screen facing upwards for 20 s in four different seated positions: (a) left hand resting on lap, (b) left hand extended out in front at shoulder height, (c) right hand resting on lap, and (d) right hand extended out in front at shoulder height; 2) finger tapping activity—alternately tap your index and middle fingers as many times as possible in 20 s. Complete once with each hand; 3) gait and balance activity—place the phone in your hip pocket and walk continuously for 30 s at your normal pace. After 30 s, turn in a full circle, then stand still with your feet shoulder-width apart and your arms at your side for another 30 s; 4) voice activity—say “Aaaaah” into the microphone for as long as possible. The smartphone data recorded during these activities were encrypted, de-identified, and uploaded via Wi-Fi to a secure database housed at the University of Rochester.

Motor data analysis

Smartphone data collection

Tri-axial accelerometer and gyroscope data were collected from the smartphone-embedded sensors for the chorea and gait and balance activities. The sampling rate varied based on resource availability and ranged from 5 to 50 Hz. For consistent processing, the tri-axial accelerometer data were interpolated to a uniform 50 Hz sampling rate prior to motor symptom analysis. In the finger tapping activity, the total number of taps per hand was recorded by the smartphone touchscreen sensor. Although we collected the audio recordings and responses to “Daily Check-in” questions, we have not analyzed this data.

Chorea analysis

Chorea analysis aimed to quantify irregular, jerky movements that are characteristic of chorea. Tri-axial accelerometer data accumulated during the chorea activity were divided into non-overlapping 2-s intervals. For each interval, a “chorea index” score normalized to 0–100 was estimated by computing the total variation in the band-pass filtered magnitude acceleration data [7]. This score represented the amplitude of the non-deliberate movements in the body observed during the interval.

Tap rate analysis

Tap rate was estimated for each hand by dividing the total number of taps by the duration of the activity (20s).

Gait analysis

Tri-axial accelerometer data obtained during the gait and balance activity were divided into non-overlapping 5-s intervals. The total number of steps per 30-s gait activity was determined by estimating and summing the number of steps in each 5-s interval [20].

Statistical analysis

To evaluate the mean differences in the chorea index, tap rate, and gait per activity between cohorts, we analyzed the data using ANOVA [21], followed by post-hoc tests for multiple pairwise comparisons. The Spearman correlation coefficient [22] was computed to assess the correlation between chorea index and tap rate and the corresponding clinician rated UHDRS scores.

RESULTS

Study population and adherence

Eight participants with manifest HD, 5 participants with prodromal HD, and 10 controls were enrolled in the study. One participant withdrew from the study before completing any remote activities because the application proved incompatible with their smartphone. The baseline characteristics of the participants are summarized in Table 1. The prodromal HD participants were much younger (38.5±12.5 years) than the HD (54.2±7.2 years) and control (57.4±12.7 years) participants. On average, participants used the application 2.1±0.7 times daily throughout the study period (mean duration = 33±3.9 days), with no significant difference in frequency based on disease status.

Table 1

Baseline characteristics of study participants

Characteristic	HD	pHD	Control
	(n = 8)	(n = 5)	(n = 10)
Sex (women), %	50	80	50
Race, % White	100	100	100
Age, y	54.2±7.2	38.5±12.5	57.4±12.7
Currently married, %	25	60	70
College-educated, %	37.5	40	20
Currently employed, %	0	80	70
Montreal Cognitive Assessment score (0-30)	23.0±2.9	28.8±2.2	27.8±1.4
Total Functional Capacity score (0-13)	9.6±2.2	12.4±1.3	13.0±0.0
UHDRS Total Motor score (0-124)	36.6±13.8	0.8±1.1	2.6±3.0
UHDRS Total Maximal Chorea score (0-28)	11.4±3.0	0.0±0.0	0.1±0.3
Timed Up and Go, s	10.0±2.5	8.5±1.6	7.6±1.3
10-Meter Walk Test, s	5.5±1.2	4.7±0.6	4.3±0.5

Chorea analysis

Figure 2 shows the distribution of chorea index in four seated chorea activity positions over the 1-month study period for each cohort. Averaged across all four positions, the mean chorea index of the manifest HD cohort was 19.5, which was significantly higher than the overall mean chorea index of the prodromal HD (4.5, p = 0.007) and control (4.3, p = 0.001) cohorts. The observation of a small but non-zero chorea index value for control participants was expected, as the chorea index is an estimate derived from smartphone sensor data and all individuals display some amount of non-deliberate movement. Figure 3a and 3b show the mean chorea index for the left and right hand, respectively, of individuals with manifest HD at each of the two in-clinic visits. At the baseline visit, the Spearman correlation coefficient between the UHDRS maximal chorea score and the mean chorea index was r = 0.53 (p = 0.216) for the left hand and r = 0.54 (p = 0.213) for the right hand. While a positive correlation between the clinical UHDRS maximal chorea score and the chorea index was observed, this effect was not found to be statistically significant in this small study.

Fig. 2

Box plots showing the distribution of chorea index in four seated chorea activity positions (left hand on lap, left hand extended, right hand on lap, and right hand extended) for Huntington’s disease (HD), prodromal HD (pHD), and control (C) participants over the full duration of the study. Number of participants = 7 HD, 5 prodromal HD, and 10 controls. The median chorea index for the HD participant group was higher than the prodromal HD and the controls for all four positions. Individuals with manifest HD demonstrated the lowest median chorea index while resting the left arm on their lap.

Fig. 3

Mean chorea index for (a) left hand and (b) right hand during baseline and final in-clinic visits and the distribution of chorea index for (c) left hand and (d) right hand over full duration of the study of individual Huntington’s Disease (HD) participants. Text annotations on the figures indicate the Unified Huntington’s Disease Rating Scale maximal chorea score for the corresponding hand assessed during the baseline (B) and final (F) in-clinic visits for each participant. For each hand, the bar plot and the box plot illustrate the average and distribution of chorea index estimates, respectively, from the on lap and hand extended position chorea activities. Number of participants = 7 HD.

Over the course of the 1-month study, considerable variability in the chorea index distribution was observed for both hands within the manifest HD cohort (Fig. 3c, d). For example, the 5^th and the 95^th percentile values for the chorea index for the left hand of one participant (ID –41) were 0.6 and 11.6, respectively, whereas for another participant (ID –48), these same values were 5.0 and 67.4. The chorea index trends observed over full duration were analogous to that observed during the in-clinic chorea activities. The daily distribution of the chorea index for both hands of each manifest HD participant is shown in Supplementary Figure 1. Throughout the study period, no specific trend (e.g., increase or decrease in chorea index) was observed for a manifest HD participant.

Tap rate analysis

Analysis of the finger tapping activity was performed separately for the left and right hands, and the tap rate distribution by cohort is illustrated in Fig. 4. The mean tap rate of manifest HD participants was 2.5 taps/s, which was significantly lower than the tap rate for prodromal HD (8.9 taps/s, p = 0.001) and control (8.1 taps/s, p = 0.001) participants. Figure 5a and 5b show the mean tap rate for the left and right hands, respectively, of individuals with manifest HD at each of the two in-clinic visits. At the baseline and final in-clinic visit, the Spearman rank correlation coefficient between the UHDRS finger tapping score and the mean tap rate for the left and right hand was r = –0.75 (p = 0.048) and r = –0.60 (p = 0.151), and r = –0.82 (p = 0.022) and r = –0.79 (p = 0.03), respectively.

Fig. 4

Box plots showing the distribution of left hand and right hand tap rate during the finger tapping activity for Huntington’s disease (HD), prodromal HD (pHD), and control (C) participants over the full duration of the study. HD participants exhibit the lowest median tap rate and the prodromal HD participants exhibit highest median tap rate. Number of participants = 7 HD, 5 prodromal HD, and 10 controls.

Fig. 5

Mean tap rate for (a) left hand and (b) right hand during the baseline and final in-clinic visit and the distribution of tap rate for (c) left hand and (d) right hand over the full duration of the study of individual Huntington’s disease (HD) participants. Text annotations on the figures indicate the Unified Huntington’s Disease Rating Scale finger tapping scores for the corresponding hand assessed during the baseline (B) and final (F) in-clinic visits for each participant. Number of participants = 7 HD.

For both hands, tap rate distribution exhibited considerable variability over full duration of the study among manifest HD participants (Fig. 5c, d). The 5^th and the 95^th percentile values for the left hand of one participant (ID –48) were 1.1 taps/s and 1.8 taps/s, respectively, whereas for another participant (ID –59), these same values were 2.4 taps/s and 4.6 taps/s. Remote tap rate observations concurred with tap rates derived from in-clinic data. Supplementary Figure 2 shows daily tap rate for the left and right hand of each manifest HD participant. A few participants (IDs –41, 46, and 48) had a consistent tap rate in the range of 1.5 –2 taps/s and exhibited very little intra- and inter-day tap rate variability. The tap rate of another participant (ID –55) increased from 1.5 –2 taps/s at the start of the study to 3.5 –4 taps/s toward its end.

Gait analysis

As shown by the boxplots in Fig. 6, no significant difference in number of steps per gait activity was observed between the cohorts. At the baseline in-clinic visit, the Pearson correlation coefficient between the number of steps per gait activity and the average Ten-Meter walk time for HD participants was r = –0.78 (p = 0.04).

Fig. 6

Box plots showing the distribution of steps per (gait) activity for Huntington’s disease (HD), prodromal HD (pHD), and control (C) participants over full duration of the study. No significant difference in number of steps per activity was observed between cohorts.

Participant feedback

The 22 participants who completed all study activities provided feedback in a usability survey. Most participants (95%) said that they would be either “very willing” (68%) or “somewhat willing” (27%) to use the GEORGE application if asked to do so in the future. In response to a question about the frequency of application use, participants said that it would be feasible to complete a mean 2.3±0.6 sets of activities per day. Twelve participants reported that the main reason why they were unable to complete three sets of activities daily throughout the study period was that other activities required their attention. The most common critique was that the gait and balance activity did not have sufficiently loud cues to know when to walk, turn, and stand still. In addition to survey responses, 21 in-clinic observations of participants using the application and 11 individual participant interviews were qualitatively coded and interpreted as part of a user experience evaluation.

DISCUSSION

In this pilot study, we demonstrated that use of a smartphone application by individuals with HD is acceptable and effective. Study retention was high, and individuals were able to complete the application tasks approximately twice daily, with no significant difference in usage between cohorts. Both in clinic and at home, GEORGE successfully differentiated individuals with and without manifest HD, demonstrating statistically significant differences in objectively measured disease features, including chorea, lack of coordination of fine motor tasks, and gait. The application was well-received among participants and presented an opportunity to apply the chorea index, a novel sensor-derived measure of chorea [7], to analysis of smartphone data.

We observed a significant difference in chorea index and speed of finger tapping between individuals with and without manifest HD. These quantitative measures also revealed substantial variability in motor symptoms for an individual HD participant over time and across all HD participants. This is consistent with wearable sensor data demonstrating variability in chorea severity in individuals with HD [7]. Surprisingly, there was no significant difference in the number of steps between cohorts. This is inconsistent with findings that individuals with HD walk slower and with smaller steps than healthy controls [23, 24]. Among HD participants, we observed a negative correlation between the number of steps per in-clinic gait activity and Ten-Meter walk time, indicating that those who finished the traditional walk more quickly took more steps in the smartphone-based gait activity. Our gait data may have been of poor quality, given that the gait and balance activity was perceived by participants as the most challenging in the GEORGE application. Multiple participants reported that they could not hear the instructional cues and consequently did not start and stop the activity at the appropriate times.

Tap rate as recorded by the touchscreen correlated strongly with the traditional UHDRS finger tapping score. However, while the chorea index appeared to correlate modestly with the UHDRS chorea rating, this association was not statistically significant, potentially due to the scale of this pilot study (7 HD participants) and the design of the chorea activity. Participants seemed able to suppress their upper limb chorea while intently focused on the brief 20-s tasks. We sought to alleviate this effect with an extended hand position, which also minimized the confounding influence of lower limb chorea, but future iterations of the application would benefit from adding a distracting element to the chorea activity (e.g., counting aloud backward from 100).

This study was limited by a small sample size (n = 22 participants) and a short period of observation (1 month). Larger scale assessments are needed to confirm the validity of these smartphone-derived measures and determine how they change over a longer period of time. In order to better characterize disease course, such efforts should focus on the recruitment of individuals with prodromal HD, which proved challenging for this study, and age-matched controls. Our prodromal HD cohort had the lowest mean chorea index and tapped the fastest, likely because their mean age was much lower than that of the other two groups. As demonstrated by Aoki & Fukuoka, maximum rate of finger tapping declines with age [25].

Additionally, the pilot GEORGE application relied on active tasks and primarily measured motor function. Future platforms would benefit from the inclusion of passive monitoring and activities that evaluate behavior, cognition, and speech. More than half of participants reported carrying their phone in either a hip or back pocket throughout everyday life. This may be a convenient placement for the continuous monitoring of gait based on the methodology and evidence presented by Kim et al. for individuals with Parkinson’s disease [26]. Unstructured measures of motor function (e.g., daily step count) could offer insight into the everyday impact of movement disorders. Another area for improvement of GEORGE is the evaluation of non-motor symptoms of HD. Although we collected 10-s audio recordings from participants, these data have yet to be analyzed, and there was no measure of cognitive function. The majority of individuals with HD experience dysarthria [27], and changes in speech production may be one of the earliest indicators of disease onset [28]. Similarly, early changes in cognition have been reported [29]. Thus, voice and cognitive tasks could be helpful in early HD diagnosis and identification of symptom onset, which will be important as new disease modifying therapies emerge.

In order to inform next steps, we conducted a user experience evaluation of GEORGE. This effort uncovered variability in capacity for mobile-specific gestures, such as scrolling and tapping, the labor-intensive and burdensome nature of repeated data entry, and the fundamental role of precise text in bridging usability and functionality. Further, the interviews led us to rethink the optimal timing and number of tasks, as some participants found multiple daily interactions to be overwhelming. While these findings are specific to HD, they have implications for the design of digital tools for other neurodegenerative disorders, demonstrating the value of a user-centered approach to ensure positive engagement.

Overall, our results support the incorporation of digital tools into HD research and care, demonstrating the acceptability and benefit of using a smartphone to measure disease features in clinic and at home. This pilot study provides valuable insights to guide the development of more targeted, user friendly smartphone applications for use in the HD community.

Footnotes

ACKNOWLEDGMENTS

Emma Waddell conducted study visits and contributed to writing the manuscript. Karthik Dinesh developed and performed data analysis and contributed to writing the manuscript. Kelsey Spear conducted study visits and contributed to writing the manuscript. Molly Elson contributed to initial study design and reviewed the manuscript. Ellen Wagner conducted a UX evaluation and contributed to editing the manuscript. Michael Curtis managed data collection and reviewed the manuscript. David Mitten contributed to initial study design and implementation and reviewed the manuscript. Christopher Tarolli conducted study visits and reviewed the manuscript. Gaurav Sharma contributed to data analysis and writing the manuscript. E. Ray Dorsey was the study’s co-principal investigator and contributed to study design and implementation and writing the manuscript. Jamie Adams was the study’s principal investigator, conducted study visits, and contributed to study implementation and writing the manuscript.

CONFLICT OF INTEREST

The authors report no conflicts of interest related to this manuscript. GEORGE^® is registered trademark of the University of Rochester.

The supplementary material is available in the electronic version of this article: .

References

Rothlind

, Bylsma

, Peyser

, Folstein

, Brandt

. Cognitive and motor correlates of everyday functioning in early Huntington’s disease. J Nerv Ment Dis. 1993;181(3):194–9. doi: 10.1097/00005053-199303000-00008

Ross

, Aylward

, Wild

, Langbehn

, Long

, Warner

, et al. Huntington disease: Natural history, biomarkers and prospects for therapeutics. Nat Rev Neurol. 2014;10(4):204–16. doi: 10.1038/nrneurol.2014.24

Huntington Study Group. Unified Huntington’s Disease Rating Scale: Reliability and consistency. Mov Disord. 1996;11(2):136–42. doi: 10.1002/mds.870110204

Dorsey

, Venuto

, Venkataraman

, Harris

, Kieburtz

. Novel methods and technologies for 21st-century clinical trials: A review. JAMA Neurol. 2015;72(5):582–8. doi: 10.1001/jamaneurol.2014.4524

Stack

, Agarwal

, King

, Burnett

, Tahavori

, Janko

, et al. Identifying balance impairments in people with Parkinson’s disease using video and wearable sensors. Gait Posture. 2018;62:321–6. doi: 10.1016/j.gaitpost.2018.03.047

Reilmann

, Schubert

. Motor outcome measures in Huntington disease clinical trials. Handb Clin Neurol. 2017;144:209–25. doi: 10.1016/b978-0-12-801893-4.00018-3

Dinesh

, Snyder

, Xiong

, Tarolli

, Sharma

, Dorsey

, et al. A longitudinal wearable sensor study in Huntington’s disease. J Huntingtons Dis. 2020;9(1):69–81. doi: 10.3233/JHD-190375

Arora

, Venkataraman

, Zhan

, Donohue

, Biglan

, Dorsey

, et al. Detecting and monitoring the symptoms of Parkinson’s disease using smartphones: A pilot study. Parkinsonism Relat Disord. 2015;21(6):650–3. doi: 10.1016/j.parkreldis.2015.02.026

Apple Inc. ResearchKit. http://researchkit.org/. Accessed April 8, 2020.

10.

ResearchStack. http://researchstack.org/. Accessed April 8, 2020.

11.

Zhan

, Mohan

, Tarolli

, Schneider

, Adams

, Sharma

, et al. Using smartphones and machine learning to quantify Parkinson disease severity: The mobile Parkinson disease score. JAMA Neurol. 2018;75(7):876–80. doi: 10.1001/jamaneurol.2018.0809

12.

Cohen

, Waks

, Elm

, Gordon

, Grachev

, Navon-Perry

, et al. Characterizing patient compliance over six months in remote digital trials of Parkinson’s and Huntington disease. BMC Med Inform Decis Mak. 2018;18(1):138. doi: 10.1186/s12911-018-0714-7

13.

Acosta-Escalante

, Beltrán-Naturi

, Boll

, Hernández-Nolasco

, García

. Meta-classifiers in Huntington’s disease patients classification, using iPhone’s movement sensors placed at the ankles. IEEE Access. . 2018;6:30942–57. doi: 10.1109/access.2018.2840327

14.

McLaren

, Andrews

, Glikmann-Johnston

, Mercieca

E-C

, Bellgrove

, Loy

, et al. F29 Validation of a novel smartphone-based cogntive assessment tool designed for remote assessment of cognitive functioning in huntington’s disease. J Neurol Neurosurg Psychiatry. 2018;89:A50.

15.

Bhattacharyya

. The story of George Huntington and his disease. Ann Indian Acad Neurol. 2016;19(1):25–8. doi: 10.4103/0972-2327.175425

16.

Huntington

. On Chorea. Medical and Surgical Reporter. Philadelphia: SW Butler; 1872.

17.

Nasreddine

, Phillips

, Bédirian

, Charbonneau

, Whitehead

, Collin

, et al. The Montreal Cognitive Assessment, MoCA: A brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005;53(4):695–9. doi: 10.1111/j.1532-5415.2005.53221.x

18.

Podsiadlo

, Richardson

. The Timed “Up & Go”: A test of basic functional mobility for frail elderly persons. J Am Geriatr Soc. 1991;39(2):142–8. doi: 10.1111/j.1532-5415.1991.tb01616.x

19.

Duncan

, Combs-Miller

, McNeely

, Leddy

, Cavanaugh

, Dibble

, et al.

Are the average gait speeds during the 10 meter and 6 minute walk tests redundant in Parkinson disease?

Gait Posture. 2017;52:178–82. doi: 10.1016/j.gaitpost.2016.11.033

20.

Dinesh

, Xiong

, Adams

, Dorsey

, Sharma

. Signal analysis for detecting motor symptoms in Parkinson’s and Huntington’s disease using multiple body-affixed sensors: A pilot study. 2016 IEEE Western New York Image and Signal Processing Workshop (WNYISPW). 2016:1-5. doi: 10.1109/WNYIPW.2016.7904834

21.

Girden

. ANOVA: Repeated measures. vol 84. Sage; 1992.

22.

Gibbons

, Chakraborti

. Nonparametric statistical inference. In: Lovric M, editor. International encyclopedia of statistical science. Springer Berlin Heidelberg; 2011. pp. 977-9.

23.

Dalton

, Khalil

, Busse

, Rosser

, van Deursen

, Ólaighin

. Analysis of gait and balance through a single triaxial accelerometer in presymptomatic and symptomatic Huntington’s disease. Gait Posture. 2013;37(1):49–54. doi: 10.1016/j.gaitpost.2012.05.028

24.

Gaßner

, Jensen

, Marxreiter

, Kletsch

, Bohlen

, Schubert

et al. Gait variability as digital biomarker of disease severity in Huntington’s disease. J Neurol. 2020;267(6):1594–601. doi: 10.1007/s00415-020-09725-3

25.

Aoki

, Fukuoka

. Finger tapping ability in healthy elderly and young adults. Med Sci Sports Exerc. 2010;42(3):449–55. doi: 10.1249/MSS.0b013e3181b7f3e1

26.

Kim

, Lee

, Kim

, Jeon

, Park

, et al. Validation of freezing-of-gait monitoring using smartphone. Telemed J E Health. 2018;24(11):899–907. doi: 10.1089/tmj.2017.0215

27.

Diehl

, Mefferd

, Lin

, Sellers

, McDonell

, de Riesthal

, et al. Motor speech patterns in Huntington disease. Neurology. 2019;93(22):e2042–52. doi: 10.1212/wnl.0000000000008541

28.

Vogel

, Shirbin

, Churchyard

, Stout

. Speech acoustic markers of early stage and prodromal Huntington’s disease: A marker of disease onset? Neuropsychologia. 2012;50(14):3273–8. doi: 10.1016/j.neuropsychologia.2012.09.011

29.

Paulsen

, Miller

, Hayes

, Shaw

. Cognitive and behavioral changes in Huntington disease before diagnosis. Handb Clin Neurol. 2017;144:69–91. doi: 10.1016/b978-0-12-801893-4.00006-7