BACKGROUND: The 6-minute walk test (6MWT) measures exercise capacity in cardiorespiratory, neurological, and musculoskeletal conditions. It consists of observing how far a patient can walk in 6 minutes and is usually performed in a corridor in a clinic. During the COVID-19 pandemic, as health care systems cancelled nonurgent outpatient appointments, many tests were conducted online. At Oxford University Hospitals National Health Service Foundation Trust, patients followed up on by cardiovascular outpatient clinics were asked to use the open-source Timed Walk app to perform the 6MWT in their community as a substitute for the regular tests in the clinic.

OBJECTIVE: This study aimed to assess the clinical usefulness of the app within the context of the pandemic.

METHODS: Consented patients were invited to perform a 6MWT outdoors using the app at least once a month and report the results through periodic telephone calls and visits. Clinical decisions made for the same cohort were registered, with a focus on the effect of the app in supporting decision-making. Data collected through the app during the study period were compared with 6MWTs performed in the prepandemic period.

RESULTS: This study was conducted between October 2021 and December 2022. A total of 55 participants consented (n=25, 45% female; mean age 44.80, SD 17.49 y). In total, 741 events were logged. A total of 51 medical decisions were made for 25 patients; in 41% (21/51) of the decisions, the app played a role, affecting 44% (11/25) of the patients. Between 2018 and 2022, a cohort of 49 patients for whom data were available performed 63 6MWTs in the clinic (18 in 2021), whereas the same patients performed 605 tests using the app in 2022 (ie, October 2021 to December 2022).

CONCLUSIONS: The use of the Timed Walk app for remote 6MWTs allowed clinicians to obtain frequent and objective indications of the status of the patients during the pandemic, compensating for the absence of regular clinic appointments and providing 33 times more tests than in the prepandemic period. These tests supported approximately half of the clinical decisions made regarding the consented patients, showing that the app is useful in clinical practice.

Functional capacity testing is essential for assessing mobility changes, which can impact independence across various populations and health conditions. This study aims to implement instrumented function tests using a combination of affordable sensors, including sensorized mats, sensorized shoes, smartphones, and smartwatches. The goal is to provide objective, reliable, and detailed data on test outcomes, such as gait analysis. We have created a dataset from 6 participants of varying ages, each performing 5 standardized functional tests: Timed Up and Go, 30-Second Chair Rise, Locomo challenge, 10-meter walk, and 40-meter walk. Alongside the dataset, we have developed a tool for visualizing the sensor signals and marking key events to facilitate data analysis. This dataset is intended to support researchers in developing algorithms for extracting test-specific parameters, and for comparing sensors in terms of quality of the signals and ease of setup.

The Timed Up and Go (TUG) test is a widely recognized and standardized mobility test to measure basic mobility and balance capabilities. Despite the possibility to derive rich information about the patient, only the total time to complete the test is conventionally measured by a professional. This work examines the use of non-wearable sensors for the measurement of parameters of the test in an accurate and objective way. The study illustrates a system specifically designed for conducting the TUG test using a set of sensorized mats. The developed system is able to identify the following 4 phases of the test, with relative timestamps: TUG-time, Sit-to-Stand, Mid-Turning, and End-Turning-Stand-to-Sit. Additionally, meaningful parameters for gait assessment are also extracted: walking speed and stride length. Two experimental iterations were conducted to assess the reliability of the developed software. Both iterations involved two different groups of six healthy participants (41.58 ± 13.32 yrs; 6 females, 6 males) performing various walking types. Our results demonstrate that sensorized mats can be used to segment the phases of the test reliably and can additionally be used to quantify gait parameters during the walk phase of the test.

Keeping track of symptoms is a familiar yet often complex task for people living with chronic conditions. In the context of Parkinson's disease, sensor-based technologies are becoming more common to track motor symptoms. These technologies typically rely on passive monitoring but can also be combined with active tests, in which users intentionally perform measuring tasks like finger-tapping or drawing. In this paper, we explore how people with Parkinson's experienced using such active tests through a smartphone app over the course of eight weeks. Drawing on 26 semi-structured interviews, our findings indicate that active tests impact bodily awareness, come with frictions of integration into daily life and may be reframed as motivations for exercises. Speculations on the resulting data suggest that these are partly seen as a useful resource for self-care, but also as a potential cause for anxiety and ambivalence when facing worsening symptoms and decline.

The Timed-Up and Go test is a simple yet effective test used to evaluate balance and mobility in conditions that affect movement, such as Parkinson’s disease. This test can inform clinicians about the monitoring and progression of the disease by measuring the time taken to complete the test. We used a smartphone app to obtain the phone’s inertial data and implemented an algorithm to automatically extract the time taken to complete the test. We considered data collected from six healthy participants performing tests at different speeds. The proposed method was further tested on twelve participants with Parkinson’s disease based on a reference measurement in clinic. We show that, for both groups, we obtain good accuracy (RMSE = 3.42 and 1.95 s) and a strong positive correlation (r = 0.85 and 0.83) between estimated duration and ground truth. We highlight limitations in our approach when the test is performed at very low speed or without a clear pause between the test and the user interaction with the phone. 

The 6-minute walk test is a standardized test used in healthcare to monitor the progress of diseases affecting physical capacity and function. Inertial sensors and positioning data from wearables or smartphones allow to conduct clinical tests in patients' home environments, thereby easing the burden for patients and reducing costs for healthcare. Computation of the 6-minute walked distance requires high accuracy to be clinically useful and current consumer technology-based approaches show that noise and interference in the data often misleads algorithms used to estimate the walk distance. In this research, we are sharing a dataset of inertial and positioning information from 203 walking tests collected with users' own smartphones and the respective ground truth distances. Ground truth is measured with trundle wheels of two types, one which only provides the final distance measurement, and one which provides continuous distance measurements to also capture changes in walking speed. The tests are performed by 19 individuals, both cardiovascular patients and healthy participants. We analyse the dataset using a state-of-the-art algorithm and observe algorithm results in relation to walking features. Based on this, we elaborate on for technology development that may provide further improvements in accuracy for walk distance estimation algorithms, including how data quality and reliability can be assessed.

This paper addresses the need for accurate and continuous measurement of walked distance in applications such as indoor localisation, gait analysis or the 6-minute walk test (6MWT). We propose a method to continuously collect ground truth data of walked distance using an IoT-based trundle wheel. The wheel is connected via Bluetooth Low Energy to a smartphone application which allows the collection of inertial sensor data and GPS location information in addition to the reference distance. We prove the usefulness of this data collection approach in a use case where we derive walked distance from inertial data. We train a 1-dimensional CNN on inertial data collected by one researcher in 15 walking sessions of 1 km length at varying speeds. The training is facilitated by the continuous nature of the reference data. The accuracy of the algorithm is then tested on holdout data of a 6-min duration for which the error of the inferred distance is within clinically significant limits. The proposed approach is useful for the efficient collection of input and reference data for the development of algorithms used to estimate walked distance, such as for the 6MWT. 

As a result of technological advancements, functional capacity assessments, such as the 6-minute walk test, can be performed remotely, at home and in the community. Current studies, however, tend to overlook the crucial aspect of data quality, often limiting their focus to idealised scenarios. Challenging conditions may arise when performing a test given the risk of collecting poor-quality GNSS signal, which can undermine the reliability of the results. This work shows the impact of applying filtering rules to avoid noisy samples in common algorithms that compute the walked distance from positioning data. Then, based on signal features, we assess the reliability of the distance estimation using logistic regression from the following two perspectives: error-based analysis, which relates to the estimated distance error, and user-based analysis, which distinguishes conventional from unconventional tests based on users' previous annotations. We highlight the impact of features associated with walked path irregularity and direction changes to establish data quality. We evaluate features within a binary classification task and reach an F1-score of 0.93 and an area under the curve of 0.97 for the user-based classification. Identifying unreliable tests is helpful to clinicians, who receive the recorded test results accompanied by quality assessments, and to patients, who can be given the opportunity to repeat tests classified as not following the instructions.

Walking speed and distance are usually collected when performing clinical tests such as the 6-Minute Walk Test (6MWT). Wearable devices and smartphones can help bring these tests to the home environment. However, there are difficulties in obtaining measures of distance indoors, where GPS cannot be relied on. Step counting is another even simpler form of data collection that can be obtained through digital technologies. In this work, we investigate the relationship between the step count variable and the standardised 6-Minute Walk Distance (6MWD) variable. By considering 176 6MWTs from 55 participants, we found a high correlation between ground truth distance and the number of steps taken during a test (0.83). Additionally, when considering low-quality outdoor tests, using the step count becomes significantly more reliable (MAE of 22.5m) compared to a state-of-the-art algorithm (MAE of 93.8m). We conclude that step count can be considered as a valid proxy to estimate 6MWD and a candidate approach for monitoring patients’ physical health in free-living conditions.

This work aims to link finger tapping and drawing tests performed on mobile phone screens with clinical ratings of Parkinson's Disease (PD). Thirty PD patients were recruited and instructed to carry out these tests in their homes. Features were extracted and used to assess the validity of the data vis a vis clinical scales (MDS-UPDRS). Statistical tests show that several features correlate with clinical scores (max correlation 0.54) and significant differences between data collected before and after medication intake (p<0.05), demonstrating the clinical validity of smartphone data. The use of Machine Learning (ML) algorithms to regress Part-3 of MDS-UPDRS further supports the validity with an absolute mean error of 6.25 (over a 0-72 scale).