Virtual Reality for advanced cancer patients assisted at home: A randomized controlled interventional study.

OBJECTIVE: Virtual Reality (VR) has been demonstrated to be an effective option for integrating psychological interventions in different therapeutic settings. This randomized controlled interventional study aims to assess the effects of VR, compared to tablet controlled intervention, on anxiety, depression, pain, and short-term psychophysical symptoms in advanced cancer patients assisted at home. METHODS: Participants were provided with a VR headset or a tablet (TAB) for 4 days. On the first and last day, anxiety and depression were measured by Hospital Anxiety and Depression Scale and pain by Brief Pain Inventory. Before and after each VR and tablet session, symptoms were collected by the Edmonton Symptom Assessment Scale (ESAS). RESULTS: Fifty-three patients (27 VR vs. 26 TAB) completed the study. Anxiety significantly decreased in the VR group after the 4-day intervention. The analysis of ESAS showed a significant improvement in pain (p = 0.013), tiredness (p < 0.001), and anxiety (p = 0.013) for TAB group, and a significant reduction in tiredness (p < 0.001) in the VR group. CONCLUSIONS: Technological and user-friendly tools, such as VR and tablets, might be integrated with traditional psychological interventions to improve anxiety and cancer-related short-term symptoms. Further studies are needed to better consolidate the possible beneficial effects of VR.

The body of evidence of late-life depression: the complex relationship between depressive symptoms, movement, dyspnea and cognition.

BACKGROUND: Physical symptoms play an important role in late-life depression and may contribute to residual symptomatology after antidepressant treatment. In this exploratory study, we examined the role of specific bodily dimensions including movement, respiratory functions, fear of falling, cognition, and physical weakness in older people with depression. METHODS: Clinically stable older patients with major depression within a Psychiatric Consultation-Liaison program for Primary Care underwent comprehensive assessment of depressive symptoms, instrumental movement analysis, dyspnea, weakness, activity limitations, cognitive function, and fear of falling. Network analysis was performed to explore the unique adjusted associations between clinical dimensions. RESULTS: Sadness was associated with worse turning and walking ability and movement transitions from walking to sitting, as well as with worse general cognitive abilities. Sadness was also connected with dyspnea, while neurovegetative depressive burden was connected with activity limitations. DISCUSSION: Limitations of motor and cognitive function, dyspnea, and weakness may contribute to the persistence of residual symptoms of late-life depression.

Using Smartphone TechnolOGy to Support an EffecTive Home ExeRcise Intervention to Prevent Falls amongst Community-Dwelling Older Adults: The TOGETHER Feasibility RCT.

INTRODUCTION: Falls have major implications for quality of life, independence, and cost of health services. Strength and balance training has been found to be effective in reducing the rate/risk of falls, as long as there is adequate fidelity to the evidence-based programme. The aims of this study were to (1) assess the feasibility of using the "Motivate Me" and "My Activity Programme" interventions to support falls rehabilitation when delivered in practice and (2) assess study design and trial procedures for the evaluation of the intervention. METHODS: A two-arm pragmatic feasibility randomized controlled trial was conducted with five health service providers in the UK. Patients aged 50+ years eligible for a falls rehabilitation exercise programme from community services were recruited and received either (1) standard service with a smartphone for outcome measurement only or (2) standard service plus the "Motivate Me" and "My Activity Programme" apps. The primary outcome was feasibility of the intervention, study design, and procedures (including recruitment rate, adherence, and dropout). Outcome measures include balance, function, falls, strength, fear of falling, health-related quality of life, resource use, and adherence, measured at baseline, three-month, and six-month post-randomization. Blinded assessors collected the outcome measures. RESULTS: Twenty four patients were randomized to control group and 26 to intervention group, with a mean age of 77.6 (range 62-92) years. We recruited 37.5% of eligible participants across the five clinical sites. 77% in the intervention group completed their full exercise programme (including the use of the app). Response rates for outcome measures at 6 months were 77-80% across outcome measures, but this was affected by the COVID-19 pandemic. There was a mean 2.6 ± 1.9 point difference between groups in change in Berg balance score from baseline to 3 months and mean 4.4 ± 2.7 point difference from baseline to 6 months in favour of the intervention group. Less falls (1.8 ± 2.8 vs. 9.1 ± 32.6) and less injurious falls (0.1 ± 0.5 vs. 0.4 ± 0.6) in the intervention group and higher adherence scores at three (17.7 ± 6.8 vs. 13.1 ± 6.5) and 6 months (15.2 ± 7.8 vs. 14.9 ± 6.1). There were no related adverse events. Health professionals and patients had few technical issues with the apps. CONCLUSIONS: The motivational apps and trial procedures were feasible for health professionals and patients. There are positive indications from outcome measures in the feasibility trial, and key criteria for progression to full trial were met.

Assessing real-world gait with digital technology? Validation, insights and recommendations from the Mobilise-D consortium.

BACKGROUND: Although digital mobility outcomes (DMOs) can be readily calculated from real-world data collected with wearable devices and ad-hoc algorithms, technical validation is still required. The aim of this paper is to comparatively assess and validate DMOs estimated using real-world gait data from six different cohorts, focusing on gait sequence detection, foot initial contact detection (ICD), cadence (CAD) and stride length (SL) estimates. METHODS: Twenty healthy older adults, 20 people with Parkinson's disease, 20 with multiple sclerosis, 19 with proximal femoral fracture, 17 with chronic obstructive pulmonary disease and 12 with congestive heart failure were monitored for 2.5 h in the real-world, using a single wearable device worn on the lower back. A reference system combining inertial modules with distance sensors and pressure insoles was used for comparison of DMOs from the single wearable device. We assessed and validated three algorithms for gait sequence detection, four for ICD, three for CAD and four for SL by concurrently comparing their performances (e.g., accuracy, specificity, sensitivity, absolute and relative errors). Additionally, the effects of walking bout (WB) speed and duration on algorithm performance were investigated. RESULTS: We identified two cohort-specific top performing algorithms for gait sequence detection and CAD, and a single best for ICD and SL. Best gait sequence detection algorithms showed good performances (sensitivity > 0.73, positive predictive values > 0.75, specificity > 0.95, accuracy > 0.94). ICD and CAD algorithms presented excellent results, with sensitivity > 0.79, positive predictive values > 0.89 and relative errors < 11% for ICD and < 8.5% for CAD. The best identified SL algorithm showed lower performances than other DMOs (absolute error < 0.21 m). Lower performances across all DMOs were found for the cohort with most severe gait impairments (proximal femoral fracture). Algorithms' performances were lower for short walking bouts; slower gait speeds (< 0.5 m/s) resulted in reduced performance of the CAD and SL algorithms. CONCLUSIONS: Overall, the identified algorithms enabled a robust estimation of key DMOs. Our findings showed that the choice of algorithm for estimation of gait sequence detection and CAD should be cohort-specific (e.g., slow walkers and with gait impairments). Short walking bout length and slow walking speed worsened algorithms' performances. Trial registration ISRCTN - 12246987.

mCrutch: A Novel m-Health Approach Supporting Continuity of Care.

This paper reports the architecture of a low-cost smart crutches system for mobile health applications. The prototype is based on a set of sensorized crutches connected to a custom Android application. Crutches were instrumented with a 6-axis inertial measurement unit, a uniaxial load cell, WiFi connectivity, and a microcontroller for data collection and processing. Crutch orientation and applied force were calibrated with a motion capture system and a force platform. Data are processed and visualized in real-time on the Android smartphone and are stored on the local memory for further offline analysis. The prototype's architecture is reported along with the post-calibration accuracy for estimating crutch orientation (5° RMSE in dynamic conditions) and applied force (10 N RMSE). The system is a mobile-health platform enabling the design and development of real-time biofeedback applications and continuity of care scenarios, such as telemonitoring and telerehabilitation.

Design and validation of a multi-task, multi-context protocol for real-world gait simulation.

BACKGROUND: Measuring mobility in daily life entails dealing with confounding factors arising from multiple sources, including pathological characteristics, patient specific walking strategies, environment/context, and purpose of the task. The primary aim of this study is to propose and validate a protocol for simulating real-world gait accounting for all these factors within a single set of observations, while ensuring minimisation of participant burden and safety. METHODS: The protocol included eight motor tasks at varying speed, incline/steps, surface, path shape, cognitive demand, and included postures that may abruptly alter the participants' strategy of walking. It was deployed in a convenience sample of 108 participants recruited from six cohorts that included older healthy adults (HA) and participants with potentially altered mobility due to Parkinson's disease (PD), multiple sclerosis (MS), proximal femoral fracture (PFF), chronic obstructive pulmonary disease (COPD) or congestive heart failure (CHF). A novelty introduced in the protocol was the tiered approach to increase difficulty both within the same task (e.g., by allowing use of aids or armrests) and across tasks. RESULTS: The protocol proved to be safe and feasible (all participants could complete it and no adverse events were recorded) and the addition of the more complex tasks allowed a much greater spread in walking speeds to be achieved compared to standard straight walking trials. Furthermore, it allowed a representation of a variety of daily life relevant mobility aspects and can therefore be used for the validation of monitoring devices used in real life. CONCLUSIONS: The protocol allowed for measuring gait in a variety of pathological conditions suggests that it can also be used to detect changes in gait due to, for example, the onset or progression of a disease, or due to therapy. TRIAL REGISTRATION: ISRCTN-12246987.

Wrist Photoplethysmography Signal Quality Assessment for Reliable Heart Rate Estimate and Morphological Analysis.

Photoplethysmographic (PPG) signals are mainly employed for heart rate estimation but are also fascinating candidates in the search for cardiovascular biomarkers. However, their high susceptibility to motion artifacts can lower their morphological quality and, hence, affect the reliability of the extracted information. Low reliability is particularly relevant when signals are recorded in a real-world context, during daily life activities. We aim to develop two classifiers to identify PPG pulses suitable for heart rate estimation (Basic-quality classifier) and morphological analysis (High-quality classifier). We collected wrist PPG data from 31 participants over a 24 h period. We defined four activity ranges based on accelerometer data and randomly selected an equal number of PPG pulses from each range to train and test the classifiers. Independent raters labeled the pulses into three quality levels. Nineteen features, including nine novel features, were extracted from PPG pulses and accelerometer signals. We conducted ten-fold cross-validation on the training set (70%) to optimize hyperparameters of five machine learning algorithms and a neural network, and the remaining 30% was used to test the algorithms. Performances were evaluated using the full features and a reduced set, obtained downstream of feature selection methods. Best performances for both Basic- and High-quality classifiers were achieved using a Support Vector Machine (Acc: 0.96 and 0.97, respectively). Both classifiers outperformed comparable state-of-the-art classifiers. Implementing automatic signal quality assessment methods is essential to improve the reliability of PPG parameters and broaden their applicability in a real-world context.

User-Centered Design Methodologies for the Prototype Development of a Smart Harness and Related System to Provide Haptic Cues to Persons with Parkinson's Disease.

This paper describes the second part of the PASSO (Parkinson smart sensory cues for older users) project, which designs and tests an innovative haptic biofeedback system based on a wireless body sensor network using a smartphone and different smartwatches specifically designed to rehabilitate postural disturbances in persons with Parkinson's disease. According to the scientific literature on the use of smart devices to transmit sensory cues, vibrotactile feedback (particularly on the trunk) seems promising for improving people's gait and posture performance; they have been used in different environments and are well accepted by users. In the PASSO project, we designed and developed a wearable device and a related system to transmit vibrations to a person's body to improve posture and combat impairments like Pisa syndrome and camptocormia. Specifically, this paper describes the methodologies and strategies used to design, develop, and test wearable prototypes and the mHealth system. The results allowed a multidisciplinary comparison among the solutions, which led to prototypes with a high degree of usability, wearability, accessibility, and effectiveness. This mHealth system is now being used in pilot trials with subjects with Parkinson's disease to verify its feasibility among patients.

Classical Machine Learning Versus Deep Learning for the Older Adults Free-Living Activity Classification.

Physical activity has a strong influence on mental and physical health and is essential in healthy ageing and wellbeing for the ever-growing elderly population. Wearable sensors can provide a reliable and economical measure of activities of daily living (ADLs) by capturing movements through, e.g., accelerometers and gyroscopes. This study explores the potential of using classical machine learning and deep learning approaches to classify the most common ADLs: walking, sitting, standing, and lying. We validate the results on the ADAPT dataset, the most detailed dataset to date of inertial sensor data, synchronised with high frame-rate video labelled data recorded in a free-living environment from older adults living independently. The findings suggest that both approaches can accurately classify ADLs, showing high potential in profiling ADL patterns of the elderly population in free-living conditions. In particular, both long short-term memory (LSTM) networks and Support Vector Machines combined with ReliefF feature selection performed equally well, achieving around 97% F-score in profiling ADLs.

The effects of cerebrospinal fluid tap-test on idiopathic normal pressure hydrocephalus: an inertial sensors based assessment.

BACKGROUND: Gait disturbances are typical of persons with idiopathic normal pressure hydrocephalus (iNPH) without signs distinctive from other neurodegenerative and vascular conditions. Cerebrospinal fluid tap-test (CSF-TT) is expected to improve the motor performance of iNPH patients and is a prognostic indicator in their surgical management. This observational prospective study aims to determine which spatio-temporal gait parameter(s), measured during instrumented motor tests, and clinical scale(s) may provide a relevant contribution in the evaluation of motor performance pre vs. post CSF-TT on iNPH patients with and without important vascular encephalopathy. METHODS: Seventy-six patients (20 with an associated vascular encephalopathy) were assessed before, and 24 and 72 h after the CSF-TT by a timed up and go test (TUG) and an 18 m walking test (18 mW) instrumented using inertial sensors. Tinetti Gait, Tinetti Balance, Gait Status Scale, and Grading Scale were fulfilled before and 72 h after the CSF-TT. Stride length, cadence and total time were selected as the outcome measures. Statistical models with mixed effects were implemented to determine the relevant contribution to response variables of each quantitative gait parameter and clinical scales. RESULTS AND CONCLUSION: From baseline to 72 h post CSF-TT patients improved significantly by increasing cadence in 18 mW and TUG (on average of 1.7 and 2.4 strides/min respectively) and stride length in 18 mW (on average of 3.1 cm). A significant reduction of gait apraxia was reflected by modifications in double support duration and in coordination index. Tinetti Gait, Tinetti Balance and Gait Status Scale were able to explain part of the variability of response variables not covered by instrumental data, especially in TUG. Grading Scale revealed the highest affinity with TUG total time and cadence when considering clinical scales alone. Patients with iNPH and an associated vascular encephalopathy showed worst performances compared to pure iNPH but without statistical significance. Gait improvement following CSF-TT was comparable in the two groups. Overall these results suggest that, in order to augment CSF-TT accuracy, is key to assess the gait pattern by analyzing the main spatio-temporal parameters and set post evaluation at 72 h. TRIAL REGISTRATION: Approved by ethics committee: CE 14131 23/02/2015.

Accelerometer-Based Fall Detection Using Machine Learning: Training and Testing on Real-World Falls.

Falling is a significant health problem. Fall detection, to alert for medical attention, has been gaining increasing attention. Still, most of the existing studies use falls simulated in a laboratory environment to test the obtained performance. We analyzed the acceleration signals recorded by an inertial sensor on the lower back during 143 real-world falls (the most extensive collection to date) from the FARSEEING repository. Such data were obtained from continuous real-world monitoring of subjects with a moderate-to-high risk of falling. We designed and tested fall detection algorithms using features inspired by a multiphase fall model and a machine learning approach. The obtained results suggest that algorithms can learn effectively from features extracted from a multiphase fall model, consistently overperforming more conventional features. The most promising method (support vector machines and features from the multiphase fall model) obtained a sensitivity higher than 80%, a false alarm rate per hour of 0.56, and an F-measure of 64.6%. The reported results and methodologies represent an advancement of knowledge on real-world fall detection and suggest useful metrics for characterizing fall detection systems for real-world use.

Simulating the effects of a clinical guidelines screening algorithm for fall risk in community dwelling older adults.

BACKGROUND: The current guidelines for fall prevention in community-dwelling older adults issued by the American Geriatrics Society and British Geriatrics Society (AGS/BGS) indicate an algorithm for identifying who is at increased risk of falling. The predictive accuracy of this algorithm has never been assessed, nor have the consequences that its introduction in clinical practice would bring about. AIMS: To evaluate this risk screening algorithm, estimating its predictive accuracy and its potential impact. METHODS: The analyses are based on 438 community-dwelling older adults, participating in the InCHIANTI study. We analysed different tests for gait and balance assessment. We compared the AGS/BGS algorithm with alternative strategies for fall prevention not based on fall risk evaluation. RESULTS: The AGS/BGS screening algorithm (using TUG, cut-off 13.5 s) has a sensitivity for single falls of 35.8% (95% confidence interval 23.2%-52.7%) and a specificity of 84.0% (79.3%-88.4%). It marks 18.0% (13.7%-22.4%) of the older population as at high risk. A policy of targeting people with preventive intervention regardless of their individual risk could be as effective as the policy based on risk screening but at the price of intervening on 17.3% (4.1%-34.0%) more people of the older population. DISCUSSION: This study is the first that validates and estimates the impact of the screening algorithm of these guidelines. Main limitations are related to some modelling assumptions. CONCLUSIONS: The AGS/BGS screening algorithm has low sensitivity. Nevertheless, its adoption would bring benefits with respect to policies of preventive interventions that act regardless of individual risk assessment.

An Exploratory Factor Analysis of Sensor-Based Physical Capability Assessment.

Physical capability (PC) is conventionally evaluated through performance-based clinical assessments. We aimed to transform a battery of sensor-based functional tests into a clinically applicable assessment tool. We used Exploratory Factor Analysis (EFA) to uncover the underlying latent structure within sensor-based measures obtained in a population-based study. Three hundred four community-dwelling older adults (163 females, 80.9 ± 6.4 years), underwent three functional tests (Quiet Stand, QS, 7-meter Walk, 7MW and Chair Stand, CST) wearing a smartphone at the lower back. Instrumented tests provided 73 sensor-based measures, out of which EFA identified a fifteen-factor model. A priori knowledge and the associations with health-related measures supported the functional interpretation and construct validity analysis of the factors, and provided the basis for developing a conceptual model of PC. For example, the "Walking Impairment" domain obtained from the 7MW test was significantly associated with measures of leg muscle power, gait speed, and overall lower extremity function. To the best of our knowledge, this is the first time that a battery of functional tests, instrumented through a smartphone, is used for outlining a sensor-based conceptual model, which could be suitable for assessing PC in older adults and tracking its changes over time.

Comparison of Standard Clinical and Instrumented Physical Performance Tests in Discriminating Functional Status of High-Functioning People Aged 61⁻70 Years Old.

Assessment of physical performance by standard clinical tests such as the 30-sec Chair Stand (30CST) and the Timed Up and Go (TUG) may allow early detection of functional decline, even in high-functioning populations, and facilitate preventive interventions. Inertial sensors are emerging to obtain instrumented measures that can provide subtle details regarding the quality of the movement while performing such tests. We compared standard clinical with instrumented measures of physical performance in their ability to distinguish between high and very high functional status, stratified by the Late-Life Function and Disability Instrument (LLFDI). We assessed 160 participants from the PreventIT study (66.3 ± 2.4 years, 87 females, median LLFDI 72.31, range: 44.33⁻100) performing the 30CST and TUG while a smartphone was attached to their lower back. The number of 30CST repetitions and the stopwatch-based TUG duration were recorded. Instrumented features were computed from the smartphone embedded inertial sensors. Four logistic regression models were fitted and the Areas Under the Receiver Operating Curve (AUC) were calculated and compared using the DeLong test. Standard clinical and instrumented measures of 30CST both showed equal moderate discriminative ability of 0.68 (95%CI 0.60⁻0.76), p = 0.97. Similarly, for TUG: AUC was 0.68 (95%CI 0.60⁻0.77) and 0.65 (95%CI 0.56⁻0.73), respectively, p = 0.26. In conclusion, both clinical and instrumented measures, recorded through a smartphone, can discriminate early functional decline in healthy adults aged 61⁻70 years.

HABITAT: An IoT Solution for Independent Elderly.

In this work, a flexible and extensive digital platform for Smart Homes is presented, exploiting the most advanced technologies of the Internet of Things, such as Radio Frequency Identification, wearable electronics, Wireless Sensor Networks, and Artificial Intelligence. Thus, the main novelty of the paper is the system-level description of the platform flexibility allowing the interoperability of different smart devices. This research was developed within the framework of the operative project HABITAT (Home Assistance Based on the Internet of Things for the Autonomy of Everybody), aiming at developing smart devices to support elderly people both in their own houses and in retirement homes, and embedding them in everyday life objects, thus reducing the expenses for healthcare due to the lower need for personal assistance, and providing a better life quality to the elderly users.

Reading from the Black Box: What Sensors Tell Us about Resting and Recovery after Real-World Falls.

BACKGROUND: Lying on the floor for a long time after falls, regardless of whether an injury results, remains an unsolved health care problem. In order to develop efficient and acceptable fall detection and reaction approaches, it is relevant to improve the understanding of the circumstances and the characteristics of post-impact responses and the return or failure to return to pre-fall activities. Falls are seldom observed by others; until now, the knowledge about movement kinematics during falls and following impact have been anecdotal. OBJECTIVE: This study aimed to analyse characteristics of the on-ground and recovery phases after real-world falls. The aim was to compare self-recovered falls (defined as returns to standing from the floor) and non-recovered falls with long lies. METHODS AND PARTICIPANTS: Data from subjects in different settings and of different populations with high fall risk were included. Real-world falls collected by inertial sensors worn on the lower back were taken from the FARSEEING database if reliable information was available from fall reports and sensor signals. Trunk pitch angle and acceleration were analysed to describe different patterns of recovery movements while standing up from the floor after the impact of a fall. RESULTS: Falls with successful recovery, where an upright posture was regained, were different from non-recovered falls in terms of resting duration (median 10.5 vs. 34.5 s, p = 0.045). A resting duration longer than 24.5 s (area under the curve = 0.796) after the fall impact was a predictor for the inability to recover to standing. Successful recovery to standing showed lower cumulative angular pitch movement than attempted recovery in fallers that did not return to a standing position (median = 76°, interquartile range 24-170° vs. median = 308°, interquartile range 30-1,209°, p = 0.06). CONCLUSION: Fall signals with and without successful returns to standing showed different patterns during the phase on the ground. Characteristics of real-world falls provided through inertial sensors are relevant to improve the classification and the sensing of falls. The findings are also important for redesigning emergency response processes after falls in order to better support individuals in case of an unrecovered fall. This is crucial for preventing long lies and other fall-related incidents that require an automated fall alarm.

The quality of turning in Parkinson's disease: a compensatory strategy to prevent postural instability?

BACKGROUND: The ability to turn while walking is essential for daily living activities. Turning is slower and more steps are required to complete a turn in people with Parkinson's disease (PD) compared to control subjects but it is unclear whether this altered strategy is pathological or compensatory. The aim of our study is to characterize the dynamics of postural stability during continuous series of turns while walking at various speeds in subjects with PD compared to control subjects. We hypothesize that people with PD slow their turns to compensate for impaired postural stability. METHOD: Motion analysis was used to compare gait kinematics between 12 subjects with PD in their ON state and 19 control subjects while walking continuously on a route composed of short, straight paths interspersed with eleven right and left turns between 30 and 180°. We asked subjects to perform the route at three different speeds: preferred, faster, and slower. Features describing gait spatio-temporal parameters and turning characteristics were extracted from marker trajectories. In addition, to quantify dynamic stability during turns we calculated the distance between the lateral edge of the base of support and the body center of mass, as well as the extrapolated body center of mass. RESULTS: Subjects with PD had slower turns and did not widen the distance between their feet for turning, compared to control subjects. Subjects with PD tended to cut short their turns compared to control subjects, resulting in a shorter walking path. Dynamic stability was smaller in the PD, compared to the healthy group, particularly for fast turning angles of 90°. CONCLUSIONS: The slower turning speeds and larger turning angles in people with PD might reflect a compensatory strategy to prevent dynamic postural instability given their narrow base of support.

Performance Evaluation of State of the Art Systems for Physical Activity Classification of Older Subjects Using Inertial Sensors in a Real Life Scenario: A Benchmark Study.

The popularity of using wearable inertial sensors for physical activity classification has dramatically increased in the last decade due to their versatility, low form factor, and low power requirements. Consequently, various systems have been developed to automatically classify daily life activities. However, the scope and implementation of such systems is limited to laboratory-based investigations. Furthermore, these systems are not directly comparable, due to the large diversity in their design (e.g., number of sensors, placement of sensors, data collection environments, data processing techniques, features set, classifiers, cross-validation methods). Hence, the aim of this study is to propose a fair and unbiased benchmark for the field-based validation of three existing systems, highlighting the gap between laboratory and real-life conditions. For this purpose, three representative state-of-the-art systems are chosen and implemented to classify the physical activities of twenty older subjects (76.4 ± 5.6 years). The performance in classifying four basic activities of daily life (sitting, standing, walking, and lying) is analyzed in controlled and free living conditions. To observe the performance of laboratory-based systems in field-based conditions, we trained the activity classification systems using data recorded in a laboratory environment and tested them in real-life conditions in the field. The findings show that the performance of all systems trained with data in the laboratory setting highly deteriorates when tested in real-life conditions, thus highlighting the need to train and test the classification systems in the real-life setting. Moreover, we tested the sensitivity of chosen systems to window size (from 1 s to 10 s) suggesting that overall accuracy decreases with increasing window size. Finally, to evaluate the impact of the number of sensors on the performance, chosen systems are modified considering only the sensing unit worn at the lower back. The results, similarly to the multi-sensor setup, indicate substantial degradation of the performance when laboratory-trained systems are tested in the real-life setting. This degradation is higher than in the multi-sensor setup. Still, the performance provided by the single-sensor approach, when trained and tested with real data, can be acceptable (with an accuracy above 80%).

Dyskinesia detection and monitoring by a single sensor in patients with Parkinson's disease.

OBJECTIVE: In current clinical practice, assessment of levodopa-induced dyskinesias (LIDs) in Parkinson's disease (PD) is based on semiquantitative scales or patients' diaries. We aimed to assess the feasibility, clinical validity, and usability of a waist-worn inertial sensor for discriminating between LIDs and physiological sway in both supervised and unsupervised settings. METHODS: Forty-six PD patients on L-dopa therapy, 18 de novo PD patients, and 18 healthy controls were enrolled. Patients underwent clinical assessment of motor signs and dyskinesias and kinetic-dynamic L-dopa monitoring, tracked by serial measurements of plasma drug concentrations and motor and postural tests. RESULTS: A subset of features was selected, which showed excellent reliability. Sensitivity and specificity of the selected features for dyskinesia recognition were assessed in both supervised and unsupervised settings with an accuracy of 95% and 86%, respectively. CONCLUSIONS: Our preliminary findings suggest that it is feasible to design a reliable sensor-based application for dyskinesia monitoring at home.