Abstract-The instrumented Timed " Up and Go" test (iTUG) has the potential for playing an important role in providing clinically useful information regarding an individual's balance and mobility that cannot be derived from the original singleoutcome Timed "Up and Go" test protocol. The purpose of this study was to determine the reliability and validity of the iTUG using body-fixed inertial sensors in people affected by stroke. For test-retest reliability analysis, 14 individuals with stroke and 25 nondisabled elderly patients were assessed. For validity analysis, an age-matched comparison of 12 patients with stroke and 12 nondisabled controls was performed. Out of the 14 computed iTUG metrics, the majority showed excellent testretest reliability expressed by high intraclass correlation coefficients (range 0.431-0.994) together with low standard error of measurement and smallest detectable difference values. BlandAltman plots demonstrated good agreement between two repeated measurements. Significant differences between patients with stroke and nondisabled controls were found in 9 of 14 iTUG parameters analyzed. Consequently, these results warrant the future application of the inertial sensor-based iTUG for the assessment of physical deficits poststroke in longitudinal study designs.

BackgroundStroke survivors often suffer from mobility deficits. Current clinical evaluation methods, including questionnaires and motor function tests, cannot provide an objective measure of the patients’ mobility in daily life. Physical activity performance in daily-life can be assessed using unobtrusive monitoring, for example with a single sensor module fixed on the trunk. Existing approaches based on inertial sensors have limited performance, particularly in detecting transitions between different activities and postures, due to the inherent inter-patient variability of kinematic patterns. To overcome these limitations, one possibility is to use additional information from a barometric pressure (BP) sensor.MethodsOur study aims at integrating BP and inertial sensor data into an activity classifier in order to improve the activity (sitting, standing, walking, lying) recognition and the corresponding body elevation (during climbing stairs or when taking an elevator). Taking into account the trunk elevation changes during postural transitions (sit-to-stand, stand-to-sit), we devised an event-driven activity classifier based on fuzzy-logic.Data were acquired from 12 stroke patients with impaired mobility, using a trunk-worn inertial and BP sensor. Events, including walking and lying periods and potential postural transitions, were first extracted. These events were then fed into a double-stage hierarchical Fuzzy Inference System (H-FIS). The first stage processed the events to infer activities and the second stage improved activity recognition by applying behavioral constraints. Finally, the body elevation was estimated using a pattern-enhancing algorithm applied on BP. The patients were videotaped for reference. The performance of the algorithm was estimated using the Correct Classification Rate (CCR) and F-score. The BP-based classification approach was benchmarked against a previously-published fuzzy-logic classifier (FIS-IMU) and a conventional epoch-based classifier (EPOCH).ResultsThe algorithm performance for posture/activity detection, in terms of CCR was 90.4 %, with 3.3 % and 5.6 % improvements against FIS-IMU and EPOCH, respectively. The proposed classifier essentially benefits from a better recognition of standing activity (70.3 % versus 61.5 % [FIS-IMU] and 42.5 % [EPOCH]) with 98.2 % CCR for body elevation estimation.ConclusionThe monitoring and recognition of daily activities in mobility-impaired stoke patients can be significantly improved using a trunk-fixed sensor that integrates BP, inertial sensors, and an event-based activity classifier.

Inertial sensor measurements reliably identify paresis and correlate with clinical measurements; they can therefore provide a complementary dimension of assessment in clinical practice and during clinical trials aimed at improving upper limb function.

FABIEN MASSÉ, Holst Center/imec MARTIEN VAN BUSSEL, Kempenhaeghe/Hobo Heeze B.V. ALINE SERTEYN, TU Eindhoven/Signal Processing Systems JOHAN ARENDS, Kempenhaeghe/Hobo Heeze B.V. JULIEN PENDERS, Holst Center/imecRecent advances in miniaturization of ultra-low power components allow for more intelligent wearable health monitors. The development and evaluation of a wireless wearable electrocardiogram (ECG) monitor to detect epileptic seizures from changes in the cardiac rhythm is described. The ECG data are analyzed by embedded algorithms: a robust beat-detection algorithm combined with a real-time epileptic seizure detector. In its current implementation, the proposed prototype is 52 × 36 × 15mm 3 , and has an autonomy of one day. Based on data collected on the first three epilepsy patients, preliminary clinical results are provided. Wireless, miniaturized and comfortable, this prototype opens new perspectives for health monitoring.

Sit-to-stand and Stand-to-sit transfers (STS) provide relevant information regarding the functional limitation of mobility-impaired patients. The characterization of STS pattern using a single trunk fixed inertial sensor has been proposed as an objective tool to assess changes in functional ability and balance due to disease. Despite significant research efforts, STS quantification remains challenging due to the high inter-and between-subject variability of this motion pattern. The present study aims to improve the performance of STS detection and classification by fusing the information from barometric pressure (BP) and inertial sensors while keeping a single sensor located at the trunk. A total number of 345 STSs were recorded from 12 post-stroke patients monitored in a semi-structured conditioned protocol. Model-based features of BP signal were combined with kinematic parameters from accelerometer and/or gyroscope and used in a logistic regression-based classifier to detect STS and then identify their types. The correct classification rate was 90.6% with full sensor (BP and inertial) configuration and 75.4% with single inertial sensor. Receiver-Operating-Characteristics analysis was carried out to characterize the robustness of the models. The results demonstrate the potential of BP sensor to improve the detection and classification of STSs when monitoring is performed unobtrusively in every-day life.