Sinziana Mazilu | Swiss Federal Institute of Technology (ETH) (original) (raw)

Papers by Sinziana Mazilu

Pervasive and Mobile Computing, 2016

Freezing of gait (FoG) is a motor impairment among patients with advanced Parkinson's disease, as... more Freezing of gait (FoG) is a motor impairment among patients with advanced Parkinson's disease, associated with falls and negative impact on patient's quality of life. Detecting such freezes allows real-time gait monitoring to reduce the risk of falls. We investigate the correlation between wrist movements and the freezing of the gait in Parkinsons disease, targeting FoG-detection from wrist-worn sensing data. While most of research focuses on placing inertial sensors on lower limb, i.e., foot, ankle, thigh, we focus on the wrist as an alternative placement. Commonly worn accessories at the wrist such as watches or wristbands are more likely to be accepted and easier to be worn by elderly users, especially subjects with motor problems. Experiments on data from 11 subjects with Parkinson's disease and FoG show there are specific features from wrist movements which are related to gait freeze, such the power on different frequency ranges and statistical information from acceleration and rotation data. Moreover, FoG can be detected by using wrist motion and machine learning models with a FoG hit rate of 0.9, and a specificity between 0.66-0.8. Compared with the state-of-the-art lower limb information used to detect FoG, the wrist increases the number of false detected events, while preserving the FoG hit-rate and detection latency. This suggests that wrist sensors can be a feasible alternative to the cumbersome placement on the legs.

Proceedings of the 10th EAI International Conference on Body Area Networks, 2015

Freezing of gait (FoG) is a motor impairment among patients with advanced Parkinson's disease whi... more Freezing of gait (FoG) is a motor impairment among patients with advanced Parkinson's disease which is associated with falls and has a negative impact on a patient's quality of life. Wearable systems have been developed to detect FoG and to
help patients resume walking by means of rhythmical cueing. A step further is to predict the FoG and start cueing a few seconds before it happens, which might help patients avoid the gait freeze entirely. We characterize the gait parameters continuously with up to 10-12 seconds prior to FoG, observe if and how they change before subjects enter FoG, and compare them with the gait before turns. Moreover, we introduce and discuss new frequency-based features to describe gait and motor anomalies prior to FoG. Using inertial units mounted on the ankles of 5 subjects, we show specific changes in the stride duration and length with up to four seconds prior to FoG on all subjects, compared with turns. Moreover, the dominant frequency migrates towards [3, 8] Hz band with up to six seconds prior to FoG on 3 subjects. These findings open the path to real-time prediction of FoG from inertial measurement units.

The smartphone is the modern equivalent of the Swiss Army knife: It is equipped with an ever-incr... more The smartphone is the modern equivalent of the Swiss Army knife: It is equipped with an ever-increasing collection of processing units, antennas and sensors. A latest addition are the ambient sensors, i.e, temperature, humidity and pressure, which can be used for continuous monitoring of the properties of the environment surrounding the phone. Motivated by the observation that different rooms in a residence tend to have specific ambient properties, we study how informative the phone's ambient data is for indoor localization. Our experiments show that ambient sensors can capture when the user is changing rooms in a residence. Moreover, combinations of ambient sensors allow to distinguish rooms in a home with an accuracy between 0.72 and 0.81, obtained from experiments on 132 hours of data collected from 3 residences.

XRDS: Crossroads, The ACM Magazine for Students, 2014

Wearable computing has the potential to fundamentally alter healthcare by enabling long-term pati... more Wearable computing has the potential to fundamentally alter healthcare by enabling long-term patient monitoring and rehabilitation outside of the lab.

IEEE Journal of Biomedical and Health Informatics, 2015

Freezing of gait (FoG) is a common gait impairment among patients with advanced Parkinson’s disea... more Freezing of gait (FoG) is a common gait impairment among patients with advanced Parkinson’s disease. FoG is associated with falls and negatively impacts the patient’s quality of life. Wearable systems that detect FoG in real time have been developed to help patients resume walking by means of rhythmic cueing. Current methods focus on detection, which require FoG events to happen first, while their prediction opens the road to preemptive cueing, which might help subjects to avoid freeze altogether. We analyzed electrocardiography (ECG) and skinconductance (SC) data from 11 subjects who experience FoG in daily-life, and found statistically significant changes in ECG and SC data just before the FoG episodes, compared to normal walking. Based on these findings, we developed an anomaly-based algorithm for predicting gait freeze from relevant SC features. We were able to predict 71.3% from 184 FoG with an average of 4.2 seconds before a freeze episode happened. Our findings enable the possibility of wearable systems which predict with few seconds before an upcoming FoG from skin conductance, and start external cues to help the user avoid the gait freeze.

The continuous monitoring of a person's location and his interactions with the environment fr... more The continuous monitoring of a person's location and his interactions with the environment from stand-alone and fully wearable sensor setups would enable a number of novel context-aware applications, such as memory assistants for demen-tia patients. In this paper, we introduce our action-observation-based Simultaneous Localization and Mapping (SLAM) system S-SMART, and discuss approaches for combining S-SMART indoor tracking with outdoor positioning via GPS. In particular, we study the re-initialization of indoor tracking when a person revisits an indoor place by means of particle filtering. For evaluation, we used 10 recordings with people walking around and interacting with their environment (total walked distance: 3.8 km). To simulate repeated place visits, we divided each recording in two parts, and introduced start position and heading uncertainty for the second half. We furthermore investigate a real-world recording with a person opening and closing windows in two separate...

Pervasive Computing and Communication Workshops (PerCom Workshops), 2015 IEEE International Conference on , 2015

We investigate the correlation between wrist movement and freezing of the gait in Parkinsons dise... more We investigate the correlation between wrist movement and freezing of the gait in Parkinsons disease. Detecting such freezes allows real-time monitoring to reduce the risk of falls in subjects with Parkinson’s. While most of research focuses on placing inertial sensors on lower limb, i.e., foot, ankle, thigh, lower back, we focus on the wrist as an alternative placement. Commonly worn accessories at the wrist such as watches or wristbands are easier to be accepted and worn by elderly users, in special subjects with motor problems. Experiments on data from 11 subjects show that freezing of gait episodes can be detected using the wrist movements, with a freeze hit-rate of 90% and 83% specificity in a subject-dependent evaluation scheme. This suggests that wrist sensors can be a feasible alternative to the cumbersome placement on the legs.

Proceedings of the 32nd annual ACM conference on Human factors in computing systems - CHI '14, 2014

Patients with Parkinson's disease often experience freezing of gait, which bears a high risk of f... more Patients with Parkinson's disease often experience freezing of gait, which bears a high risk of falling, a prevalent cause for morbidity and mortality. In this work we present GaitAssist, a wearable system for freezing of gait support in daily life. The system provides real-time auditory cueing after the onset of freezing episodes. Furthermore, GaitAssist implements training exercises to learn how to handle freezing situations. GaitAssist is the result of a design process where we considered the input of engineers, clinicians and 18 Parkinson's disease patients, in order to find an optimal trade-off between system wearability and performance. We tested the final system in a user study with 5 additional patients. They reported a reduction in the freezing of gait duration as a result of the auditory stimulation provided, and that they feel the system enhanced their confidence during walking.

ACM Transactions on Interactive Intelligent Systems, 2015

People with Parkinson’s disease (PD) suffer from declining mobility capabilities, which cause... more People with Parkinson’s disease (PD) suffer from declining mobility capabilities, which cause a prevalent risk of falling. Commonly, short periods of motor blocks occur during walking, known as freezing of gait (FoG). To slow the progressive decline of motor abilities, people with PD usually undertake stationary motor-training exercises in the clinics or supervised by physiotherapists. We present a wearable system for the support of people with PD and FoG. The system is designed for independent use. It enables motor training and gait assistance at home and other unsupervised environments. The system consists of three components. First, FoG episodes are detected in real time using wearable inertial sensors and a smartphone as the processing unit. Second, a feedback mechanism triggers a rhythmic auditory signal to the user to alleviate freeze episodes in an assistive mode. Third, the smartphone-based application features support for training exercises. Moreover, the system allows unobtrusive and long-term monitoring of the user’s clinical condition by transmitting sensing data and statistics to a telemedicine service.

We investigate the at-home acceptance of the wearable system in a study with nine PD subjects. Participants deployed and used the system on their own, without any clinical support, at their homes during three protocol sessions in 1 week. Users’ feedback suggests an overall positive attitude toward adopting and using the system in their daily life, indicating that the system supports them in improving their gait. Further, in a data-driven analysis with sensing data from five participants, we study whether there is an observable effect on the gait during use of the system. In three out of five subjects, we observed a decrease in FoG duration distributions over the protocol days during gait-training exercises. Moreover, sensing data-driven analysis shows a decrease in FoG duration and FoG number in four out of five participants when they use the system as a gait-assistive tool during normal daily life activities at home.

Lecture Notes in Computer Science, 2013

Freezing of gait (FoG) is a common gait impairment among patients with advanced Parkinson's disea... more Freezing of gait (FoG) is a common gait impairment among patients with advanced Parkinson's disease. FoG is associated with falls and negatively impact the patient's quality of life. Wearable systems that detect FoG have been developed to help patients resume walking by means of auditory cueing. However, current methods for automated detection are not yet ideal. In this paper, we first compare feature learning approaches based on time-domain and statistical features to unsupervised ones based on principal components analysis. The latter systematically outperforms the former and also the standard in the field -Freezing Index by up to 8.1% in terms of F1-measure for FoG detection. We go a step further by analyzing FoG prediction, i.e., identification of patterns (pre-FoG) occurring before FoG episodes, based only on motion data. Until now this was only attempted using electroencephalography. With respect to the three-class problem (FoG vs. pre-FoG vs. normal locomotion), we show that FoG prediction performance is highly patient-dependent, reaching an F1-measure of 56% in the pre-FoG class for patients who exhibit enough gait degradation before FoG.

International Conference on Indoor Positioning and Indoor Navigation, 2013

We present SmartActionSLAM, an Android smartphone application that performs location tracking in ... more We present SmartActionSLAM, an Android smartphone application that performs location tracking in home and office environments. It uses the integrated motion sensors of the smartphone and an optional foot-mounted inertial measurement unit to track a person. The application implements an instance of the ActionSLAM algorithmic framework. ActionSLAM combines pedestrian dead reckoning with the observation of activities (in SmartActionSLAM: sitting and standing still) to build and update a local landmark map of the user's environment. This map is used to compensate for error accumulation of dead reckoning in a particle filter framework.

2014 IEEE International Conference on Pervasive Computing and Communication Workshops (PERCOM WORKSHOPS), 2014

Many patients with Parkinson's disease suffer from short periods during which they cannot continu... more Many patients with Parkinson's disease suffer from short periods during which they cannot continue walking, the so-called freezing of gait. Patients can learn to use rhythmic auditory sounds as support during these episodes. We developed GaitAssist, a personalized wearable system for freezing of gait support, that enables training in unsupervised environments. GaitAssist detects freezing episodes from ankle-mounted motion sensors, which stream data via Bluetooth to an Android phone. In response, the system plays a rhythmic auditory sound that adapts to the patient's regular gait speed. While GaitAssist can be used as a daily-life assistant, it also provides support for three types of training and rehabilitation exercises. The user can create personalized training sessions by adjusting the exercise and feedback parameters.

Proceedings of the 4th Augmented Human International Conference on - AH '13, 2013

Many people with Parkinson's disease suffer from freezing of gait, a debilitating temporary inabi... more Many people with Parkinson's disease suffer from freezing of gait, a debilitating temporary inability to pursue walking. Rehabilitation with wearable technology is promising. State of the art approaches face difficulties in providing the needed bio-feedback with a sufficient low-latency and high accuracy, as they rely solely on the crude analysis of movement patterns allowed by commercial motion sensors. Yet the medical literature hints at more sophisticated approaches. In this work we present our first step to address this with a rich multimodal approach combining physical and physiological sensors. We present the experimental recordings including 35 motion and 3 physiological sensors we conducted on 18 patients, collecting 23 hours of data. We provide best practices to ensure a robust data collection that considers real requirements for real world patients. To this end we show evidence from a user questionnaire that the system is low-invasive and that a multimodal view can leverage cross modal correlations for detection or even prediction of gait freeze episodes.

Lecture Notes in Computer Science, 2013

Extracting semantic meaning of locations enables a large range of applications including automati... more Extracting semantic meaning of locations enables a large range of applications including automatic daily activity logging, assisted living for elderly, as well as the adaptation of phone user profiles according to user needs. Traditional location recognition approaches often rely on power-hungry sensor modalities such as GPS, network localization or audio to identify semantic locations, e.g., at home, or in a shop. To enable a continuous observation with minimal impact on power consumption, we propose to use low-power ambient sensors -pressure, temperature, humidity and light -integrated in phones. Ambient fingerprints allow the recognition of routinely visited places without requiring traditional localization sensing modalities. We show the feasibility of our approach on 250 hours of data collected in realistic settings by five users during their daily transition patterns, in the course of 49 days. To this end, we employ a prototype smartphone with integrated humidity and temperature sensor. We achieve up to 80% accuracy for recognition of five location categories in a user-specific setting, while saving up to 85% of the battery power consumed by traditional sensing modalities.

2011 10th International Conference on Machine Learning and Applications and Workshops, 2011

In this paper we study the problem of building document classifiers using labeled features and un... more In this paper we study the problem of building document classifiers using labeled features and unlabeled documents, where not all the features are helpful for the process of learning. This is an important setting, since building classifiers using labeled words has been recently shown to require considerably less human labeling effort than building classifiers using labeled documents. We propose the use of Generalized Expectation (GE) criteria combined with a L1 regularization term for learning from labeled features. This lets the feature labels guide model expectation constraints, while approaching feature selection from a regularization perspective. We show that GE criteria combined with L1 regularization consistently outperforms - up to 12% increase in accuracy - the best previously reported results in the literature under the same setting, obtained using L2 regularization. Furthermore, the results obtained with GE criteria and L1 regularizer are competitive to those obtained in the traditional instance-labeling setting, with the same labeling cost.

Proceedings of the 6th International Conference on Pervasive Computing Technologies for Healthcare, 2012

Freezing of gait (FoG) is a common gait deficit in advanced Parkinson's disease (PD). FoG events ... more Freezing of gait (FoG) is a common gait deficit in advanced Parkinson's disease (PD). FoG events are associated with falls, interfere with daily life activities and impair quality of life. FoG is often resistant to pharmacologic treatment; therefore effective non-pharmacologic assistance is needed. We propose a wearable assistant, composed of a smartphone and wearable accelerometers, for online detection of FoG. The system is based on machine learning techniques for automatic detection of FoG episodes. When FoG is detected, the assistant provides rhythmic auditory cueing or vibrotactile feedback that stimulates the patient to resume walking. We tested our solution on more than 8h of recorded lab data from PD patients that experience FoG in daily life. We characterize the system performance on user-dependent and user-independent experiments, with respect to different machine learning algorithms, sensor placement and preprocessing window size. The final system was able to detect FoG events with an average sensitivity and specificity of more than 95%, and mean detection latency of 0.34s in user-dependent settings.

2011 International Conference on P2P, Parallel, Grid, Cloud and Internet Computing, 2011

Vehicular ad-hoc networks (VANETs) have a great potential to improve road safety, traffic jam... more Vehicular ad-hoc networks (VANETs) have a great potential to improve road safety, traffic jams, fuel consumption, and to increase passenger convenience in vehicles. However, VANETs use an open medium for communication and, therefore, are exposed to security threats that influence their reliability. We propose a data-trust security model designed for VANETs based on social network theories. Drivers receiving data about traffic congestion or safety warnings can use the model to evaluate the trust in the received information. The model computes a trust index for each message based on the relevance of the event. It also uses a gossiping approach to disseminate data-trust indexes between vehicles, increasing the accuracy in the trustworthiness of an event and assuring the privacy by hiding the original event sources. The approach is evaluated through modeling and simulation, and we present results that proof its validity.

2012 International Conference on Indoor Positioning and Indoor Navigation (IPIN), 2012

Indoor localization at minimal deployment effort and with low costs is relevant for many ambient ... more Indoor localization at minimal deployment effort and with low costs is relevant for many ambient intelligence and mobile computing applications. This paper presents Ac-tionSLAM, a novel approach to Simultaneous Localization And Mapping (SLAM) for pedestrian indoor tracking that makes use of body-mounted sensors. ActionSLAM iteratively builds a map of the environment and localizes the user within this map. A footmounted Inertial Measurement Unit (IMU) keeps track of the user's path, while observations of location-related actions (e.g. door-opening or sitting on a chair) are used to compensate for drift error accumulation in a particle filter framework. Locationrelated actions are recognizable from body-mounted IMUs that are often used in ambient-assisted living scenarios for context awareness. Thus localization relies only on on-body sensing and requires no ambient infrastructure such as Wi-Fi access points or radio beacons.

Pervasive and Mobile Computing, 2016

Freezing of gait (FoG) is a motor impairment among patients with advanced Parkinson's disease, as... more Freezing of gait (FoG) is a motor impairment among patients with advanced Parkinson's disease, associated with falls and negative impact on patient's quality of life. Detecting such freezes allows real-time gait monitoring to reduce the risk of falls. We investigate the correlation between wrist movements and the freezing of the gait in Parkinsons disease, targeting FoG-detection from wrist-worn sensing data. While most of research focuses on placing inertial sensors on lower limb, i.e., foot, ankle, thigh, we focus on the wrist as an alternative placement. Commonly worn accessories at the wrist such as watches or wristbands are more likely to be accepted and easier to be worn by elderly users, especially subjects with motor problems. Experiments on data from 11 subjects with Parkinson's disease and FoG show there are specific features from wrist movements which are related to gait freeze, such the power on different frequency ranges and statistical information from acceleration and rotation data. Moreover, FoG can be detected by using wrist motion and machine learning models with a FoG hit rate of 0.9, and a specificity between 0.66-0.8. Compared with the state-of-the-art lower limb information used to detect FoG, the wrist increases the number of false detected events, while preserving the FoG hit-rate and detection latency. This suggests that wrist sensors can be a feasible alternative to the cumbersome placement on the legs.

Proceedings of the 10th EAI International Conference on Body Area Networks, 2015

Freezing of gait (FoG) is a motor impairment among patients with advanced Parkinson's disease whi... more Freezing of gait (FoG) is a motor impairment among patients with advanced Parkinson's disease which is associated with falls and has a negative impact on a patient's quality of life. Wearable systems have been developed to detect FoG and to
help patients resume walking by means of rhythmical cueing. A step further is to predict the FoG and start cueing a few seconds before it happens, which might help patients avoid the gait freeze entirely. We characterize the gait parameters continuously with up to 10-12 seconds prior to FoG, observe if and how they change before subjects enter FoG, and compare them with the gait before turns. Moreover, we introduce and discuss new frequency-based features to describe gait and motor anomalies prior to FoG. Using inertial units mounted on the ankles of 5 subjects, we show specific changes in the stride duration and length with up to four seconds prior to FoG on all subjects, compared with turns. Moreover, the dominant frequency migrates towards [3, 8] Hz band with up to six seconds prior to FoG on 3 subjects. These findings open the path to real-time prediction of FoG from inertial measurement units.

The smartphone is the modern equivalent of the Swiss Army knife: It is equipped with an ever-incr... more The smartphone is the modern equivalent of the Swiss Army knife: It is equipped with an ever-increasing collection of processing units, antennas and sensors. A latest addition are the ambient sensors, i.e, temperature, humidity and pressure, which can be used for continuous monitoring of the properties of the environment surrounding the phone. Motivated by the observation that different rooms in a residence tend to have specific ambient properties, we study how informative the phone's ambient data is for indoor localization. Our experiments show that ambient sensors can capture when the user is changing rooms in a residence. Moreover, combinations of ambient sensors allow to distinguish rooms in a home with an accuracy between 0.72 and 0.81, obtained from experiments on 132 hours of data collected from 3 residences.

XRDS: Crossroads, The ACM Magazine for Students, 2014

Wearable computing has the potential to fundamentally alter healthcare by enabling long-term pati... more Wearable computing has the potential to fundamentally alter healthcare by enabling long-term patient monitoring and rehabilitation outside of the lab.

IEEE Journal of Biomedical and Health Informatics, 2015

Freezing of gait (FoG) is a common gait impairment among patients with advanced Parkinson’s disea... more Freezing of gait (FoG) is a common gait impairment among patients with advanced Parkinson’s disease. FoG is associated with falls and negatively impacts the patient’s quality of life. Wearable systems that detect FoG in real time have been developed to help patients resume walking by means of rhythmic cueing. Current methods focus on detection, which require FoG events to happen first, while their prediction opens the road to preemptive cueing, which might help subjects to avoid freeze altogether. We analyzed electrocardiography (ECG) and skinconductance (SC) data from 11 subjects who experience FoG in daily-life, and found statistically significant changes in ECG and SC data just before the FoG episodes, compared to normal walking. Based on these findings, we developed an anomaly-based algorithm for predicting gait freeze from relevant SC features. We were able to predict 71.3% from 184 FoG with an average of 4.2 seconds before a freeze episode happened. Our findings enable the possibility of wearable systems which predict with few seconds before an upcoming FoG from skin conductance, and start external cues to help the user avoid the gait freeze.

The continuous monitoring of a person's location and his interactions with the environment fr... more The continuous monitoring of a person's location and his interactions with the environment from stand-alone and fully wearable sensor setups would enable a number of novel context-aware applications, such as memory assistants for demen-tia patients. In this paper, we introduce our action-observation-based Simultaneous Localization and Mapping (SLAM) system S-SMART, and discuss approaches for combining S-SMART indoor tracking with outdoor positioning via GPS. In particular, we study the re-initialization of indoor tracking when a person revisits an indoor place by means of particle filtering. For evaluation, we used 10 recordings with people walking around and interacting with their environment (total walked distance: 3.8 km). To simulate repeated place visits, we divided each recording in two parts, and introduced start position and heading uncertainty for the second half. We furthermore investigate a real-world recording with a person opening and closing windows in two separate...

Pervasive Computing and Communication Workshops (PerCom Workshops), 2015 IEEE International Conference on , 2015

We investigate the correlation between wrist movement and freezing of the gait in Parkinsons dise... more We investigate the correlation between wrist movement and freezing of the gait in Parkinsons disease. Detecting such freezes allows real-time monitoring to reduce the risk of falls in subjects with Parkinson’s. While most of research focuses on placing inertial sensors on lower limb, i.e., foot, ankle, thigh, lower back, we focus on the wrist as an alternative placement. Commonly worn accessories at the wrist such as watches or wristbands are easier to be accepted and worn by elderly users, in special subjects with motor problems. Experiments on data from 11 subjects show that freezing of gait episodes can be detected using the wrist movements, with a freeze hit-rate of 90% and 83% specificity in a subject-dependent evaluation scheme. This suggests that wrist sensors can be a feasible alternative to the cumbersome placement on the legs.

Proceedings of the 32nd annual ACM conference on Human factors in computing systems - CHI '14, 2014

Patients with Parkinson's disease often experience freezing of gait, which bears a high risk of f... more Patients with Parkinson's disease often experience freezing of gait, which bears a high risk of falling, a prevalent cause for morbidity and mortality. In this work we present GaitAssist, a wearable system for freezing of gait support in daily life. The system provides real-time auditory cueing after the onset of freezing episodes. Furthermore, GaitAssist implements training exercises to learn how to handle freezing situations. GaitAssist is the result of a design process where we considered the input of engineers, clinicians and 18 Parkinson's disease patients, in order to find an optimal trade-off between system wearability and performance. We tested the final system in a user study with 5 additional patients. They reported a reduction in the freezing of gait duration as a result of the auditory stimulation provided, and that they feel the system enhanced their confidence during walking.

ACM Transactions on Interactive Intelligent Systems, 2015

People with Parkinson’s disease (PD) suffer from declining mobility capabilities, which cause... more People with Parkinson’s disease (PD) suffer from declining mobility capabilities, which cause a prevalent risk of falling. Commonly, short periods of motor blocks occur during walking, known as freezing of gait (FoG). To slow the progressive decline of motor abilities, people with PD usually undertake stationary motor-training exercises in the clinics or supervised by physiotherapists. We present a wearable system for the support of people with PD and FoG. The system is designed for independent use. It enables motor training and gait assistance at home and other unsupervised environments. The system consists of three components. First, FoG episodes are detected in real time using wearable inertial sensors and a smartphone as the processing unit. Second, a feedback mechanism triggers a rhythmic auditory signal to the user to alleviate freeze episodes in an assistive mode. Third, the smartphone-based application features support for training exercises. Moreover, the system allows unobtrusive and long-term monitoring of the user’s clinical condition by transmitting sensing data and statistics to a telemedicine service.

We investigate the at-home acceptance of the wearable system in a study with nine PD subjects. Participants deployed and used the system on their own, without any clinical support, at their homes during three protocol sessions in 1 week. Users’ feedback suggests an overall positive attitude toward adopting and using the system in their daily life, indicating that the system supports them in improving their gait. Further, in a data-driven analysis with sensing data from five participants, we study whether there is an observable effect on the gait during use of the system. In three out of five subjects, we observed a decrease in FoG duration distributions over the protocol days during gait-training exercises. Moreover, sensing data-driven analysis shows a decrease in FoG duration and FoG number in four out of five participants when they use the system as a gait-assistive tool during normal daily life activities at home.

Lecture Notes in Computer Science, 2013

Freezing of gait (FoG) is a common gait impairment among patients with advanced Parkinson's disea... more Freezing of gait (FoG) is a common gait impairment among patients with advanced Parkinson's disease. FoG is associated with falls and negatively impact the patient's quality of life. Wearable systems that detect FoG have been developed to help patients resume walking by means of auditory cueing. However, current methods for automated detection are not yet ideal. In this paper, we first compare feature learning approaches based on time-domain and statistical features to unsupervised ones based on principal components analysis. The latter systematically outperforms the former and also the standard in the field -Freezing Index by up to 8.1% in terms of F1-measure for FoG detection. We go a step further by analyzing FoG prediction, i.e., identification of patterns (pre-FoG) occurring before FoG episodes, based only on motion data. Until now this was only attempted using electroencephalography. With respect to the three-class problem (FoG vs. pre-FoG vs. normal locomotion), we show that FoG prediction performance is highly patient-dependent, reaching an F1-measure of 56% in the pre-FoG class for patients who exhibit enough gait degradation before FoG.

International Conference on Indoor Positioning and Indoor Navigation, 2013

We present SmartActionSLAM, an Android smartphone application that performs location tracking in ... more We present SmartActionSLAM, an Android smartphone application that performs location tracking in home and office environments. It uses the integrated motion sensors of the smartphone and an optional foot-mounted inertial measurement unit to track a person. The application implements an instance of the ActionSLAM algorithmic framework. ActionSLAM combines pedestrian dead reckoning with the observation of activities (in SmartActionSLAM: sitting and standing still) to build and update a local landmark map of the user's environment. This map is used to compensate for error accumulation of dead reckoning in a particle filter framework.

2014 IEEE International Conference on Pervasive Computing and Communication Workshops (PERCOM WORKSHOPS), 2014

Many patients with Parkinson's disease suffer from short periods during which they cannot continu... more Many patients with Parkinson's disease suffer from short periods during which they cannot continue walking, the so-called freezing of gait. Patients can learn to use rhythmic auditory sounds as support during these episodes. We developed GaitAssist, a personalized wearable system for freezing of gait support, that enables training in unsupervised environments. GaitAssist detects freezing episodes from ankle-mounted motion sensors, which stream data via Bluetooth to an Android phone. In response, the system plays a rhythmic auditory sound that adapts to the patient's regular gait speed. While GaitAssist can be used as a daily-life assistant, it also provides support for three types of training and rehabilitation exercises. The user can create personalized training sessions by adjusting the exercise and feedback parameters.

Proceedings of the 4th Augmented Human International Conference on - AH '13, 2013

Many people with Parkinson's disease suffer from freezing of gait, a debilitating temporary inabi... more Many people with Parkinson's disease suffer from freezing of gait, a debilitating temporary inability to pursue walking. Rehabilitation with wearable technology is promising. State of the art approaches face difficulties in providing the needed bio-feedback with a sufficient low-latency and high accuracy, as they rely solely on the crude analysis of movement patterns allowed by commercial motion sensors. Yet the medical literature hints at more sophisticated approaches. In this work we present our first step to address this with a rich multimodal approach combining physical and physiological sensors. We present the experimental recordings including 35 motion and 3 physiological sensors we conducted on 18 patients, collecting 23 hours of data. We provide best practices to ensure a robust data collection that considers real requirements for real world patients. To this end we show evidence from a user questionnaire that the system is low-invasive and that a multimodal view can leverage cross modal correlations for detection or even prediction of gait freeze episodes.

Lecture Notes in Computer Science, 2013

Extracting semantic meaning of locations enables a large range of applications including automati... more Extracting semantic meaning of locations enables a large range of applications including automatic daily activity logging, assisted living for elderly, as well as the adaptation of phone user profiles according to user needs. Traditional location recognition approaches often rely on power-hungry sensor modalities such as GPS, network localization or audio to identify semantic locations, e.g., at home, or in a shop. To enable a continuous observation with minimal impact on power consumption, we propose to use low-power ambient sensors -pressure, temperature, humidity and light -integrated in phones. Ambient fingerprints allow the recognition of routinely visited places without requiring traditional localization sensing modalities. We show the feasibility of our approach on 250 hours of data collected in realistic settings by five users during their daily transition patterns, in the course of 49 days. To this end, we employ a prototype smartphone with integrated humidity and temperature sensor. We achieve up to 80% accuracy for recognition of five location categories in a user-specific setting, while saving up to 85% of the battery power consumed by traditional sensing modalities.

2011 10th International Conference on Machine Learning and Applications and Workshops, 2011

In this paper we study the problem of building document classifiers using labeled features and un... more In this paper we study the problem of building document classifiers using labeled features and unlabeled documents, where not all the features are helpful for the process of learning. This is an important setting, since building classifiers using labeled words has been recently shown to require considerably less human labeling effort than building classifiers using labeled documents. We propose the use of Generalized Expectation (GE) criteria combined with a L1 regularization term for learning from labeled features. This lets the feature labels guide model expectation constraints, while approaching feature selection from a regularization perspective. We show that GE criteria combined with L1 regularization consistently outperforms - up to 12% increase in accuracy - the best previously reported results in the literature under the same setting, obtained using L2 regularization. Furthermore, the results obtained with GE criteria and L1 regularizer are competitive to those obtained in the traditional instance-labeling setting, with the same labeling cost.

Proceedings of the 6th International Conference on Pervasive Computing Technologies for Healthcare, 2012

Freezing of gait (FoG) is a common gait deficit in advanced Parkinson's disease (PD). FoG events ... more Freezing of gait (FoG) is a common gait deficit in advanced Parkinson's disease (PD). FoG events are associated with falls, interfere with daily life activities and impair quality of life. FoG is often resistant to pharmacologic treatment; therefore effective non-pharmacologic assistance is needed. We propose a wearable assistant, composed of a smartphone and wearable accelerometers, for online detection of FoG. The system is based on machine learning techniques for automatic detection of FoG episodes. When FoG is detected, the assistant provides rhythmic auditory cueing or vibrotactile feedback that stimulates the patient to resume walking. We tested our solution on more than 8h of recorded lab data from PD patients that experience FoG in daily life. We characterize the system performance on user-dependent and user-independent experiments, with respect to different machine learning algorithms, sensor placement and preprocessing window size. The final system was able to detect FoG events with an average sensitivity and specificity of more than 95%, and mean detection latency of 0.34s in user-dependent settings.

2011 International Conference on P2P, Parallel, Grid, Cloud and Internet Computing, 2011

Vehicular ad-hoc networks (VANETs) have a great potential to improve road safety, traffic jam... more Vehicular ad-hoc networks (VANETs) have a great potential to improve road safety, traffic jams, fuel consumption, and to increase passenger convenience in vehicles. However, VANETs use an open medium for communication and, therefore, are exposed to security threats that influence their reliability. We propose a data-trust security model designed for VANETs based on social network theories. Drivers receiving data about traffic congestion or safety warnings can use the model to evaluate the trust in the received information. The model computes a trust index for each message based on the relevance of the event. It also uses a gossiping approach to disseminate data-trust indexes between vehicles, increasing the accuracy in the trustworthiness of an event and assuring the privacy by hiding the original event sources. The approach is evaluated through modeling and simulation, and we present results that proof its validity.

2012 International Conference on Indoor Positioning and Indoor Navigation (IPIN), 2012

Indoor localization at minimal deployment effort and with low costs is relevant for many ambient ... more Indoor localization at minimal deployment effort and with low costs is relevant for many ambient intelligence and mobile computing applications. This paper presents Ac-tionSLAM, a novel approach to Simultaneous Localization And Mapping (SLAM) for pedestrian indoor tracking that makes use of body-mounted sensors. ActionSLAM iteratively builds a map of the environment and localizes the user within this map. A footmounted Inertial Measurement Unit (IMU) keeps track of the user's path, while observations of location-related actions (e.g. door-opening or sitting on a chair) are used to compensate for drift error accumulation in a particle filter framework. Locationrelated actions are recognizable from body-mounted IMUs that are often used in ambient-assisted living scenarios for context awareness. Thus localization relies only on on-body sensing and requires no ambient infrastructure such as Wi-Fi access points or radio beacons.