Vidéosurveillance intelligente pour la détection de chutes chez les personnes âgées

Les pays industrialisés comme le Canada doivent faire face au vieillissement de leur population. En particulier, la majorité des personnes âgées, vivant à domicile et souvent seules, font face à des situations à risques telles que des chutes. Dans ce contexte, la vidéosurveillance est une solution innovante qui peut leur permettre de vivre normalement dans un environnement sécurisé. L’idée serait de placer un réseau de caméras dans l’appartement de la personne pour détecter automatiquement une chute. En cas de problème, un message pourrait être envoyé suivant l’urgence aux secours ou à la famille via une connexion internet sécurisée. Pour un système bas coût, nous avons limité le nombre de caméras à une seule par pièce ce qui nous a poussé à explorer les méthodes monoculaires de détection de chutes. Nous avons d’abord exploré le problème d’un point de vue 2D (image) en nous intéressant aux changements importants de la silhouette de la personne lors d’une chute. Les données d’activités normales d’une personne âgée ont été modélisées par un mélange de gaussiennes nous permettant de détecter tout événement anormal. Notre méthode a été validée à l’aide d’une vidéothèque de chutes simulées et d’activités normales réalistes. Cependant, une information 3D telle que la localisation de la personne par rapport à son environnement peut être très intéressante pour un système d’analyse de comportement. Bien qu’il soit préférable d’utiliser un système multi-caméras pour obtenir une information 3D, nous avons prouvé qu’avec une seule caméra calibrée, il était possible de localiser une personne dans son environnement grâce à sa tête. Concrêtement, la tête de la personne, modélisée par une ellipsoide, est suivie dans la séquence d’images à l’aide d’un filtre à particules. La précision de la localisation 3D de la tête a été évaluée avec une bibliothèque de séquence vidéos contenant les vraies localisations 3D obtenues par un système de capture de mouvement (Motion Capture). Un exemple d’application utilisant la trajectoire 3D de la tête est proposée dans le cadre de la détection de chutes. En conclusion, un système de vidéosurveillance pour la détection de chutes avec une seule caméra par pièce est parfaitement envisageable. Pour réduire au maximum les risques de fausses alarmes, une méthode hybride combinant des informations 2D et 3D pourrait être envisagée.

Développement et validation d'un modèle de simulation numérique personnalisé à une athlète de plongeon

Les entraîneurs en sports acrobatiques disposent de peu d’outils permettant d’améliorer leur compréhension des saltos vrillés et la performance des athlètes. L’objectif de ce mémoire était de développer un environnement graphique de simulation numérique réaliste et utile des acrobaties aériennes. Un modèle composé de 17 segments et de 42 degrés de liberté a été développé et personnalisé à une athlète de plongeon. Un système optoélectronique échantillonné à 300 Hz a permis l’acquisition de huit plongeons en situation réelle d’entraînement. La cinématique articulaire reconstruite avec un filtre de Kalman étendu a été utilisée comme entrée du modèle. Des erreurs quadratiques moyennes de 20° (salto) et de 9° (vrille) entre les performances simulées et réelles ont permis de valider le modèle. Enfin, une formation basée sur le simulateur a été offerte à 14 entraîneurs en sports acrobatiques. Une augmentation moyenne de 11 % des résultats aux questionnaires post-test a permis de constater le potentiel pédagogique de l’outil pour la formation.

Expérience subjective et différences individuelles dans l'intégration d'informations visuelle et kinesthésique

L’expérience subjective accompagnant un mouvement se construit a posteriori en intégrant différentes sources d’informations qui s’inter-influencent à différents moments tant avant qu’après le mouvement. Cette expérience subjective est interprétée par un modèle d’attribution bayésien afin de créer une expérience d’agentivité et de contrôle sur les mouvements de son propre corps. Afin de déterminer l’apport de l’interaction entre les paramètres considérés par le modèle d’attribution et d’investiguer la présence de disparités inter-individuelles dans la formation de l’expérience subjective du mouvement, une série de 90 pulsations simples de stimulation magnétique transcrânienne (SMT) sur le cortex moteur primaire (M1) suivi de multiples questions sur l’expérience subjective reliée au mouvement provoqué a été effectuée chez 20 participants normaux. Les données objectives du mouvement ont été recueillies par électromyographie (EMG) et capture du mouvement. Un modèle de régression a entre autres été effectué pour chaque participant afin de voir quelle proportion du jugement subjectif pouvait être expliqué par des indices objectifs et cette proportion variait grandement entre les participants. Les résultats de la présente étude indiquent la présence d’une capacité individuelle à se former des jugements subjectifs reflétant adéquatement la réalité comme en témoigne la cohérence entre les différentes mesures d’acuité et plusieurs variables mesurant l’expérience subjective.

Video-based analysis of Gait pathologies

L’analyse de la marche a émergé comme l’un des domaines médicaux le plus im- portants récemment. Les systèmes à base de marqueurs sont les méthodes les plus fa- vorisées par l’évaluation du mouvement humain et l’analyse de la marche, cependant, ces systèmes nécessitent des équipements et de l’expertise spécifiques et sont lourds, coûteux et difficiles à utiliser. De nombreuses approches récentes basées sur la vision par ordinateur ont été développées pour réduire le coût des systèmes de capture de mou- vement tout en assurant un résultat de haute précision. Dans cette thèse, nous présentons notre nouveau système d’analyse de la démarche à faible coût, qui est composé de deux caméras vidéo monoculaire placées sur le côté gauche et droit d’un tapis roulant. Chaque modèle 2D de la moitié du squelette humain est reconstruit à partir de chaque vue sur la base de la segmentation dynamique de la couleur, l’analyse de la marche est alors effectuée sur ces deux modèles. La validation avec l’état de l’art basée sur la vision du système de capture de mouvement (en utilisant le Microsoft Kinect) et la réalité du ter- rain (avec des marqueurs) a été faite pour démontrer la robustesse et l’efficacité de notre système. L’erreur moyenne de l’estimation du modèle de squelette humain par rapport à la réalité du terrain entre notre méthode vs Kinect est très prometteur: les joints des angles de cuisses (6,29◦ contre 9,68◦), jambes (7,68◦ contre 11,47◦), pieds (6,14◦ contre 13,63◦), la longueur de la foulée (6.14cm rapport de 13.63cm) sont meilleurs et plus stables que ceux de la Kinect, alors que le système peut maintenir une précision assez proche de la Kinect pour les bras (7,29◦ contre 6,12◦), les bras inférieurs (8,33◦ contre 8,04◦), et le torse (8,69◦contre 6,47◦). Basé sur le modèle de squelette obtenu par chaque méthode, nous avons réalisé une étude de symétrie sur différentes articulations (coude, genou et cheville) en utilisant chaque méthode sur trois sujets différents pour voir quelle méthode permet de distinguer plus efficacement la caractéristique symétrie / asymétrie de la marche. Dans notre test, notre système a un angle de genou au maximum de 8,97◦ et 13,86◦ pour des promenades normale et asymétrique respectivement, tandis que la Kinect a donné 10,58◦et 11,94◦. Par rapport à la réalité de terrain, 7,64◦et 14,34◦, notre système a montré une plus grande précision et pouvoir discriminant entre les deux cas.

Design and Development of a Frame Based MT System for English-to-ISL

This paper presents the design and development of a frame based approach for speech to sign language machine translation system in the domain of railways and banking. This work aims to utilize the capability of Artificial intelligence for the improvement of physically challenged, deaf-mute people. Our work concentrates on the sign language used by the deaf community of Indian subcontinent which is called Indian Sign Language (ISL). Input to the system is the clerk’s speech and the output of this system is a 3D virtual human character playing the signs for the uttered phrases. The system builds up 3D animation from pre-recorded motion capture data. Our work proposes to build a Malayalam to ISL

The Role of Kinaesthetic Motor Imagery in Promoting Activation of the Lumbrical Muscles in the Hemiparetic Hand: a Randomised Controlled Study

The objective of the investigation who gave as result this work was to investigate the effectiveness of kinaesthetic motor imagery in the activation of the hemiplegic hand muscles following stroke. The experiment consisted of two random groups. Movements were measured after treatment. The participants were ten patients with hemiplegic hands (men who mean age was 74.4 years; mean time since stroke 3.05 months). All patients received three sessions of physical treatment based on an identical treatment protocol. Five patients were randomly assigned to an experimental group practising kinaesthetic motor imagery of a grasp using the 'lumbrical action' (experimental group). The others five (control group) followed a relaxation script. All the patients were then asked to grasp an object using the 'lumbrical action'. The grasps were recorded using an optoelectronic motion capture system. The magnitude of the extension of the index finger and the correlation of the angular displacement of the proximal phalangeal joints and the metacarpophalangeal joints were calculated. The movement time for the whole grip was calculated. The experimental group demonstrated higher extension in the index finger (p = < 0.01) and they had a higher correlation coefficient (0.99) than the control group (0.77) for the displacement of the proximal interphalangeal joint and the metacarpophalangeal joints. The movement time for the experimental group was faster, although the difference was not significant.

Low-cost optical tracking for immersive collaboration in the CAVE using the Wii remote

We present a novel way of interacting with an immersive virtual environment which involves inexpensive motion-capture using the Wii Remote®. A software framework is also presented to visualize and share this information across two remote CAVETM-like environments. The resulting applications can be used to assist rehabilitation by sending motion information across remote sites. The application’s software and hardware components are scalable enough to be used on desktop computer when home-based rehabilitation is preferred.

Marker placement to describe the wrist movements during activities of daily living in cyclical tasks

Objective. To describe the wrist kinematics during movement through free range of motion and activities of daily living using a cyclical task. Design. The wrist angles were initially calculated in a calibration trial and then in two selected activities of daily living (jar opening and carton pouring). Background. Existing studies which describe the wrist movement do not address the specific application of daily activities. Moreover, the data presented from subject to subject may differ simply because of the non-cyclical nature of the upper limbs movements. Methods. The coordinates of external markers attached to bone references on the forearm and dorsal side of the hand were obtained using an optical motion capture system. The wrist angles were derived from free motion trials and successively calculated in four healthy subjects for two specific cyclical daily activities (opening a jar and pouring from a carton). Results. The free motions trial highlighted the interaction between the wrist angles. Both the jar opening and the carton pouring activity showed a repetitive pattern for the three angles within the cycle length. In the jar-opening task, the standard deviation for the whole population was 10.8degrees for flexion-extension, 5.3degrees for radial-ulnar deviation and 10.4degrees for pronation-supination. In the carton-pouring task, the standard deviation for the whole population was 16.0degrees for flexion-extension, 3.4degrees for radial-ulnar deviation and 10.7degrees for pro nation-supination. Conclusion. Wrist kinematics in healthy subjects can be successfully described by the rotations about the axes of marker-defined coordinates systems during free range of motion and daily activities using cyclical tasks.

Preliminary observations on the presence of sustained tendon strain and eccentric contractions of the wrist extensors during a common manual task: implications for lateral epicondylitis

Lateral epicondylitis (LE) is hypothesized to occur as a result of repetitive, strenuous and abnormal postural activities of the elbow and wrist. There is still a lack of understanding of how wrist and forearm positions contribute to this condition during common manual tasks. In this study the wrist kinematics and the wrist extensors’ musculotendon patterns were investigated during a manual task believed to elicit LE symptoms in susceptible subjects. A 42-year-old right-handed male, with no history of LE, performed a repetitive movement involving pushing and turning a spring-loaded mechanism. Motion capture data were acquired for the upper limb and an inverse kinematic and dynamic analysis was subsequently carried out. Results illustrated the presence of eccentric contractions sustained by the extensor carpi radialis longus (ECRL), together with an almost constant level of tendon strain of both extensor carpi radialis brevis (ECRB) and extensor digitorum communis lateral (EDCL) branch. It is believed that these factors may partly contribute to the onset of LE as they are both responsible for the creation of microtears at the tendons’ origins. The methodology of this study can be used to explore muscle actions during movements that might cause or exacerbate LE.

Low-cost optical tracking for collaborative applications in immersive virtual environments

We present a novel way of interacting with an immersive virtual environment which involves inexpensive motion-capture using the Wii Remote®. A software framework is also presented to visualize and share this information across two remote CAVETM-like environments. The resulting application can be used to assist rehabilitation by sending motion information across remote sites. The application’s software and hardware components are scalable enough to be used on a desktop computer when home-based rehabilitation is preferred.

Efeito da estimulação transcraniana por corrente contínua anódica sobre o controle do movimento em para-halterofilistas

Introduction: The sport practiced by people with disabilities has been growing in recent years. Consequently, advances in assessment and training methods have emerged. However, the paralympic sport keeps in tow these advances, with few specific studies that consider disability as intervening factor. The transcranial direct current stimulation (tDCS) is a technique that has proven to be capable of modulating brain function. Studies show beneficial effects of tDCS on muscle strength, power and fatigue during exercise. Objective: Investigate de the effect of tDCS on movement control in para-powerlifters. Methods: Eight subjects underwent two sessions of motion capture, which previously applied the anodic tDCS or sham sessions in the cerebellum. Three movements were performed with increasing load between 90-95% of 1MR. The movements were recorded by an 10 infrared cameras system which reconstructed the 3D trajectory of markers placed on the bar. Results: There have been changes between the anodic and sham conditions over bar level (initial, final, maximum during the eccentric and concentric phase) and in the difference between the final and initial bar level. Moreover, there was difference in bar level (final and during the eccentric phase) comparing athletes amputees and les autres. Conclusion: The findings of this study suggest that tDCS applied prior to the exercise over the cerebellum in para-powerlifters acts differently according to disability

Efeitos da adição de carga no treino de marcha na esteira em indivíduos com Doença de Parkinson: ensaio clínico controlado randomizado

Introduction: The intrinsic gait disorders in individuals with Parkinson's disease (PD) are one of the most disabling motor symptoms. Among the therapeutic approaches used in attempts to improve the motor function, especially the gait pattern of individuals, stands out the treadmill gait training associated with the addition of load. However, there are few findings that elucidate the benefits arising from such practice. Objective: To assess the effects of adding load on the treadmill gait training in individuals with PD. Material and Methods: A controlled, randomized and blinded clinical trial, was performed with a sample of 27 individuals (18 men and 9 women) with PD, randomly assigned to three experimental conditions, namely: treadmill gait training (n=9), treadmill gait training associated with addition of 5% load (n=9) and treadmill gait training associated with addition of 10% load (n=9). All volunteers were assessed, during phase on of Parkinson's medication, regarding to demographic, clinical and anthropometric (identification form) data, level of disability (Hoehn and Yahr Modified Scale), cognitive function (Mini Mental State Examination), clinical functional - in those areas activity of daily living and motor examination (Unified Parkinson's Disease Rating Scale - UPDRS) and gait cinematic analysis was performed through Qualisys Motion Capture System®. The intervention protocol consisted of gait training in a period of 4 consecutive weeks, with three weekly sessions, lasting 30 minutes each. The post-intervention assessment occurred the next day after the last training session, which was performed cinematic analysis of gait and the UPDRS. Data analysis was performed using the software Statistical Package for Social Sciences® (SPSS) 17.0. Results: The age of volunteers ranged from 41 to 75 years old (62,26 ± 9,07) and the time of clinical diagnosis of PD between 2 to 9 years (4,56 ± 2,42). There was a reduction regarding the score from motor exam domain (p=0,005), only when training with the addition of a 5% load. As for the space-time variables there was no significant difference between groups (p>0,120); however, the training with addition of 5% load presented the following changes: increase in stride length (p=0,028), in step length (p=0,006), in time balance of the most affected member (p=0,006) and reduction in support time of the referred member (p=0,007). Regarding angular variables significant differences between groups submitted to treadmill gait training without addition load and with 5% of load were observed in angle of the ankle at initial contact (p=0,019), in plantar flexion at toe-off (p=0,003) and in the maximum dorsiflexion in swing (p=0,005). While within groups, there was a reduction in amplitude of motion of the ankle (p=0,048), the only workout on the treadmill. Conclusion: The treadmill gait training with addition of 5% load proved to be a better experimental condition than the others because it provided greater gains in a number of variables (space-time and angular gait) and in the motion function, becoming a therapy capable of effectively improving the progress of individuals with PD

Efeitos do treino em esteira na marcha com dupla tarefa de indivíduos com doença de Parkinson: ensaio clínico controlado randomizado

Background: The gait automaticity loss difficults realization of concurrent activities - Dual Task (DT). In these situations, individuals with Parkinson`s disease (PD) show a significant reduction in gait velocity and stride length, as strides variability and asymmetry increased, factors predisposing to falls. However, recent studies have shown that training involving DT may cause subsequent improvements in gait variables with DT in individuals with PD. The treadmill use was adopted by this study, by promoting greater regularity in step and enhance training. Objective:To investigate immediate effects of gait training associated with cognitive tasks on gait in individuals with PD. Methods: Twenty-two volunteers were randomly divided into two groups: control group (n = 11), who performed gait training on a treadmill for 20 minutes, and the experimental group (n = 11), who performed treadmill gait training for 20 minutes associated with cognitive tasks of verbal fluency, memory, and spatial planning. Participants were evaluated in phase on of antiparkinsonian medication as the demographic, clinical and anthropometric (identification form), cognitive status (Montreal Cognitive Assessment - MoCA), executive function (Frontal Assessment Battery), level of physical disability (Hoehn and Yahr Modified), motor and functional status (Unified Rating Scale for Parkinson`s Disease - UPDRS), and kinematics (Qualisys Motion Capture System). Results: There were not differences between groups, but both showed improvement after the intervention. The control group had an increase in velocity (p = 0.008), stride length (p = 0.04), step length (p = 0.02) and decreased double support time(p = 0.03). The experimental group showed an increase in speed (p = 0.002), stride length (p = 0.008), step length (p = 0.02) and cadence (p = 0.01), as well as a decrease in the width stride (p = 0.001) and total support time (p = 0.02). As the angular variables, the experimental group had a significant increase in the initial contact angle of ankle (p = 0.01). Conclusion: The gait training combined with cognitive activities didn`t provide significant improvements in gait variables with DT, but this study was the first to demonstrate that gait training on treadmill as simple task minimized the negative interference of DT in PD

Does proprioceptive system stimulation improve sit-to-walk performance in healthy young adults?

[Purpose] Sit-to-walk performance is linked to proper proprioceptive information processing. Therefore, it is believed that an increase of proprioceptive inflow (using muscle vibration) might improve sit-to-walk performance. However, before testing muscle vibration effects on a frail population, assessment of its effects on healthy young people is necessary. Thus, the aim of this study was to investigate the effects of muscle vibration on sit-to-walk performance in healthy young adults. [Subjects and Methods] Fifteen young adults performed the sit-to-walk task under three conditions: without vibration, with vibration applied before movement onset, and with vibration applied during the movement. Vibration was applied bilaterally for 30 s to the tibialis anterior, rectus femoris, and upper trapezius muscles bellies. The vibration parameters were as follows: 120 Hz and 1.2 mm. Kinematics and kinetic data were assessed using a 3D motion capture system and two force plates. The coordinates of reflective markers were used to define the center-of-mass velocities and displacements. In addition, the first step spatiotemporal variables were assessed. [Results] No vibration effect was observed on any dependent variables. [Conclusion] The results show that stimulation of the proprioceptive system with local muscle vibration does not improve sit-towalk performance in healthy young adults.

CID: an efficient complexity-invariant distance for time series

The ubiquity of time series data across almost all human endeavors has produced a great interest in time series data mining in the last decade. While dozens of classification algorithms have been applied to time series, recent empirical evidence strongly suggests that simple nearest neighbor classification is exceptionally difficult to beat. The choice of distance measure used by the nearest neighbor algorithm is important, and depends on the invariances required by the domain. For example, motion capture data typically requires invariance to warping, and cardiology data requires invariance to the baseline (the mean value). Similarly, recent work suggests that for time series clustering, the choice of clustering algorithm is much less important than the choice of distance measure used.In this work we make a somewhat surprising claim. There is an invariance that the community seems to have missed, complexity invariance. Intuitively, the problem is that in many domains the different classes may have different complexities, and pairs of complex objects, even those which subjectively may seem very similar to the human eye, tend to be further apart under current distance measures than pairs of simple objects. This fact introduces errors in nearest neighbor classification, where some complex objects may be incorrectly assigned to a simpler class. Similarly, for clustering this effect can introduce errors by “suggesting” to the clustering algorithm that subjectively similar, but complex objects belong in a sparser and larger diameter cluster than is truly warranted.We introduce the first complexity-invariant distance measure for time series, and show that it generally produces significant improvements in classification and clustering accuracy. We further show that this improvement does not compromise efficiency, since we can lower bound the measure and use a modification of triangular inequality, thus making use of most existing indexing and data mining algorithms. We evaluate our ideas with the largest and most comprehensive set of time series mining experiments ever attempted in a single work, and show that complexity-invariant distance measures can produce improvements in classification and clustering in the vast majority of cases.