>Magister Scientiae - MSc / Hand shape recognition is a pivotal part of any system that attempts to implement Sign Language recognition. This thesis presents a novel system which recognises hand shapes from a single camera view in 2D. By mapping the recognised hand shape from 2D to 3D,it is possible to obtain 3D co-ordinates for each of the joints within the hand using the kinematics embedded in a 3D hand avatar and smooth the transformation in 3D space between any given hand shapes. The novelty in this system is that it does not require a hand pose to be recognised at every frame, but rather that hand shapes be detected at a given step size. This architecture allows for a more efficient system with better accuracy than other related systems. Moreover, a real-time hand tracking strategy was developed that works efficiently for any skin tone and a complex background.

The British Columbia [B.C.] Polydactyly study was designed to investigate the occurrence of Polydactyly in the B.C. population. Polydactylous infants born in the province in the years 1964 - 1967 were ascertained principally through the Registry for Handicapped Children and Adults; additional cases were found-through disease indexes of B.C. hospitals and through referrals from medical personnel. Data on the probands were obtained by family interviews and review of various medical records. Information compiled for a case included: sex of proband, racial background, parental age, parity of mother, notes on the pregnancy, description of accessory digits, genetic family history, and nature of additional congenital anomalies when rpresent. Incidence of Polydactyly in B.C. for 1964 - 1967 was 1.02 per 1,000 live births. This figure includes patients with multiple congenital anomalies. Of the four years covered by the study, 1966 had the highest incidence (1.35), a year in which a seemingly disproportionate number of ulnar cases occurred. Ninety-five males and 43 females, a sex ratio of approximately 2:1, presented the following types of duplication: radial (31), tibial (6), ulnar (54), fibular (24) and ulnar and fibular (16). It is proposed that cases with both ulnar and fibular accessory digits and cases with ulnar or fibular accessory digits in whose family both ulnar and fibular duplication occur present ulnar-fibular type Polydactyly. Based on cases unassociated with multiple congenital anomalies, specific types of Polydactyly had the following incidences: radial 0.20; tibial, 0.04; polysyndactyly, 0.01; ulnar, 0.32; fibular, 0.16; and ulnar-fibular, 0.09. Genetic histories indicate in 32 families out of 138 there is another individual with Polydactyly. For specific types of Polydactyly the following proportion of cases without multiple congenital anomalies had a positive family history: radial, 5 - 19 %; tibial, 0 %; ulnar, 29 %• fibular, 25 and ulnar-fibular, 73 f°~ Polygenic inheritance of ulnar and of fibular types of Polydactyly is suggested by the finding that the percentage of affected sibs is greater when there is a history of Polydactyly in previous generations (approximately 6.6 %) than for probands with a negative history (1.2 %). The number of affected sibs of the probands (3/107) is in agreement with the number which would be expected for a polygenic trait. Data for ulnar-fibular Polydactyly is more compatible with dominant inheritance with reduced penetrance among females. Fifty percent of the offspring of polydactylous males were affected; a smaller percentage of the children of affected mothers were polydactylous (27.7 %). Nonpenetrance of the Polydactyly gene(s) would seem to be fairly common among females since unaffected mothers are known to pass the trait to their children whereas such an occurrence is unknown for males. . Among ulnar-fibular cases with no history of accessory digits in previous generations, 17.4 % of the sibs were polydactylous. Major congenital anomaly was noted in 33 probands and occurs approximately 3½ times more frequently among the polydactylous infants than the general population. Multiple major anomalies are approximately 25 times more frequent. A polydactylous child found to have one major malformation is twice as likely to have additional major anomalies than a child with one major anomaly who has the normal number of digits. Multiple congenital anomalies were most commonly seen with ulnar-fibular Polydactyly. This may reflect the fact that ulnar-fibular duplication is the type most commonly seen as a feature of various syndromes. In addition to the B.C. findings, literature on Polydactyly has been reviewed. Included are: data on the incidence of Polydactyly in other populations, description of different types of duplication, hypotheses regarding the inheritance of accessory digits, tabulation of the types of Polydactyly seen in various syndromes, and discussion of the morphogenesis of accessory digits in relation to limb embryogenesis. / Medicine, Faculty of / Medical Genetics, Department of / Graduate

Magister Curationis - MCur / Studies have shown that no great emphasis has been put on hand washing practices in psychiatric health facilities, despite the fact that nosocomial infection outbreaks have been reported for decades. Most studies have focused on hand washing practices among general health personnel; however, little is known about hand washing practices among nurses working at psychiatric hospitals.

Lateral biases are evident in a number of behaviors across many organisms. The present work was concerned with the particular lateral phenomenon known as handedness. Previous research has suggested that handedness is not a one-dimensional trait. This study evaluated handedness using two factors: hand preference and hand performance. Hand preference refers to the hand chosen to carry out a given action whereas hand performance refers to each hand’s ability, or skill, at carrying out that action. The relationship between hand preference and hand performance has been studied extensively in adults, but the larger body of work with human infants has only assessed hand preference. The goals of this study were to develop a methodology to measure infant hand performance and to begin to examine the relationship between hand preference and hand performance in development. To this end, thirty-six 11-month-old infants were videotaped completing three tasks. The first task assessed hand preference and consisted of a free-play period during which infants were presented with a series of toys that afforded different types of manipulation. The second and third tasks were novel measures of infant hand performance. The second task assessed the infant’s gross motor skills and involved fitting a ball into the top aperture of a toy. The third task assessed the infant’s fine motor skills by requiring infants to retrieve a Cheerio from a stationary plastic cup. Overall, the majority of infants were found to be right-preferent. This was in agreement with previous studies of hand preference in 11-month-olds as well as the pattern of hand preference seen in adults. There was no group-level asymmetry on either measure of hand performance. Hand preference was regressed on hand performance in an overall model of handedness. The right hand’s performance on each task significantly predicted hand preference scores. This was the first study to demonstrate that hand preference can be predicted by hand performance in infants. Future work will examine infant hand proficiency in greater detail as well as the relationship between hand preference and hand performance in nonhuman primate infants.

There are many challenges facing society today, both socially and industrially. Whether it is to improve productivity in factories or with the intention of improving the quality of life of people in their homes, technological advances in robotics and computing have led to solutions to many problems in modern society. These areas are of great interest and are in constant development, especially in societies with a relatively ageing population. In this thesis, we address different challenges in which robotics, artificial intelligence and computer vision are used as tools to propose solutions oriented to home assistance. These tools can be organised into three main groups: “Grasping Challenges”, where we have addressed the problem of performing robot grasping in domestic environments; “Hand Interaction Challenges”, where we have addressed the detection of static and dynamic hand gestures, using approaches based on DeepLearning and GeometricLearning; and finally, “Human Behaviour Recognition”, where using a machine learning model based on hyperbolic geometry, we seek to group the actions that performed in a video sequence.

<p>  </p> <p>To regain a healthy degree of hand function, injured patients require strenuous rehabilitation therapies with the expectation of gaining the full range of motion and strength necessary for performing activities of daily living (ADLs). Metacarpal fractures are one of the most common musculoskeletal injuries and require occupational therapy after the immobilization phase. Obstacles, such as longer recovery times, high costs, or lack of trained physiotherapists, often present a barrier for individuals seeking adequate treatment. Repetitive extension and flexion therapy routines improve grasping functionalities when performed correctly and repetitively. Robotic devices, such as hand exoskeletons, have been found to make up for the lack of hand motor function and assist in grasping tasks performed in ADLs, improving users’ independence. To increase robot acceptability, wearable robots have been recently proposed as part of rehabilitation technologies. Hand rehabilitation systems are an active research interest; however, most studies focus on rehabilitating central nerve injuries. There is a lack of research on systems treating hand fracture injuries, explicitly focusing on function recovery involving the fingers. Integrating systems that provide the necessary dexterity in a user-friendly manner while keeping a compact and lightweight fashion remains challenging. This thesis describes the development of a Soft Hand Exoskeleton (SHE) for robotic hand rehabilitation. The system integrates a flexible glove-like body and a bio-inspired cable-driven transmission system for motion assistance. The exoskeleton’s usage effects were evaluated through a user study experiment. An electromyography (EMG) based analysis allowed us to assess the muscular effort demands of ADLs. Experimental results and evaluation metrics demonstrated a reduction in the total integrated muscular activity (TIMA) in the performance of common ADLs when wearing the SHE system. </p>

The ability to perform precise and relatively independent movements of the fingers is an important feature of the primate's motor repetoire. A substantial amount of evidence suggests a special role for the hand area of the motor cortex in the execution of these movements, and in particular, thoses cells which make direct cortico-motoneuronal connections with motoneurones innervating the hand muscles. The axons of these corticomotoneuronal (CM) cells form part of the pyramidal tract. The subject of this study is the influence of these CM cells upon the activity of hand muscles in the conscious monkey performing a precision grip task between thumb and index fingers. Three aspects of the cortico-motoneuronal connection have been investigated; 1. The effect of discharges in individual CM cells on the gross e.m.g. activity of intrinsic hand and forearm muscles, and the distribution of these effects over various muscles. 2. The effect of individual CM cell discharge on the single motor unit activity in the intrinsic thumb muscles. 3. The significance of differences in firing frequency of individual CM cells upon their modulation of gross e.m.g. The spike-triggered averaging technique was employed to study the overall effect of identified pyramidal tract neurones (PTN) on gross e.m.g. activity. Rectified gross e.m.g. was averaged with respect to the discharges of single PTNs. The influence of the triggering cortical cell was revealed in the averages as a transient increase in muscle activity at the appropriate time after the cell had fired. This effect is called post-spike facilitation (PSF). The occurance of PSF in the spike-triggered average was taken as evidence that the triggering cells made a direct, presumably monosynaptic, connection with the motoneurones of the muscle in question. Quantitative and qualitative aspects of PSF are discussed. The activity of 5 to 10 intrinsic hand and forearm muscles were averaged with respect to individual PTNs. In this way the PSF distribution over different muscles, produced by one cortical cell, could be determined. It was established that CM cells produce PSF in a relatively restricted number of muscles (2-3 out of 10 tested). The impact of CM cells on the discharge activity of single motor units was studied by cross-correlating both spike trains. To overcome the considerable sampling problem, the occurrence of PSF in the spike-triggered gross e.m.g. average was used as a criterion for further cross-correlation analysis. The cross-correlogram peaks half-width and onset latency were suggestive for monosynaptic excitation of the motoneurones. Different approaches are discussed to estimate the strength of these connections and to relate these strengths to the number of synpatic boutons that the CM cell makes upon its target motoneurones. In terms of the number of extra discharges fired by the single motor unit in response to CM cell firing, the effects can be described as weak (5-20 extra motor unit spikes per 1000 CM cell discharges). It was possible to study the connectivity of single CM cells with different motor units of the same muscle by simultaneously recording several (2-5) motor units. It was found that if a CM cell showed a correlogram peak with one of the concurrently sampled motor units, most other motor units sampled from the same muscle would give correlogram peaks with the same CM cell. This finding is suggestive for a rather diffuse collateralization of the CM cell axons within the motoneurone pool of the target muscle. Finally the influence of different firing frequencies of CM cells on the form and strength of PSF effects was investigated. A striking result was that discharges preceded by long interspike intervals (> 50ms)could still exert quite powerful excitatory effects. The results of this series of experiments further elucidate the nature of the corticomotoneuronal connection and give an impression of the facilitatory capacity of single CM cells on motoneurons innervating the hand muscles.

There are two main hypotheses examined in the dissertation. The first is that the physiology of the hand provides motivation for the claim that sign language handshapes can be considered easy, hard or even impossible to articulate. The second hypothesis is that easy handshapes occur more often than expected, hard handshapes occur less often than expected and impossible handshapes don't occur at all within a single sign language. These hypotheses are examined in the following ways: first, I provide a detailed explanation of the physiology of the hand from which I conclude that not all fingers are equal in skill and not all configurations a hand may assume are equally easy. Second, based on the physiology, I propose a metric for determining which handshapes are "easy" and which are "difficult". Third, I examine whether the "easy" handshapes occur more often than expected, while the "hard" handshapes occur less often than expected in the signs of two languages, American Sign Language (ASL) and Taiwan Sign Language (TSL). I conclude that the hypothesis that the "easy" handshapes occur more often than expected and the "hard" handshapes occur less often than expected is supported in approximately half of the cases.

No description available.

No description available.