Analysis of intentional head gestures to assist computer access by physically disabled people

Any human-computer interface requires both a means of transducing information flowing from the person and a way of classifying this information in a form that can be used by an application program. Since several interface devices exploit the head movements of disabled people to control computers, this paper includes a discussion of existing technologies based on head movements. As an alternative to simple techniques based on pointing to classify this information, this paper studies the possibility of using a combination of pointing and movement gestures to control an application program. By using hidden Markov models to classify movements into ‘yes’, ‘no’ and spurious gestures, it was possible to control a simple graphics application program. Subsequent analysis showed that the hidden Markov models achieved a 74% success rate.     

Predicting the voluntary arm forces in FES-assisted standing up using neural networks

 For individuals with paraplegia, standing up requires activation of paralyzed leg muscles by an artificial functional electrical stimulation (FES) controller and voluntary control of arm forces by the individual. Any knowledge of such voluntary control, particularly its prediction, could be used to design more effective FES controllers. Therefore, artificial neural network models were developed to predict voluntary arm forces from measured angular positions of the ankle, knee, and hip joints during FES-assisted standing up in paraplegia. The training data were collected from eight paraplegic subjects in repeated standing-up trials, and divided into two categories for training and validation. The predictions of the models closely followed both the training and validation data, showing good accuracy and generalization. The comparison of the models showed that, although there are striking similarities among the voluntary controls adopted by different subjects, each subject develops his/her own `personal strategy' to control the arm forces, which is consistent from trial to trial. The level of consistency was dependent on the experience in using FES, injury level, body weight, and other subject-specific parameters. Received: 5 January 1999 / Accepted in revised form: 29 January 2001     

Response of thirsty rats to absence of water: Frustration,disinhibition or compensation?

Water-deprived rats were allowed to drink in an experimental chamber during a series of training trials, and were then presented with an empty water spout during a single test trial. Their initial response to absence of water was to bite, sniff and paw at the empty spout (spout-directed behaviour), after which they engaged in activities such as eating, grooming and exploration of the chamber. By comparison with control animals that had no expectation of water, experimental rats performed these latter activities in a bizarre ‘speeded-up’ manner, which was quantifiable as an increase in local rate. Both the occurrence of spout-directed behaviour and the increase in local rate of other activities were enhanced by increasing the prior level of water deprivation and by offering a more palatable fluid during training. The results are discussed in relation to current models of behavioural switching.     

Rats more readily acquire a time-of-day go no-go discrimination than a time-of-day choice discrimination

Male Sprague-Dawley rats were used to compare six time-place training procedures that differed with respect to housing or training conditions. All procedures involved training food-deprived rats to enter one choice arm of a T-maze during a morning test session and to enter the other choice arm during an afternoon session to obtain Cocoa Puffs(R). The task proved to be difficult. Only 39 of 49 rats attained a criterion of nine correct choices on ten consecutive trials within a total of 120 trials. Making one choice arm distinct, limiting consecutive same correct choices to two, giving one session during the light and one during the dark portion of the light cycle, extinguishing perseveration of the same choice responses, and housing the rats with a natural light cycle all failed to significantly decrease the errors or trials to a 90% correct choice criterion. In contrast, all responding rats (n=7) showed significantly better than chance performance within 48 trials when the task was a go no-go discrimination based on time of day. Learning to make a response during a session when a choice to either choice arm is reinforced and to withhold responding during a session when a choice to neither choice arm is reinforced was relatively easy for the rats to acquire. Continued high level performance after the light cycle was eliminated, a random feeding schedule was initiated, and other time-related cues were masked suggests that the rats used internal cues or an internal clock to make the correct go or no-go response. It was concluded that rats are prepared to use time of day as an occasion-setting stimulus, but have difficulty using time of day as a signal for a specific response.     

Hidden Markov Models and Animal Behaviour

This paper proposes the use of hidden Markov time series models for the analysis of the behaviour sequences of one or more animals under observation. These models have advantages over the Markov chain models commonly used for behaviour sequences, as they can allow for time-trend or expansion to several subjects without sacrificing parsimony. Furthermore, they provide an alternative to higher-order Markov chain models if a first-order Markov chain is unsatisfactory as a model. To illustrate the use of such models, we fit multivariate and univariate hidden Markov models allowing for time-trend to data from an experiment investigating the effects of feeding on the locomotory behaviour of locusts (Locusta migratoria).     

A two-step procedure for coupling development and usage of a pair of human neck models

Both finite element models and multi-body models of human head-neck complex had been widely used in neck injuries analysis, as the former could be used to generate detailed stress strain information and the later could generate dynamic responses with high efficiency. Sometimes, detailed stress and strain information were hoped to be obtained more efficiently, but current methods were not effective enough when they were used to analyze responses of human head neck complex to long duration undulate accelerations. In this paper, a two-step procedure for ‘parallel’ development and ‘sequential’ usage of a pair of human head neck models was discussed. The pair of models contained a finite element model and a multi-body model, which were developed based on the coupling ‘parallel’ procedure using the same bio-realistic geometry. After being validated using available data, the pair of human neck models were applied to analyze biomechanical responses of pilot’s neck during arrested landing operation according to the ‘sequential’ procedure, because typical sustained undulate accelerations usually appeared during such processes. The results, including both kinematic and detailed biomechanical responses of human head-neck complex, were obtained with preferable efficiency. This research provided an effective way for biomechanical analysis of human head neck responses to sustained undulate accelerations.     

Stimulus sequence effects on human express saccades described by a Markov model

Express saccades predominantly occur in experiments employing the gap paradigm where the target onset is separated from the fixation point offset by a blank period. Their relative frequency is distinctly influenced by catch trials (i.e. trials without a saccadic target) mixed into the stream of regular target trials. Generalizing this concept for other stimulus uncertainties (direction, amplitude), we found that the preparation time of a saccade depends on both the type of uncertainty used and the sequence of trial type (e.g., target vs catch, left vs right) in the experiment. This stimulus sequence effect is most prominent for catch trials. A similar but less pronounced effect can still be observed in the case of direction uncertainty but not in that of amplitude uncertainty. A two-state Markov process model is proposed which is based on the dichotomy of express and regular saccades in the gap paradigm. According to this model the actual state of the saccadic system, which determines the type of saccade just in preparation, depends on the "trial history". The implications for models of saccade programming are discussed. Received: 14 April 1993/Accepted in revised form: 2 July 1993     

Induction of habits in rats by a forced-choice procedure in T-maze and the effect of pre-test free exploration

A forced-choice procedure in T-maze designed for the induction of habits was used to induce strong habits in rats. The response choices of rats in 20 free-choice trials were compared after the rats had been subjected to 1 or 200 forced-choice trials to one side of the T-maze. After 200 forced-choice trials the rats showed a significant (p < .001) propensity for the habitual arm of the maze in the subsequent free-choice trials. The habit was at least as pronounced when analysed over the last 10 free-choice trials as when it was analysed over the first 10 free-choice trials. When the rats were given the opportunity to explore the entire maze immediately before the free-choice challenge after 200 forced-choice trials, this resulted in a large variation in the choice pattern of the individual rats, and a subgroup of rats choose the newly opened maze arm in 95-100% of the 20 free-choice trials.     

Behavioural correlate of choice confidence in a discrete trial paradigm

How animals make choices in a changing and often uncertain environment is a central theme in the behavioural sciences. There is a substantial literature on how animals make choices in various experimental paradigms but less is known about the way they assess a choice after it has been made in terms of the expected outcome. Here, we used a discrete trial paradigm to characterise how the reward history shaped the behaviour on a trial by trial basis. Rats initiated each trial which consisted of a choice between two drinking spouts that differed in their probability of delivering a sucrose solution. Critically, sucrose was delivered after a delay from the first lick at the spouts--this allowed us to characterise the behavioural profile during the window between the time of choice and its outcome. Rats' behaviour converged to optimum choice, both during the acquisition phase and after the reversal of contingencies. We monitored the post-choice behaviour at a temporal precision of 1 millisecond; lick-response profiles revealed that rats spent more time at the spout with the higher reward probability and exhibited a sparser lick pattern. This was the case when we exclusively examined the unrewarded trials, where the outcome was identical. The differential licking profiles preceded the differential choice ratios and could thus predict the changes in choice behaviour.     

Effects of Artichoke Latent Virus Infection on the Production of Artichoke Heads

Virus-free and artichoke latent virus (ALV) infected plants of ‘Brindisino’ artichoke obtained by in vitro propagation, were studied over a period of three years to evaluate the effect of ALV infection on artichoke field performance and to determine the infection rate of healthy plants. ALV infection caused qualitative and quantitative changes ‘Brindisino’ artichoke such as leaf and bract discoloration, opening of head apex, delay of first harvest, shortening of head stalk, reduction of head width and a dramatic decrease of yield. Due to the high infection rate of healthy plants during the trial period, the differences in the production of heads between the two groups of plants decreased in the last year of experiment.     

A fruit in hand is worth many more in the bush: steep spatial discounting by free-ranging rhesus macaques (Macaca mulatta)

Decision making is one of the principal cognitive processes underlying goal-directed behaviour and thus there is justifiably strong interest in modeling it. However, many of these models have yet to be tested outside of the laboratory. At the same time, field work would benefit from the use of experimental methods developed in the laboratory to determine the causal relationships between environmental variables and behaviour. We therefore adapted a laboratory-derived experimental paradigm to test decision making in the wild. The experiment used an indifference-point procedure to determine the influence of both the amount and distance of food on choice behaviour. Free-ranging rhesus monkeys were given the choice between a smaller amount of food at a closer distance and a larger amount farther away. In four conditions, we held the closer amount constant across trials and varied the farther amount to determine the point at which the monkeys were indifferent to the choice alternatives. For example, in condition one, we used one piece of food at the closer location, and determined how many pieces would be equivalent in the farther location. Four different closer amounts were tested to obtain an indifference point curve, with the indifference amounts at the farther location plotted against the closer amounts. The slope of the obtained linear indifference curve was surprisingly high, suggesting that rhesus monkeys significantly discount food that is farther away. Possible reasons for this steep spatial discounting are discussed.     

Aversive stimulation and rats' preference for areas differing in novelty-value and brightness

Hooded rats were presented with black-white maze arms differing in novelty-value. Rats shocked prior to choice entered the familiar arm when this arm was black and displayed random choice when this arm was white. During an observation period following initial choice, the shocked rats that entered the familiar arm showed a greater tendency to remain in the arm initially entered than the shocked rats that entered the novel arm. The unshocked rats preferred the novel arm, irrespective of brightness, during the initial choice; this tendency was attenuated during the period following initial choice in the condition where the novel arm was also the brighter (white) alternative. The results are discussed with reference to two-factor (approach/avoidance) theories of exploratory behaviour.     

Dynamics of pre- and post-choice behaviour: rats approximate optimal strategy in a discrete-trial decision task

We simulate two types of environments to investigate how closely rats approximate optimal foraging. Rats initiated a trial where they chose between two spouts for sucrose, which was delivered at distinct probabilities. The discrete trial procedure used allowed us to observe the relationship between choice proportions, response latencies and obtained rewards. Our results show that rats approximate the optimal strategy across a range of environments that differ in the average probability of reward as well as the dynamics of the depletion-renewal cycle. We found that the constituent components of a single choice differentially reflect environmental contingencies. Post-choice behaviour, measured as the duration of time rats spent licking at the spouts on unrewarded trials, was the most sensitive index of environmental variables, adjusting most rapidly to changes in the environment. These findings have implications for the role of confidence in choice outcomes for guiding future choices.     

The Feeding Behaviour of Starlings (Sturnus vulgaris) in the Presence of ‘Eyes’

Previous studies that have looked at the aversive properties of simple, predator-related, eye-like patterns presented in an artificial context, have yielded conflicting results. The aim of the present series of experiments was to investigate whether such stimuli had potential for use as bird scarers. The starling (Sturnus vulgaris) was the subject species. The experimental procedure was designed to provide a relevant and objective, interval scale of aversiveness based upon the ability of eye patterns to deter hungry birds from a feeding area. Each subject was deprived of food for 5 h and then tested in an apparatus which exposed an eye pattern over a food trough immediately after the bird alighted on that trough. 10 measures of the starlings' feeding behaviour were monitored over a 1-h trial. It was predicted that five of these variables would be positively and five negatively correlated with an increase in the fear evoking properties of the stimulus. The 10 measures were analysed using principal component analysis and the first component in every experiment had the signs of the latent vectors of the variables exactly in the predicted pattern and therefore appeared to be an ‘aversiveness index’. The scores from this first component were then used in a Latin square ANOVA to distinguish subject, test order and stimulus effects. The main findings from the eight experiments are as follows. Simple eye-like patterns can indeed deter hungry starlings from feeding in their vicinity. The presence of a pair of eyes painted on a white card caused an 88% reduction in time spent on the nearby food trough and a 65% reduction in actual feeding time. The presence of ‘pupils’ in the patterns is essential for simple circles alone are not significantly more aversive than the control card. Once a pattern has the ‘pupil/iris’ distinction then a circular outline becomes important although the shape of the ‘pupil’ appears relatively unimportant. Eyes with coloured irises are more effective than black and white patterns but the degree of contrast between the ‘pupil’ and ‘iris’ appears irrelevant. Increasing the number of simultaneously presented eyes from one to three is correlated with an increasing trend in aversiveness which is destroyed if the same stimuli are presented within the context of a simple head outline. Changing the orientation of a pair of eyes from horizontal to vertical only slightly reduces their aversiveness. The size of the patterns appears to be unimportant within the range tested. The aversive properties of eyespots and of broadcast starling distress calls are positively additive. The effectiveness of a stimulus combining all the important features mentioned is as aversive as is a realistic model of the head of an owl.     

Development of ‘strategies’ utilized by albino rats in operant conditioning

A study concerning an operant conditioning of albino rats over a period of 6 months showed that different subjects progressively adopted a highly stable behaviour which could be called a ‘strategy’. Each one of these ‘strategies’ enables the subject to obtain, in a defined manner, a certain number of reinforcements within a given period; despite their different efficiencies, each one of these ‘strategies’ can be considered as ‘correct’. We have shown here that the duration of the conditioning session and the mean yield per stimulus depend on the choice of a given ‘strategy’.     

Mapping the complexity of ecological models

We propose to define the complexity of an ecological model as the statistical complexity of the output it produces. This allows for a direct comparison between data and model complexity. Working with univariate time series, we show that this measure ‘blindly’ discriminates among the different dynamical behaviours a model can exhibit. We then search a model parameter space in order to segment it into areas of different dynamical behaviour and calculate the maximum complexity a model can generate. Given a time series, and the problem of choosing among a number of ecological models to study it, we suggest that models whose maximum complexity is lower than the time series complexity should be disregarded because they are unable to reconstruct some of the structures contained in the data. Similar reasoning could be used to disregard models’ subdomains as well as areas of unnecessary high complexity. We suggest that model complexity so defined better captures the difficulty faced by a user in managing and understanding the behaviour of an ecological model than measures based on a model ‘size’.     

Towards a theory of thresholds

Some authors studying feeding or drinking have assumed that the activity starts or stops when causal factors rise or fall to a given point. Apparently non-homeostatic behaviour is then explained by adding more complex control mechanisms. Other authors, studying the interaction between activities, have assumed that activities compete so that one activity effectively sets the threshold for another. The latter view was embodied in a ‘model animal’ described previously. The model animal has now been used to simulate rat feeding and drinking behaviour. Although its feedback loops are over-simple, the model exhibits some of the non-homeostatic behaviour shown by rats and it is argued that the complexities arising from competition between activities need to be understood before further progress can be made in exploring homeostatic mechanisms.     

Paradigms of expected failure

This study explores the outsourcing and offshoring of clinical trials and how they interconnect with the dynamics of drug development and regulation in the United States. I focus on the activities of United States-based contract research organizations, which make up a specialized global clinical trials industry focusing on the recruitment of human subjects and investigators. Tracking this industry’s activities in eastern Europe and Latin America, two clinical trial market ‘growth regions,’ I address the strategies of evidence-making that inform clinical trial offshoring. I also show how aspects of the clinical trial model—in which failures to predict safety outcomes or a paradigm of expected failure—are being exported along with the offshored trial. The clinical trials industry is a crucial, highly mobile, and profitable arm of the global pharmaceutical industry. Where state agencies furnish limited or no health care, drug developers claim that trial expansion and experiments have become social goods in themselves. But questions remain: How is drug value and research integrity maintained? And how do the results of clinical trials strengthen or undermine the delivery of affordable and effective interventions? As this essay shows, clinical trials are not only hypothesis-testing instruments; they are operative environments redistributing resources and occasioning tense medical and social fields. In highlighting the inefficiencies and uncertainties of global drug development, this study points to problems in the operational model of drug development and in systems of human protection. It also considers new forms of accountability at the nexus of private sector science and public health.     

Spontaneous light or dark preference in albino mice?

The present experiment examined spontaneous visual choice behaviour and acquisition of a positively reinforced visual discrimination task in Swiss albino mice. In experiment I animals were given 4 consecutive trials in which they could freely enter either a dimly illuminated or a darkened arm of a Y-maze; the position of the light stimulus was randomized across trials. D groups and L groups were tested during the dark and the light period of the day respectively. Results revealed a significant spontaneous preference for the illuminated arm of the maze, independent of the testing period. It is suggested that the dim light has a reinforcing value because it provides additional information about a novel environment. In a second experiment an appetitive visual discrimination task was carried out in the same Y-maze. After a pretraining period, half the animals were reinforced in the illuminated arm and half were reinforced in the darkened one, on five consecutive days. On the first test session all groups of animals chose the illuminated arm significantly more frequently, whereas light/dark choices reached chance level on the last test session. Discrimination learning was not acquired and a behavioural analysis revealed an increasing tendency to a side preference across testing.