OKN眼动跟踪在运动文字识别中的作用

用运动文字的阅读眼动实验来研究运动图像识别与眼动控制的关系,并与一般OKN眼动及一般正常阅读进行对比,探讨了速度及位置信息处理与内容信息处理的关系.实验结果表明;(1)一般正常文字阅读的眼动是Saccades眼动与注视停顿;运动文字识别阅读的眼动中没有注视停顿,而是快、慢交替的OKN眼动,在其慢相期间即运动文字与视网膜相对静止期间采集内容信息进行处理,慢相期间既处理文字内容信息又处理运动速度信息,说明对运动速度与内容信息的处理是并行的.(2)在运动文字运动速度高达80°/s以上时,已不能阅读甚至不能识别单字意义,但仍可产生OKN眼动;这一方面证实阅读速度的受限不在于眼球运动的跟踪能力,而在于高级识别中枢的解码速度,另一方面也说明OKN眼动不是在识别后才产生,而是进行运动图象识别的必要条件.(3)运动文字识别阅读的速度不低于一般正常文字阅读的速度.本文的结果还证实OKN眼动的快相眼动有别于Foveating Saccades.     

Abnormal Fixational Eye Movements in Amblyopia

Purpose

Fixational saccades shift the foveal image to counteract visual fading related to neural adaptation. Drifts are slow eye movements between two adjacent fixational saccades. We quantified fixational saccades and asked whether their changes could be attributed to pathologic drifts seen in amblyopia, one of the most common causes of blindness in childhood.

Methods

Thirty-six pediatric subjects with varying severity of amblyopia and eleven healthy age-matched controls held their gaze on a visual target. Eye movements were measured with high-resolution video-oculography during fellow eye-viewing and amblyopic eye-viewing conditions. Fixational saccades and drifts were analyzed in the amblyopic and fellow eye and compared with controls.

Results

We found an increase in the amplitude with decreased frequency of fixational saccades in children with amblyopia. These alterations in fixational eye movements correlated with the severity of their amblyopia. There was also an increase in eye position variance during drifts in amblyopes. There was no correlation between the eye position variance or the eye velocity during ocular drifts and the amplitude of subsequent fixational saccade. Our findings suggest that abnormalities in fixational saccades in amblyopia are independent of the ocular drift.

Discussion

This investigation of amblyopia in pediatric age group quantitatively characterizes the fixation instability. Impaired properties of fixational saccades could be the consequence of abnormal processing and reorganization of the visual system in amblyopia. Paucity in the visual feedback during amblyopic eye-viewing condition can attribute to the increased eye position variance and drift velocity.     

Influence of Retinal Image Shifts and Extra-Retinal Eye Movement Signals on Binocular Rivalry Alternations

Previous studies have indicated that saccadic eye movements correlate positively with perceptual alternations in binocular rivalry, presumably because the foveal image changes resulting from saccades, rather than the eye movement themselves, cause switches in awareness. Recently, however, we found evidence that retinal image shifts elicit so-called onset rivalry and not percept switches as such. These findings raise the interesting question whether onset rivalry may account for correlations between saccades and percept switches.We therefore studied binocular rivalry when subjects made eye movements across a visual stimulus and compared it with the rivalry in a ‘replay’ condition in which subjects maintained fixation while the same retinal displacements were reproduced by stimulus displacements on the screen. We used dichoptic random-dot motion stimuli viewed through a stereoscope, and measured eye and eyelid movements with scleral search-coils.Positive correlations between retinal image shifts and perceptual switches were observed for both saccades and stimulus jumps, but only for switches towards the subjects'' preferred eye at stimulus onset. A similar asymmetry was observed for blink-induced stimulus interruptions. Moreover, for saccades, amplitude appeared crucial as the positive correlation persisted for small stimulus jumps, but not for small saccades (amplitudes < 1°). These findings corroborate our tenet that saccades elicit a form of onset rivalry, and that rivalry is modulated by extra-retinal eye movement signals.     

Modeling dynamical interactions between fast and slow movements: fast saccadic eye movement behavior in the presence of the slower VOR

An ongoing controversy has to do with the interactions between “fast” (saccadic, quick phase) and “slow” (all other) eye movements. By attacking such issues with both experimental and especially simulation studies using our nonlinear sixth order reciprocally innervated model of the eye mechanical system, insights can be gained into the nature of these nontrivial phenomena. In our present study we relied both (1) on simulation of saccades under a wide range of experimental conditions [vestibular ocular reflex (VOR) velocities from -100 to 100 deg/sec, VOR induced position ranges from -30 to 30 degrees, time-optimal saccades ranging from 2 to 40 degrees], and (2) on using a wide variety of computer simulation of eye movement models, ranging from nonlinear ones with first and especially second order multipulse step controller signal structures, to different controller signal interaction schemes, to simulation using linearized models. We have isolated two important nonlinear phenomena: a level I nonlinear mechanical interaction, dependent not only on the initial velocity but also on the “position effect,” a new finding; and a level II nonlinear neurological interaction, close to “squelching” of the VOR controller signals by the dominating saccadic signal. Furthermore, we have used our simulation findings to reinterpret others' experimental data on eye movement interactions, including saccadic-smooth pursuit, saccadic-vergence, and vestibular nystagmus.     

Automatic detection and removal of fast phases from nystagmographic recordings by optimal thresholding

Recording of ocular nystagmus during vestibular tests does not measure the true response of the vestibulo-ocular reflex (VOR), because the VOR response (so-called slow phase of nystagmus) is interrupted by resetting saccades (so-called fast phase of nystagmus). In order to extract the real VOR contribution, saccades must be removed. In most of the nystagmus processing algorithms, saccade removal requires a human operator to choose a suitable eye velocity threshold able to separate fast from slow nystagmus phases. In the present report a fully automatic removal system is presented which selects an optimal velocity threshold by computing the VOR frequency response and maximizing its coherence function.     

Multiple imputation for discrete data: Evaluation of the joint latent normal model

Missing data are ubiquitous in clinical and social research, and multiple imputation (MI) is increasingly the methodology of choice for practitioners. Two principal strategies for imputation have been proposed in the literature: joint modelling multiple imputation (JM‐MI) and full conditional specification multiple imputation (FCS‐MI). While JM‐MI is arguably a preferable approach, because it involves specification of an explicit imputation model, FCS‐MI is pragmatically appealing, because of its flexibility in handling different types of variables. JM‐MI has developed from the multivariate normal model, and latent normal variables have been proposed as a natural way to extend this model to handle categorical variables. In this article, we evaluate the latent normal model through an extensive simulation study and an application on data from the German Breast Cancer Study Group, comparing the results with FCS‐MI. We divide our investigation in four sections, focusing on (i) binary, (ii) categorical, (iii) ordinal, and (iv) count data. Using data simulated from both the latent normal model and the general location model, we find that in all but one extreme general location model setting JM‐MI works very well, and sometimes outperforms FCS‐MI. We conclude the latent normal model, implemented in the R package jomo , can be used with confidence by researchers, both for single and multilevel multiple imputation.     

A random walk model of fast-phase timing during optokinetic nystagmus

 Most vertebrate animals produce optokinetic nystagmus in response to rotation of their visual surround. Nystagmus consists of an alternation of slow-phase eye rotations, which follow the surround, and fast-phase eye rotations, which quickly reset eye position. The time intervals between fast phases vary stochastically, even during optokinetic nystagmus produced by constant velocity rotation of a uniform surround. The inter-fast-phase interval distribution has a long tail, and intervals that are long relative to the mode become even more likely as constant surround velocity is decreased. This paper provides insight into fast-phase timing by showing that the process of fast-phase generation during constant velocity optokinetic nystagmus is analogous to a random walk with drift toward a threshold. Neurophysiologically, the output of vestibular nucleus neurons, which drive the slow phase, would approximate a random walk with drift because they integrate the noisy, constant surround velocity signal they receive from the visual system. Burst neurons, which fire a burst to drive the fast phase and reset the slow phase, are brought to threshold by the vestibular nucleus neurons. Such a nystagmic process produces stochastically varying inter-fast-phase intervals, and long intervals emerge naturally because, as drift rate (related to surround velocity) decreases, it becomes more likely that any random walk can meander for a long time before it crosses the threshold. The theoretical probability density function of the first threshold crossing times of random walks with drift is known to be that of an inverse Gaussian distribution. This probability density function describes well the distributions of the intervals between fast phases that were either determined experimentally, or simulated using a neurophysiologically plausible neural network model of fast-phase generation, during constant velocity optokinetic nystagmus. Received: 1 June 1995/Accepted in revised form: 15 February 1996     

The assembly of microtubule protein in vitro. The kinetic role in microtubule elongation of oligomeric fragments containing microtubule-associated proteins.

The kinetics of assembly were studied for bovine and pig microtubule protein in vitro over a range of conditions of pH, temperature, nucleotide and protein concentration. The kinetics are in general biphasic with two major processes of similar amplitude but separated in rate by one order of magnitude. Rates and amplitudes are complex functions of solution conditions. The rates of the fast phase and the slow phase attain limiting values as a function of increasing protein concentration, and are more stringently limited at pH 6.5 than pH 6.95. Such behaviour indicates that mechanisms other than the condensation polymerization of tubulin dimer become rate-limiting at higher protein concentration. The constancy of the wavelength-dependence of light-scattering and ultrastructural criteria indicate that microtubules of normal morphology are formed in both phases of the assembly process. Electrophoretic analysis of assembling microtubule protein shows that MAP- (microtubule-associated-protein-)rich microtubules are formed during the fast phase. The rate of dissociation of oligomeric species on dilution of microtubule protein closely parallels the fast-phase rate in magnitude and temperature-dependence. We propose that the rate of this process constitutes an upper limit to the rate of the fast phase of assembly. The kinetics of redistribution of MAPs from MAP-rich microtubules may be a factor limiting the slow-phase rate. A working model is derived for the self-assembly of microtubule protein incorporating the dissociation and redistribution mechanisms that impose upper limits to the rates of assembly attainable by bimolecular addition reactions. Key roles are assigned to MAP-containing fragments in both phases of microtubule elongation. Variations in kinetic behaviour with solution conditions are inferred to derive from the nature and properties of fragments formed from oligomeric species after the rapid temperature jump. The model accounts for the limiting rate behaviour and indicates experimental criteria to be applied in evaluating the relative contributions of alternative pathways.     

Latent period of antisaccades in various conditions of visual stimulation in man

The latent periods (LP) of normal saccades and antisaccades were studied in 10 right-handed healthy subjects in two series of experiments. Peripheral visual stimuli were located at an angle of 10 degrees with respect to the central fixation stimulus in the left and right visual semifields. Two standard schemes of visual stimulation: 1) SS (single step), i.e., switching the peripheral stimulus on immediately after switching the central stimulus of; 2) GAP, i.e., the same with the interstimulus interval in 200 ms. It was shown that in the GAP stimulation condition, the LP of both saccades and antisaccades was 30-50 shorter than in the SS condition. The LP of antisaccades was longer than that of saccades by 145-300 ms. The LP of the leftward antisaccades was by 10-100 ms shorter than that of the rightward ones. Probably, this phenomenon reflects the dominance of the right hemisphere in spatial attention.     

Random walks, random sequences, and nonlinear dynamics in human optokinetic nystagmus.

Optokinetic nystagmus (OKN) is a reflexive eye movement with target-following slow phases (SP) alternating with oppositely directed fast phases (FP). We measured the following from OKN in three humans: FP beginning and ending positions, amplitudes, and intervals and SP amplitudes and velocities. We sought to predict future values of each parameter on the basis of past values, using state-space representation of the sequence (time-delay embedding) and local second-order approximation of trajectories. Predictability is an indication of determinism: this approach allows us to investigate the relative contributions of random and deterministic dynamics in OKN. FP beginning and ending positions showed good predictability, but SP velocity was less predictable. FP and SP amplitudes and FP intervals had little or no predictability. FP beginnings and endings were as predictable as randomized versions that retain linear autocorrelation; this is typical of random walks. Predictability of FP intervals did not change under random rearrangement, which is characteristic of a random process. Only linear determinism was demonstrated; nonlinear interactions may exist that would not be detected by our present approach.     

The role of background movement in goldfish vision

In experiments described in the literature objects presented to restrained goldfish failed to induce eye movements like fixation and/or tracking. We show here that eye movements can be induced only if the background (visual surround) is not stationary relative to the fish but moving. We investigated the influence of background motion on eye movements in the range of angular velocities of 5–20° s⁻¹. The response to presentation of an object is a transient shift in mean horizontal eye position which lasts for some 10 s. If an object is presented in front of the fish the eyes move in a direction such that it is seen more or less symmetrically by both eyes. If it is presented at ±70° from the fish's long axis the eye on the side of the object moves in the direction that the object falls more centrally on its retina. During these object induced eye responses the typical optokinetic nystagmus of amplitude of some 5° with alternating fast and slow phases is maintained, and the eye velocity during the slow phase is not modified by presentation of the object. Presenting an object in front of stationary or moving backgrounds leads to transient suppression of respiration which shows habituation to repeated object presentations. Accepted: 14 April 2000     

A model of the smooth pursuit eye movement system

Human, horizontal, smooth-pursuit eye movements were recorded by the search coil method in response to Rashbass step-ramp stimuli of 5 to 30 deg/s. Eye velocity records were analyzed by measuring features such as the time, velocity and acceleration of the point of peak acceleration, the time and velocity of the peaks and troughs of ringing and steady-state velocity. These values were averaged and mean responses reconstructed. Three normal subjects were studied and their responses averaged. All showed a peak acceleration-velocity saturation. All had ringing frequencies near 3.8 Hz and the mean steady-state gain was 0.95.It is argued that a single, linear forward path with any transfer function G(s) and a 100 ms delay (latency) cannot simultaneously simulate the initial rise of acceleration and ring at 3.8 Hz based on a Bode analysis. Also such a simple negative feedback model cannot have a steady-state gain greater than 1.0; a situation that occurs frequently experimentally. L.R. Young's model, which employs internal positive feedback to eliminate the built-in unity negative feedback, was felt necessary to resolve this problem and a modification of that model is proposed which simulates the data base. Acceleration saturation is achieved by borrowing the idea of the local feedback model for saccades so that one nonlinearity can account for the acceleration-velocity saturation: the main sequence for pursuit. Motor plasticity or motor learning, recently demonstrated for pursuit, is also incorporated and simulated.It was noticed that the offset of pursuit did not show the ringing seen in the onset so this was quantified in one subject. Offset velocity could be characterized by a single exponential with a time constant of about 90 ms. This observation suggests that fixation is not pursuit at zero velocity and that the pursuit system is turned on when needed and off during fixation.     

Saccadic head movements of the praying mantis, with particular reference to visual and proprioceptive information

ABSTRACT. Horizontal head movements of the praying mantis, Sphodromantis lineola Burm., were recorded continuously. They responded to the presence of a live blowfly prey in the antero-lateral visual field with a rapid saccadic head movement. The angular movement of a fixation saccade was correlated positively to the displacement of the prey from the prothoracic midline. Saccade magnitude and velocity are related. After the stimulus moved out of the visual field, the mantis made a second saccadic head movement, a return saccade towards the body midline. We observed return saccades in which the head overshot or undershot the body midline, as well as saccades which returned the head exactly to its initial position. In 92% of trials with intact mantids, the return movement succeeded eventually in rotating the head back to its initial position, whereas after removal of the neck hair plates this occurred in only 47% of trials. There is a consistent relation between saccade extent and velocity. Velocities of return saccades were slower than those of fixation saccades. It is suggested that sensory inputs from the neck hair plate proprioceptors modify both the magnitude and the angular velocity of fixation and return saccadic head movements.     

Evoked potentials of the cat inferior colliculus to acoustic stimuli simulating sound source movement with different velocities in opposite directions

Amplitude changes of inferior colliculus evoked potentials (EPs) in anaesthetized adult cats were studied under presentation of acoustic stimuli simulating both azimuth-moving and stationary sound source. The movement was simulated with gradual changes of interaural time delay between binaurally presented click trains. It was shown that the amplitude of EPs elicited by "moving" signals depended on the velocity of movement. Amplitude differences between EPs to "moving" and stationary stimuli were observed under motion velocities up to 320 deg./s. The greatest response amplitudes in different experiments took place under velocities within the range of 67-320 deg./s with most of them recorded under velocities of 170 and 125 deg./s. Amplitude of the responses to lateral-medial movement with any velocity were always greater than those to opposite direction of movement with the same velocity.     

Unequal saccades generated by velocity interactions in the peripheral oculomotor system

Experiments with precision eye movement recordings show binocularly unequal saccades to be present under several stimulus conditions having as a common theme ongoing low velocities at the times of the saccades. Simulations using a model of eye muscles and eyeball dynamics reproduce these unequal saccades in quantitative agreement with the experimental findings. The model uses equal innervation for the saccades, and demonstrates a peripheral interaction between the muscle forces and the eye velocities to be the cause of the large inequality of the simulated binocular saccades. Thus, the simulations provide evidence that Hering's law continues to describe the innervation patterns to corresponding muscles producing these binocularly unequal saccades found in the experimental situation.     

Dyslexic children are confronted with unstable binocular fixation while reading

Reading requires three-dimensional motor control: saccades bring the eyes from left to right, fixating word after word; and oblique saccades bring the eyes to the next line of the text. The angle of vergence of the two optic axes should be adjusted to the depth of the book or screen and--most importantly--should be maintained in a sustained manner during saccades and fixations. Maintenance of vergence is important as it is a prerequisite for a single clear image of each word to be projected onto the fovea of the eyes. Deficits in the binocular control of saccades and of vergence in dyslexics have been reported previously but only for tasks using single targets. This study examines saccades and vergence control during real text reading. Thirteen dyslexic and seven non-dyslexic children read the French text "L'Allouette" in two viewing distances (40 cm vs. 100 cm), while binocular eye movements were measured with the Chronos Eye-tracking system. We found that the binocular yoking of reading saccades was poor in dyslexic children (relative to non-dyslexics) resulting in vergence errors; their disconjugate drift during fixations was not correlated with the disconjugacy during their saccades, causing considerable variability of vergence angle from fixation to fixation. Due to such poor oculomotor adjustments during reading, the overall fixation disparity was larger for dyslexic children, putting larger demand on their sensory fusion processes. Moreover, for dyslexics the standard deviation of fixation disparity was larger particularly when reading at near distance. We conclude that besides documented phoneme processing disorders, visual/ocular motor imperfections may exist in dyslexics that lead to fixation instability and thus, to instability of the letters or words during reading; such instability may perturb fusional processes and might--in part--complicate letter/word identification.     

Positive potentials of the human brain at different stages of preparation of a visually triggered saccade

The EEG of 10 right-handed subjects preceding saccades with mean values of latent periods were selected and averaged. Two standard paradigms of presentation of visual stimuli (central fixation stimulus-peripheral target succession): with a 200-ms inerstimulus interval (GAP) and successive single step (SS). During the period of central fixation, two kinds of positive potentials were observed: fast potentials of "inermediate" positivity (IP) developing 600-400 ms prior to saccade onset and fast potentials of "leading" positivity (LP), which immediately preceded the offset of the central fixation stimulus. Peak latency of the LP potentials was 300 ms prior to saccade onset in the SS paradigm and 400 ms in the GAP paradigm. These potentials were predominantly recorded in the frontal and frontosagittal cortical areas. Decrease in the latency by 30-50 ms in the GAP paradigm was associated with more pronounced positive potentials during the fixation period and absence of the initiation potential P-1' (or decrease in its amplitude). The obtained evidence suggest that the fast positive presaccadic potentials are of a complex nature related to attention, anticipation, motor preparation, decision making, saccadic initiation, and backward afferentation.     

Open-loop simulations of the primate saccadic system using burst cell discharge from the superior colliculus

 Saccade-related burst neurons (SRBNs) in the monkey superior colliculus (SC) have been hypothesized to provide the brainstem saccadic burst generator with the dynamic error signal and the movement initiating trigger signal. To test this claim, we performed two sets of open-loop simulations on a burst generator model with the local feedback disconnected using experimentally obtained SRBN activity as both the driving and trigger signal inputs to the model. First, using neural data obtained from cells located near the middle of the rostral to caudal extent of the SC, the internal parameters of the model were optimized by means of a stochastic hill-climbing algorithm to produce an intermediate-sized saccade. The parameter values obtained from the optimization were then fixed and additional simulations were done using the experimental data from rostral collicular neurons (small saccades) and from more caudal neurons (large saccades); the model generated realistic saccades, matching both position and velocity profiles of real saccades to the centers of the movement fields of all these cells. Second, the model was driven by SRBN activity affiliated with interrupted saccades, the resumed eye movements observed following electrical stimulation of the omnipause region. Once again, the model produced eye movements that closely resembled the interrupted saccades produced by such simulations, but minor readjustment of parameters reflecting the weight of the projection of the trigger signal was required. Our study demonstrates that a model of the burst generator produces reasonably realistic saccades when driven with actual samples of SRBN discharges. Received: 25 October 1994/Accepted in revised form: 20 June 1995     

Neural network simulations of the primate oculomotor system III. An one-dimensional, one-directional model of the superior colliculus

This report evaluates the performance of a biologically motivated neural network model of the primate superior colliculus (SC). Consistent with known anatomy and physiology, its major features include excitatory connections between its output elements, nigral gating mechanisms, and an eye displacement feedback of reticular origin to recalculate the metrics of saccades to memorized targets in retinotopic coordinates. Despite the fact that it makes no use of eye position or eye velocity information, the model can account for the accuracy of saccades in double step stimulation experiments. Further, the model accounts for the effects of focal SC lesions. Finally, it accounts for the properties of saccades evoked in response to the electrical stimulation of the SC. These include the approximate size constancy of evoked saccades despite increases of stimulus intensity, the fact that the size of evoked saccades depends on the time that has elapsed from a previous saccade, the fact that staircases of saccades are evoked in response to prolonged stimuli, and the fact that the size of saccades evoked in response to the simultaneous stimulation of two SC sites is the average of the saccades that are evoked when the two sites are separately stimulated. Received: 3 November 1997 / Accepted in revised form: 30 June 1998