Search

Visual search can be facilitated by the learning of spatial configurations that predict the location of a target among distractors. Neuropsychological and functional magnetic resonance imaging (fMRI) evidence implicates the medial temporal lobe (MTL) memory system in this contextual cueing effect, and electroencephalography (EEG) studies have identified the involvement of visual cortical regions related to attention. This work investigated two questions: (1) how memory and attention systems are related in contextual cueing; and (2) how these systems are involved in both short- and long-term contextual learning. In one session, EEG and fMRI data were acquired simultaneously in a contextual cueing task. In a second session conducted 1 week later, EEG data were recorded in isolation. The fMRI results revealed MTL contextual modulations that were correlated with short- and long-term behavioral context enhancements and attention-related effects measured with EEG. An fMRI-seeded EEG source analysis revealed that the MTL contributed the most variance to the variability in the attention enhancements measured with EEG. These results support the notion that memory and attention systems interact to facilitate search when spatial context is implicitly learned.

Rapid serial visual presentation (RSVP) tasks, in which participants are presented with a continuous sequence of images in one location, have been used in combination with electroencephalography (EEG) in a variety of Brain-Machine Interface (BMI) applications. The RSVP task is advantageous because it can be performed at a high temporal rate. The rate of the RSVP sequence is controlled by the stimulus onset asynchrony (SOA) between subsequent stimuli. When used within the context of a BMI, an RSVP task with short SOA could increase the information throughput of the system while also allowing for stimulus repetitions. However, reducing the SOA also increases the perceptual degradation caused by presenting two stimuli in close succession, and it decreases the target-to-target interval (TTI), which can increase the cognitive demands of the task. These negative consequences of decreasing the SOA could affect on the EEG signal measured in the task and degrade the performance of the BMI. Here we systematically investigate the effects of SOA and stimulus repetition (r) on single-trial target detection in an RSVP task. Ten healthy volunteers participated in an RSVP task in four conditions that varied in SOA and repetitions (SOA=500 ms, r=1; SOA=250 ms, r=2; SOA=166 ms, r=3; and SOA=100 ms, r=5) while processing time across conditions was controlled. There were two key results: First, when controlling for the number of repetitions, single-trial performance increases when the SOA decreases. Second, when the repetitions were combined, the best performance (AUC=0.967) was obtained with the shortest SOA (100 ms). These results suggest that shortening the SOA in an RSVP task has the benefit of increasing the performance relative to longer SOAs, and it also allows a higher number of repetitions of the stimuli in a limited amount of time.

Images derived from a “phantom” are useful for characterizing the performance of imaging systems. In particular, the modulation transfer properties of imaging detectors are traditionally assessed by physical phantoms consisting of an edge. More recently researchers have come to realize that quantifying the effects of object variability can also be accomplished with phantoms in modalities such as breast imaging where anatomical structure may be the principal limitation in performance. This has driven development of virtual phantoms that can be used in simulation environments. In breast imaging, several such phantoms have been proposed. In this work, we analyze non-Gaussian statistical properties of virtual phantoms, and compare them to similar statistics from a database of breast images. The virtual phantoms assessed consist of three classes. The first is known as clustered-blob lumpy backgrounds. The second class is “binarized” textures which typically apply some sort of threshold to a stochastic 3D texture intended to represent the distribution of adipose and glandular tissue in the breast. The third approach comes from efforts at the University of Pennsylvania to directly simulate the 3D anatomy of the breast. We use Laplacian fractional entropy (LFE) as a measure of the non-Gaussian statistical properties of each simulation. Our results show that the simulation approaches differ considerably in LFE with very low scores for the clustered-blob lumpy background to very high values for the UPenn phantom. These results suggest that LFE may have value in developing and tuning virtual phantom simulation procedures.

In previous studies of saccadic targeting, the issue how visually guided saccades to unambiguous targets are programmed and executed has been examined. These studies have found different degrees of guidance for saccades depending on the task and task difficulty. In this study, we use ideal-observer analysis to estimate the visual information used for the first saccade during a search for a target disk in noise. We quantitatively compare the performance of the first saccadic decision to that of the ideal observer (ie absolute efficiency of the first saccade) and to that of the associated final perceptual decision at the end of the search (ie relative efficiency of the first saccade). Our results show, first, that at all levels of salience tested, the first saccade is based on visual information from the stimulus display, and its highest absolute efficiency is ∼20%. Second, the efficiency of the first saccade is lower than that of the final perceptual decision after active search (with eye movements) and has a minimum relative efficiency of 19% at the lowest level of saliency investigated. Third, we found that requiring observers to maintain central fixation (no saccades allowed) decreased the absolute efficiency of their perceptual decision by up to a factor of two, but that the magnitude of this effect depended on target salience. Our results demonstrate that ideal-observer analysis can be extended to measure the visual information mediating saccadic target-selection decisions during visual search, which enables direct comparison of saccadic and perceptual efficiencies.

The speed of processes underlying lightness constancy was studied by having observers discriminate small differences in simulated lightness under an oscillating illumination. The period of oscillation varied from 0.25 to 120 sec. The target was a 1 degrees square which appeared for 150 msec at random intervals either directly against a uniform background or separated from the background by a 1 degrees dark gap. When the target and background were adjacent to each other, discrimination accuracy approached control levels (fixed illumination) at all but the shortest periods of oscillation. When the gap was introduced, accuracy increased as the period of oscillation increased, but never approached control levels. The results suggest that a fast local contrast mechanism is the primary mediator of lightness constancy for this task, but that there is also a slower mechanism that may be related to adaptation.

Recently, quantitative models based on signal detection theory have been successfully applied to the prediction of human accuracy in visual search for a target that differs from distractors along a single attribute (feature search). The present paper extends these models for visual search accuracy to multidimensional search displays in which the target differs from the distractors along more than one feature dimension (conjunction, disjunction, and triple conjunction displays). The model assumes that each element in the display elicits a noisy representation for each of the relevant feature dimensions. The observer combines the representations across feature dimensions to obtain a single decision variable, and the stimulus with the maximum value determines the response. The model accurately predicts human experimental data on visual search accuracy in conjunctions and disjunctions of contrast and orientation. The model accounts for performance degradation without resorting to a limited-capacity spatially localized and temporally serial mechanism by which to bind information across feature dimensions.

Models of human visual processing start with an initial stage with parallel independent processing of different physical attributes or features (e.g., color, orientation, motion). A second stage in these models is a temporally serial mechanism (visual attention) that combines or binds information across feature dimensions. Evidence for this serial mechanism is based on experimental results for visual search. I conducted a study of visual search accuracy that carefully controlled for low-level effects: physical similarity of target and distractor, element eccentricity, and eye movements. The larger set-size effects in visual search accuracy for briefly flashed conjunction displays, compared with feature displays, are quantitatively predicted by a simple model in which each feature dimension is processed independently with inherent neural noise and information is combined linearly across feature dimensions. The data are not predicted by a temporally serial mechanism or by a hybrid model with temporally serial and noisy processing. The results do not support the idea that a temporally serial mechanism, visual attention, binds information across feature dimensions and show that the conjunction-feature dichotomy is due to the noisy independent processing of features in the human visual system.

Human vision is an active process in which information is sampled during brief periods of stable fixation in between gaze shifts. Foveal analysis serves to identify the currently fixated object and has to be coordinated with a peripheral selection process of the next fixation location. Models of visual search and scene perception typically focus on the latter, without considering foveal processing requirements. We developed a dual-task noise classification technique that enables identification of the information uptake for foveal analysis and peripheral selection within a single fixation. Human observers had to use foveal vision to extract visual feature information (orientation) from different locations for a psychophysical comparison. The selection of to-be-fixated locations was guided by a different feature (luminance contrast). We inserted noise in both visual features and identified the uptake of information by looking at correlations between the noise at different points in time and behavior. Our data show that foveal analysis and peripheral selection proceeded completely in parallel. Peripheral processing stopped some time before the onset of an eye movement, but foveal analysis continued during this period. Variations in the difficulty of foveal processing did not influence the uptake of peripheral information and the efficacy of peripheral selection, suggesting that foveal analysis and peripheral selection operated independently. These results provide important theoretical constraints on how to model target selection in conjunction with foveal object identification: in parallel and independently.

In general, humans tend to first look just below the eyes when identifying another person. Does everybody look at the same place on a face during identification, and, if not, does this variability in fixation behavior lead to functional consequences? In two conditions, observers had their free eye movements recorded while they performed a face-identification task. In another condition, the same observers identified faces while their gaze was restricted to specific locations on each face. We found substantial differences, which persisted over time, in where individuals chose to first move their eyes. Observers’ systematic departure from a canonical, theoretically optimal fixation point did not correlate with performance degradation. Instead, each individual’s looking preference corresponded to an idiosyncratic performance-maximizing point of fixation: Those who looked lower on the face performed better when forced to fixate the lower part of the face. The results suggest an observer-specific synergy between the face-recognition and eye movement systems that optimizes face-identification performance.

Exploiting scene context and object–object co-occurrence is critical in guiding eye movements and facilitating visual search, yet the mediating neural mechanisms are unknown. We used functional magnetic resonance imaging while observers searched for target objects in scenes and used multivariate pattern analyses (MVPA) to show that the lateral occipital complex (LOC) can predict the coarse spatial location of observers' expectations about the likely location of 213 different targets absent from the scenes. In addition, we found weaker but significant representations of context location in an area related to the orienting of attention (intraparietal sulcus, IPS) as well as a region related to scene processing (retrosplenial cortex, RSC). Importantly, the degree of agreement among 100 independent raters about the likely location to contain a target object in a scene correlated with LOC's ability to predict the contextual location while weaker but significant effects were found in IPS, RSC, the human motion area, and early visual areas (V1, V3v). When contextual information was made irrelevant to observers' behavioral task, the MVPA analysis of LOC and the other areas' activity ceased to predict the location of context. Thus, our findings suggest that the likely locations of targets in scenes are represented in various visual areas with LOC playing a key role in contextual guidance during visual search of objects in real scenes.

The detection of event-related potentials (ERPs) in the electroencephalogram (EEG) signal has several real-world applications, from cognitive state monitoring to brain-computer interfaces. Current systems based on the detection of ERPs only consider a single type of response to detect. Hence, the classification methods that are considered for ERP detection are binary classifiers (target vs. non target). Here we investigated multiclass classification of single-trial evoked responses during a rapid serial visual presentation task in which short video clips were presented to fifteen observers. Each trial contained potential targets that were human or non-human, stationary or moving. The goal of the classification analysis was to discriminate between three classes: moving human targets, moving non-human targets, and non-moving human targets. The analysis revealed that the mean volume under the ROC surface of 0.878. These results suggest that it is possible to efficiently discriminate between more than two types of evoked responses using single-trial detection.

Background. The concept of diagnostic utility is a fundamental component of signal detection theory, going back to some of its earliest works. Attaching utility values to the various possible outcomes of a diagnostic test should, in principle, lead to meaningful approaches to evaluating and comparing such systems. However, in many areas of medical imaging, utility is not used because it is presumed to be unknown. Methods. In this work, we estimate relative utility (the utility benefit of a detection relative to that of a correct rejection) for screening mammography using its known relation to the slope of a receiver operating characteristic (ROC) curve at the optimal operating point. The approach assumes that the clinical operating point is optimal for the goal of maximizing expected utility and therefore the slope at this point implies a value of relative utility for the diagnostic task, for known disease prevalence. We examine utility estimation in the context of screening mammography using the Digital Mammographic Imaging Screening Trials (DMIST) data. Results. We show how various conditions can influence the estimated relative utility, including characteristics of the rating scale, verification time, probability model, and scope of the ROC curve fit. Relative utility estimates range from 66 to 227. Conclusions. We argue for one particular set of conditions that results in a relative utility estimate of 162 (±14%). This is broadly consistent with values in screening mammography determined previously by other means. At the disease prevalence found in the DMIST study (0.59% at 365-day verification), optimal ROC slopes are near unity, suggesting that utility-based assessments of screening mammography will be similar to those found using Youden’s index.

When viewing a human face, people often look toward the eyes. Maintaining good eye contact carries significant social value and allows for the extraction of information about gaze direction. When identifying faces, humans also look toward the eyes, but it is unclear whether this behavior is solely a byproduct of the socially important eye movement behavior or whether it has functional importance in basic perceptual tasks. Here, we propose that gaze behavior while determining a person’s identity, emotional state, or gender can be explained as an adaptive brain strategy to learn eye movement plans that optimize performance in these evolutionarily important perceptual tasks. We show that humans move their eyes to locations that maximize perceptual performance determining the identity, gender, and emotional state of a face. These optimal fixation points, which differ moderately across tasks, are predicted correctly by a Bayesian ideal observer that integrates information optimally across the face but is constrained by the decrease in resolution and sensitivity from the fovea toward the visual periphery (foveated ideal observer). Neither a model that disregards the foveated nature of the visual system and makes fixations on the local region with maximal information, nor a model that makes center-of-gravity fixations correctly predict human eye movements. Extension of the foveated ideal observer framework to a large database of real-world faces shows that the optimality of these strategies generalizes across the population. These results suggest that the human visual system optimizes face recognition performance through guidance of eye movements not only toward but, more precisely, just below the eyes.

Visual search and cueing tasks have been employed extensively in attentional research, with each having a standard effect (visual search: set size effects, cueing: cue validity). Generally these effects have been treated with different (but often similar) attentional theories. The present study aims to consolidate cueing and set size effects within an ideal observer approach. Four observers performed a yes/no contrast discrimination of a gaussian signal in a task combining cueing with visual search. The signal appeared in half the trials, and effective set size (M, 2 to 8) was determined by one primary precue (having 50% validity in signal present trials) and M-1 secondary precues. There were two stimulus durations: 1 second (eye movements allowed), and the first-saccade latency (in the 1 second duration condition) minus 80 milliseconds. Simulations found that an ideal observer for the perceptual yes/no decisions and the first saccadic localization decisions predicted both set size and cueing effects with a single weighting mechanism, providing a unifying account. For the human observer results, a modified ideal observer (with performance matched to human performance) fit the yes/no perceptual decisions well. For the first saccadic decisions, there was evidence of use of the primary cue, but the modified ideal observer was not a good fit, indicating a suboptimal use of the cue. We discuss possible underlying assumptions about the task that might explain the apparent suboptimal nature of saccadic decisions and the overall utility of the ideal observer for cueing and visual search studies in visual attention and saccades.

The great amount of slices in volumetric data sets and limited time prevents human observers from exhaustively pointing their high resolution processing fovea to all regions in the images. Thus, many image-regions are processed with nonfoveal peripheral visual processing. Yet, most studies quantifying human detectability of signals in computer simulated textures and medical image backgrounds, have measured performance without consideration of the location of the signal in the observer's eye relative to the fovea (retinal eccentricity). Here, we measure human observer detectability of signals in CT images as a function of retinal eccentricity. A representative signal was extracted from a liver image and was added to healthy liver backgrounds at random positions. The retinal eccentricities of the signal were manipulated by varying the position of the point at which observers fixated with their eyes. Real-time video-based eye tracking was used to ensure steady fixation. High contrast fiduciary marks indicated the only possible location of the signal which was present in 50% of the images. Single CT slices were presented for 200 ms or 1 second. The observer was instructed to decide whether the image contained a signal (yes/no task). We probed 6 eccentricities with 420 decision trials per eccentricity. We found a large detectability degradation with retinal eccentricity with d' degrading by 50% at an eccentricity of 9 degrees for a 200 ms display time.

The detection of event-related potentials in the electroencephalogram signal is a common way for creating a brain-computer interface (BCI). Successful detection of evoked responses can be enhanced by the user selectively attending to specific stimuli presented in the BCI task. Because BCI users need a system that performs well in a variety of contexts, even ones that may impair selective attention, it is critical to understand how single trial detection is affected by attention. We tested 16 participants using a rapid serial visual/auditory presentation paradigm under three conditions, one in which they detected the presence of a visual target, one in which they detected the presence of an auditory target, and one in which they detected both visual and auditory targets. The behavioral performance indicates that the visual task was more difficult than the auditory task. Consistent with the higher behavioral difficulty of the visual task, single trial performance showed no difference between single and dual-task for the visual target detection (mean=0.76). However, the area under the curve for the auditory target detection was significantly lower than the dual-task (mean=0.81 for single task, 0.75 for dual-task). The results support the conclusion that single-trial target detection is impaired when attention is divided between multiple tasks.

In non-invasive brain-computer interface (BCI), the analysis of event-related potentials (ERP) has typically focused on averaged trials, a current trend is to analyze single-trial evoked response individually with new approaches in pattern recognition and signal processing. Such single trial detection requires a robust response that can be detected in a variety task conditions. Here, we investigated the influence of target probability, a key factor known to influence the amplitude of the evoked response, on single trial target classification in a difficult rapid serial visual presentation (RSVP) task. Our classification approach for detecting target vs. non target responses, considers spatial filters obtained through the maximization of the signal to signal-plus-noise ratio, and then uses the resulting information as inputs to a Bayesian discriminant analysis. The method is evaluated across eight healthy subjects, on four probability conditions (P=0.05, 0.10, 0.25, 0.50). We show that the target probability has a statistically significant effect on both the behavioral performance and the target detection. The best mean area under the ROC curve is achieved with P=0.10, AUC=0.82. These results suggest that optimal performance of ERP detection in RSVP tasks is critically dependent on target probability.

It is generally well known that the appearance of breast tissue in a mammogram is considerably more complex in a statistical sense than a simple random Gaussian texture, even when the correlation structure of the Gaussian has been set to match the power-law power spectrum of mammograms. However there has not been a systematic way to characterize the extent of departure from a Gaussian process. We address this topic here by proposing a noisy-Laplacian distribution to model response histograms derived from digital (or digitized) mammograms. We describe the distribution in terms of the probability density function and cumulative density function, as well as moments up to fourth order. We also demonstrate the usefulness of the new distribution by fitting it to responses from digital mammography.

The application of multivariate techniques to neuroimaging and electrophysiological data has greatly enhanced the ability to detect where, when, and how functional neural information is processed during a variety of behavioral tasks. With the extension to single-trial analysis, neuroscientists are able to relate brain states to perceptual, cognitive, and motor processes. Using pattern classification methods, the neuroscientist can extract neural performance measures in a manner analogous to human behavioral performance, allowing for a consistent information content metric across measurement modalities. However, as with behavioral psychophysical performance, pattern classifier performances are a product of both the task-relevant information inherent in the brain and in the task/stimuli. Here, we argue for the use of an ideal observer framework with which the researcher can effectively normalize the observed neural performance given the task’s inherent objective difficulty. We use data from a face versus car discrimination task and compare classifier performance applied to electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) data with corresponding human behavior through the absolute and relative efficiency metrics. We show that confounding variables that can lead to erroneous interpretations of information content can be accounted for through comparisons to an ideal observer, allowing for more confident interpretation of the neural mechanisms involved in the task of interest. Finally, we discuss limitations of interpretation due to the transduction of indirect measures of neural activity, underlying assumptions in the optimality of the pattern classifiers, and dependence of efficiency results on signal contrast.

When an investigator is developing a new image-processing technique or manipulating an image-acquisition technique they are confronted with the question of whether the new technique will improve clinical diagnosis. A first approach is to look at individual physical properties of the image such as image contrast and resolution. Although these properties might be useful, it has long been known that the noise characteristics of the image system need to be taken into consideration to appropriately evaluate the quality of an image whether it will be used to detect, classify, and/ˆ•or estimate a signal (Cunningham and Shaw, 1999). One useful measure of the noise characteristics is the noise-equivalent quanta (NEQ) that expresses the image noise in terms of the number of Poisson-distributed input photons per unit area at each spatial frequency (Wagner and Brown, 1985). The NEQ can be thought of as a measure inversely related to the amount of noise as a function of spatial frequency. However, when the diagnostic decision involves a human observer, medical image quality can be defined in terms of human performance in visual tasks that are relevant to clinical diagnosis (Barrett, 1993). In this context, receiver operating characteristics (ROC) studies are the standard method of evaluating the impact of a particular image manipulation on clinical diagnosis. In these studies, the physicians scrutinize a set of medical images (under the different image-acquisition or processing conditions) and rate their confidence about the presence of the lesion. The investigator infers from these ratings a measure of performance known as the area under the ROC curve. Often, the number of possible conditions is large and ROC studies become time consuming and costly because they require a large number of human observations. Other times, the investigator might want to optimize a parameter or a set of parameters. In such cases, the number of conditions suffers a combinatorial explosion, and therefore ROC studies become unfeasible. Thus it is desirable to develop a metric of image quality that could be used for fast evaluation and optimization of image quality but also would have the predictive power of ROC studies. Computer-model observers are algorithms that attempt to predict human visual performance in noisy images and might represent the desired metric of image quality when the diagnostic decision involves a human observer and a visual task. Development of models to predict human visual signal detection in noise goes back to work by Rose (1948) who studied the detectability of a flat-topped disk embedded in white noise (see Burgess, 1999a, for a review). In the last two decades, many studies have concentrated on finding a model observer that can predict human performance across many types of synthetic backgrounds. More recently, model observers have been applied to real medical-image backgrounds. The hope is that eventually model observers will become common metrics of task-based image quality for evaluation of medical-image quality as well as optimization of imaging systems.

We investigate signal-known-exactly (SKE) detection and free localization tasks in correlated noise having a power-law power spectrum. In all cases the target is an additive focal Gaussian "bump" signal. We compare the performance of human observers in psychophysical studies to that of the Bayesian ideal observer for both tasks in the form of the observer efficiency. We find efficiencies that range from 40% to 60% in the SKE detection task, consistent with previous works. Observer efficiency is considerably higher in the free-localization task, ranging from 60% to 80%. Direct estimation of the human-observer spatial-frequency weights shows clear evidence of a shift to higher frequencies with increasing power-law exponent in both tasks. Our results suggest that human observers are performing some sort of limited prewhitening of stimuli in both tasks.

As an important clinical task, evaluating the placement of multiple coronary stents requires fine judgments of distance between stents. However, making these judgments is limited by low system resolution, noise, low contrast of the deployed stent, and stent motion during the cardiac cycle. We use task performance as a figure of merit for optimizing image display parameters. In previous work, we described our simulation procedure in detail, and also reported results of human observers for a visual task involving discrimination of 4 gap sizes under various frame rates and number of frames. Here, we report the results of three spatial model observers (i.e. NPW, NPWE, and PWMF) and two temporal sensitivity functions (i.e. transient and sustained) for the same task. Under signal known exactly conditions, we find that model observers can be used to predict human observers in terms of discrimination accuracy by adding internal noise.

We investigate the potential of using EEG recordings of observers performing a rapid visual categorization task for person identification. We examine a 0.5 s epoch of EEG data using machine learning techniques that, unlike most previous studies, analyze the data in a holistic manner and extracts discriminative spatio-temporal filters. The analysis of the filters suggest sparse feature representation spatially as well as temporally. The filters reveal that the neural activity that discriminates individuals is spatially localized to occipital electrodes located on the scalp above the visual cortex and temporally localized in the interval of 120–200 ms after presentation of the visual stimulus. The results demonstrate the feasibility of EEG-based person identification based on difficult perceptual tasks.

We compared the ability of three model observers (nonprewhitening matched filter with an eye filter, Hotelling and channelized Hotelling) in predicting the effect of JPEG and wavelet-Crewcode image compression on human visual detection of a simulated lesion in single frame digital x-ray coronary angiograms. All three model observers predicted the JPEG superiority present in human performance, although the nonprewhitening matched filter with an eye filter (NPWE) and the channelized Hotelling models were better predictors than the Hotelling model. The commonly used root mean square error and related peak signal to noise ratio metrics incorrectly predicted a JPEG inferiority. A particular image discrimination/perceptual difference model correctly predicted a JPEG advantage at low compression ratios but incorrectly predicted a JPEG inferiority at high compression ratios. In the second part of the paper, the NPWE model was used to perform automated simulated annealing optimization of the quantization matrix of the JPEG algorithm at 25:1 compression ratio. A subsequent psychophysical study resulted in improved human detection performance for images compressed with the NPWE optimized quantization matrix over the JPEG default quantization matrix. Together, our results show how model observers can be successfully used to perform automated evaluation and optimization of diagnostic performance in clinically relevant visual tasks using real anatomic backgrounds.

We evaluate the effect on human visual detection of a display method that stabilizes the motion of the feature of interest in sequential JPEG compressed x-ray coronary angiogram images. At all levels of image compression the feature stabilized display significantly improved performance with respect to the standard display where the artery is moving. In addition, for both the moving artery and stabilized display, human performance with images compressed at 15:1 was not significantly different from performance with the uncompressed images.

The placement of multiple coronary stents requires fine judgments of distance between a deployed stent and stent/guidewire assembly. The goal of this deployment is to achieve continuous and gapless coverage between them. However, making these judgments is difficult because of limited system resolution, noise, relatively low contrast of the deployed stent, and stent motion during the cardiac cycle. In this work, we extend our previous work by investigating wider range of conditions associated with this task. The present studies consider number of frames and frame rate separately, and include stabilization of the stents as a way to quantify the performance effects of stent motion. We find that (1) stabilization reduces the uncertainty when detecting the gap size; (2) observer performance increases with the number of frames; (3) the effect of display frame rate is highly dependent on the motion of the target.

Researchers studying human and computer vision have found description and construction of these systems greatly aided by analysis of the statistical properties of naturally occurring scenes. More specifically, it has been found that receptive fields with directional selectivity and bandwidth properties similar to mammalian visual systems are more closely matched to the statistics of natural scenes. It is argued that this allows for sparse representation of the independent components of natural images [Olshausen and Field, Nature, 1996]. These theories have important implications for medical image perception. For example, will a system that is designed to represent the independent components of natural scenes, where objects occlude one another and illumination is typically reflected, be appropriate for X-ray imaging, where features superimpose on one another and illumination is transmissive? In this research we begin to examine these issues by evaluating higher-order statistical properties of breast images from X-ray projection mammography (PM) and dedicated breast computed tomography (bCT). We evaluate kurtosis in responses of octave bandwidth Gabor filters applied to PM and to coronal slices of bCT scans. We find that kurtosis in PM rises and quickly saturates for filter center frequencies with an average value above 0.95. By contrast, kurtosis in bCT peaks near 0.20 cyc/mm with kurtosis of approximately 2. Our findings suggest that the human visual system may be tuned to represent breast tissue more effectively in bCT over a specific range of spatial frequencies.

In this study, we adapt and apply model observers within the framework of realistic detection tasks in breast tomosynthesis (BT). We use images consisting of realistic masses digitally embedded in real patient anatomical backgrounds, and we adapt specific model observers that have been previously applied to digital mammography (DM). We design alternative forced-choice experiments (AFC) studies for DM and BT tasks in the signal known exactly but variable (SKEV) framework. We compare performance of various linear model observers (non-prewhitening matched filter with an eye filter, and several channelized Hotelling observers (CHO) against human. A good agreement in performance between human and model observers can be obtained when an appropriate internal noise level is adopted. Models achieve the same detection performance across BT and DM with about three times less projected signal intensity in BT than in DM (humans: 3.8), due to the anatomical noise reduction in BT. We suggest that, in the future, model observers can potentially be used as an objective tool for automating the optimization of BT acquisition parameters or reconstruction algorithms, or narrowing a wide span of possible parameter combinations, without requiring human observers studies.

We studied the influence of signal variability on human and model observers for detection tasks with realistic simulated masses superimposed on real patient mammographic backgrounds and synthesized mammographic backgrounds (clustered lumpy backgrounds, CLB). Results under the signal-known-exactly (SKE) paradigm were compared with signal-known-statistically (SKS) tasks for which the observers did not have prior knowledge of the shape or size of the signal. Human observers' performance did not vary significantly when benign masses were superimposed on real images or on CLB. Uncertainty and variability in signal shape did not degrade human performance significantly compared with the SKE task, while variability in signal size did. Implementation of appropriate internal noise components allowed the fit of model observers to human performance.

Synthetic yet realistic images are valuable for many applications in visual sciences and medical imaging. Typically, investigators develop algorithms and adjust their parameters to generate images that are visually similar to real images. In this study, we used a genetic algorithm and an objective, statistical similarity measure to optimize a particular texture generation algorithm, the clustered lumpy backgrounds (CLB) technique, and synthesize images mimicking real mammograms textures. We combined this approach with psychophysical experiments involving the judgment of radiologists, who were asked to qualify the visual realism of the images. Both objective and psychophysical approaches show that the optimized versions are significantly more realistic than the previous CLB model. Anatomical structures are well reproduced, and arbitrary large databases of mammographic texture with visual and statistical realism can be generated. Potential applications include detection experiments, where large amounts of statistically traceable yet realistic images are needed. (C) 2008 Optical Society of America.

Evaluating the placement of multiple coronary stents requires fine judgments of distance between two or more deployed stents in order to determine if there is continuous coverage without a gap or overlap between the two. These judgments are made difficult by limited system resolution, noise, relatively low contrast of the deployed stent, and stent motion during the cardiac cycle. In this work, we assess the effect of frame rate and number of frames used in a sequence on the detection accuracy of gaps between the stents. Both of these factors can be used to reduce patient dose. We use real X-ray coronary angiograms as backgrounds along with stents imaged separately with Lucite for similar beam attenuation. Stents and simulated guidewires are embedded in the angiograms by adding optical densities after scatter subtraction. Realistic motion is rendered by manually synchronizing the stent densities to vascular features in each image. We find no significant difference the different frame rates or sequence lengths, indicating potential savings in dose.

We use the classification image technique to investigate the effect of white noise and various correlated Gaussian noise textures (low-pass, high-pass, and band-pass) on observer performance in detection and discrimination tasks. For these tasks, performance is generally enhanced by an observer's ability to "prewhiten" correlated noise as part of the formation of a decision variable. We find that observer efficiency in these tasks is well represented by the measured classification images and that human observers show strong evidence of adaptation to different correlated noise textures. This adaptation is captured in the frequency weighting of the classification images.

Targets of a visual search are often not randomly positioned within a scene, but may be more likely to co-occur adjacent to other objects or background properties. Studies on target-cue co-occurrence (e.g. cue validity) suggest that observers can exploit this knowledge to increase performance in detection and localization tasks. However, little is known regarding how observers learn this co-occurrence. The present experiment sought to determine if observers were capable of learning the probability of cue validity, and determine how this learning is shaped by feedback. Separate groups of subjects performed a search task using one of three different feedback conditions providing varying degrees of information: unsupervised feedback, response reinforcement, or supervised feedback. Results show that saccadic and perceptual decisions reflect larger cueing effects as feedback information increased. This suggests that internal signals generated from response selection are insufficient for exploiting cue validity, but that reinforcement may be sufficient. However, final explicit estimates of cue validity were independent of feedback condition, suggesting that implicit behaviors are subject to unique learning constraints. Comparison to an ideal observer reveals that the rate at which participants learned cue validity was suboptimal, which may have impaired performance during initial familiarization with scene statistics.

The inclusion of internal noise in model observers is a common method to allow for quantitative comparisons between human and model observer performance in visual detection tasks. In this article, we studied two different strategies for inserting internal noise into Hotelling model observers. In the first strategy, internal noise was added to the output of individual channels: (a) Independent nonuniform channel noise, (b) independent uniform channel noise. In the second strategy, internal noise was added to the decision variable arising from the combination of channel responses. The standard deviation of the zero mean internal noise was either constant or proportional to: (a) the decision variable's standard deviation due to the external noise, (b) the decision variable's variance caused by the external noise, (c) the decision variable magnitude on a trial to trial basis. We tested three model observers: square window Hotelling observer (HO), channelized Hotelling observer (CHO), and Laguerre-Gauss Hotelling observer (LGHO) using a four alternative forced choice (4AFC) signal known exactly but variable task with a simulated signal embedded in real x-ray coronary angiogram backgrounds. The results showed that the internal noise method that led to the best prediction of human performance differed across the studied model observers. The CHO model best predicted human observer performance with the channel internal noise. The HO and LGHO best predicted human observer performance with the decision variable internal noise. The present results might guide researchers with the choice of methods to include internal noise into Hotelling model observers when evaluating and optimizing medical image quality.

Rationale and Objectives Our project was to investigate a complete methodology for the semiautomatic assessment of digital mammograms according to their density, an indicator known to be correlated to breast cancer risk. The BI-RADS four-grade density scale is usually employed by radiologists for reporting breast density, but it allows for a certain degree of subjective input, and an objective qualification of density has therefore often been reported hard to assess. The goal of this study was to design an objective technique for determining breast BI-RADS density. Materials and Methods The proposed semiautomatic method makes use of complementary pattern recognition techniques to describe manually selected regions of interest (ROIs) in the breast with 36 statistical features. Three different classifiers based on a linear discriminant analysis or Bayesian theories were designed and tested on a database consisting of 1408 ROIs from 88 patients, using a leave-one-ROI-out technique. Classifications in optimal feature subspaces with lower dimensionality and reduction to a two-class problem were studied as well. Results Comparison with a reference established by the classifications of three radiologists shows excellent performance of the classifiers, even though extremely dense breasts continue to remain more difficult to classify accurately. For the two best classifiers, the exact agreement percentages are 76% and above, and weighted κ values are 0.78 and 0.83. Furthermore, classification in lower dimensional spaces and two-class problems give excellent results. Conclusion The proposed semiautomatic classifiers method provides an objective and reproducible method for characterizing breast density, especially for the two-class case. It represents a simple and valuable tool that could be used in screening programs, training, education, or for optimizing image processing in diagnostic tasks.

In this study we estimated human observer templates associated with the detection of a realistic mass signal superimposed on real and simulated but realistic synthetic mammographic backgrounds. Five trained naive observers participated in two-alternative forced-choice (2-AFC) experiments in which they were asked to detect a spherical mass signal extracted from a mammographic phantom. This signal was superimposed on statistically stationary clustered lumpy backgrounds (CLB) in one instance, and on nonstationary real mammographic backgrounds in another. Human observer linear templates were estimated using a genetic algorithm. An additional 2-AFC experiment was conducted with twin noise in order to determine which local statistical properties of the real backgrounds influenced the ability of the human observers to detect the signal. Results show that the estimated linear templates are not significantly different for stationary and nonstationary backgrounds. The estimated performance of the linear template compared with the human observer is within 5% in terms of percent correct (Pc) for the 2-AFC task. Detection efficiency is significantly higher on nonstationary real backgrounds than on globally stationary synthetic CLB. Using the twin-noise experiment and a new method to relate image features to observers trial to trial decisions, we found that the local statistical properties preventing or making the detection task easier were the standard deviation and three features derived from the neighborhood gray-tone difference matrix: coarseness, contrast and strength. These statistical features showed a dependency with the human performance only when they are estimated within an area sufficiently small around the searched location. These findings emphasize that nonstationary backgrounds need to be described by their local statistics and not by global ones like the noise Wiener spectrum.

Are the body's actions and the mind's perceptions the result of shared neural processing, or are they performed largely independently? The brain has two major processing streams, and some have proposed that this division segregates visual processing for action and perception. The ventral pathway is claimed to support conscious experience (perception), whereas the dorsal pathway is claimed to support the control of movement (action). Others have argued that perception and action share much of their visual processing within the primate cortex. During visual search, the brain performs a sophisticated deployment of eye movements (saccadic actions) to gather information to subserve perceptual judgments. The relationship between the neural mechanisms mediating perception and action in visual search remains unexplored. Here, we investigate the visual representation of target information in the human brain, both for perceptual decisions and for saccadic actions during visual search. We use classification image analysis, a form of reverse correlation, to estimate the behavioral receptive fields of the visual mechanisms responsible for saccadic and perceptual responses during the same visual search task. Results show that the behavioral receptive fields mediating the perceptual decisions are indistinguishable from those driving the oculomotor decisions, suggesting that similar neural mechanisms are responsible for both perception and oculomotor action during search. Diverging target representations would result in an inefficient coupling between eye movement planning and perceptual judgments. Thus, a common target representation would be more optimal and might be expected to have evolved through natural selection in the neural systems responsible for visual search.

X-ray medical imaging is increasingly becoming three-dimensional (3-D). The dose to the population and its management are of special concern in computed tomography (CT). Task-based methods with model observers to assess the dose-image quality trade-off are promising tools, but they still need to be validated for real volumetric images. The purpose of the present work is to evaluate anthropomorphic model observers in 3-D detection tasks for low-contrast CT images. We scanned a low-contrast phantom containing four types of signals at three dose levels and used two reconstruction algorithms. We implemented a multislice model observer based on the channelized Hotelling observer (msCHO) with anthropomorphic channels and investigated different internal noise methods. We found a good correlation for all tested model observers. These results suggest that the msCHO can be used as a relevant task-based method to evaluate low-contrast detection for CT and optimize scan protocols to lower dose in an efficient way.

Culture influences not only human high-level cognitive processes but also low-level perceptual operations. Some perceptual operations, such as initial eye movements to faces, are critical for extraction of information supporting evolutionarily important tasks such as face identification. The extent of cultural effects on these crucial perceptual processes is unknown. Here, we report that the first gaze location for face identification was similar across East Asian and Western Caucasian cultural groups: Both fixated a featureless point between the eyes and the nose, with smaller between-group than within-group differences and with a small horizontal difference across cultures (8% of the interocular distance). We also show that individuals of both cultural groups initially fixated at a slightly higher point on Asian faces than on Caucasian faces. The initial fixations were found to be both fundamental in acquiring the majority of information for face identification and optimal, as accuracy deteriorated when observers held their gaze away from their preferred fixations. An ideal observer that integrated facial information with the human visual system's varying spatial resolution across the visual field showed a similar information distribution across faces of both races and predicted initial human fixations. The model consistently replicated the small vertical difference between human fixations to Asian and Caucasian faces but did not predict the small horizontal leftward bias of Caucasian observers. Together, the results suggest that initial eye movements during face identification may be driven by brain mechanisms aimed at maximizing accuracy, and less influenced by culture. The findings increase our understanding of the interplay between the brain's aims to optimally accomplish basic perceptual functions and to respond to sociocultural influences.