Errors in visual search: Are they stochastic or deterministic?

In any visual search task in the lab or in the world, observers will make errors. Those errors can be categorized as "deterministic": If you miss this target in this display once, you will definitely miss it again. Alternatively, errors can be "stochastic", occurring randomly with some probability from trial to trial. Researchers and practitioners have sought to reduce errors in visual search, but different types of errors might require different techniques for mitigation. To empirically categorize errors in a simple search task, our observers searched for the letter "T" among "L" distractors, with each display presented twice. When the letters were clearly visible (white letters on a gray background), the errors were almost completely stochastic (Exp 1). An error made on the first appearance of a display did not predict that an error would be made on the second appearance. When the visibility of the letters was manipulated (letters of different gray levels on a noisy background), the errors became a mix of stochastic and deterministic. Unsurprisingly, lower contrast targets produced more deterministic errors. (Exp 2). Using the stimuli of Exp 2, we tested whether errors could be reduced using cues that guided attention around the display but knew nothing about the content of that display (Exp3a, b). This had no effect, but cueing all item locations did succeed in reducing deterministic errors (Exp3c).

On the usefulness of graph-theoretic properties in the study of perceived numerosity.

Observers can quickly estimate the quantity of sets of visual elements. Many aspects of this ability have been studied and the underlying system has been called the Approximate Number Sense (Dehaene, 2011). Specific visual properties, such as size and clustering of the elements, can bias an estimate. For intermediate numerical quantities at low density (above five, but before texturization), human performance is predicted by a model based on the region of influence of elements (occupancy model: Allïk & Tuulmets, 1991). For random 2D configurations we computed ten indices based on graph theory, and we compared them with the occupancy model: independence number, domination, connected components, local clustering coefficient, global clustering coefficient, random walk, eigenvector centrality, maximum clique, total degree of connectivity, and total edge length. We made comparisons across a range of parameters, and we varied the size of the region of influence around each element. The analysis of the pattern of correlations suggests two main groups of graph-based measures. The first group is sensitive to the presence of local clustering of elements, the second seems more sensitive to density and the way information spreads in graphs. Empirical work on perception of numerosity may benefit from comparing, or controlling for, these properties.

Cognitive and Electrophysiological Correlates of Working Memory Impairments in Neurofibromatosis Type 1.

Neurofibromatosis 1 (NF1) is a single gene disorder associated with working Memory (WM) impairments. The aim of this study was to investigate P300 event-related potential (ERP) associated with WM in NF1. Sixteen adolescents with NF1 were compared with controls on measures of WM and EEG was recorded during a WM nback task. The NF1 group showed poorer performance on measures of WM as compared to the control group. No group differences were observed in P300 amplitude at Pz, but P300 latency was shorter in the NF1 group. Topographic analyses of P300 amplitude showed group differences indicating neural processing differences in the NF1 group relative to controls, which possibly contribute to the cognitive deficits seen in this population.

Quantitative and qualitative differences in the top-down guiding attributes of visual search.

It is generally assumed that there is a category of undoubted guiding attributes in visual search: color, motion, orientation. Any differences between these attributes are a matter of degree, rather than kind. This assumption has led to a preferential use of color in experiments that involve top-down guidance, because it provides the strongest effects. Yet, results observed for color are considered representative for the other undoubted attributes. This article reports 7 experiments that compare the top-down guiding strengths of color, motion, and orientation by adding them to a T versus L search. Giving some of the Ls a different color, motion status, or orientation should make the T easier to find, because those Ls cannot possibly be the target. The results show that whereas adding color or motion does indeed improve search performance, adding differently oriented Ls actually makes search harder, especially on absent trials and even when there are only very few items that could be the target. There were also some subtle differences between color and motion. So, rather than a single category of undoubted guiding attributes, there seems to be a clear ranking, with color at the top and orientation at the bottom. It may therefore be unwise to think that results found with color will translate one-on-one to other attributes like motion and orientation. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Is there a safety-net effect with computer-aided detection?

Computer-aided detection (CAD) systems are used to aid readers interpreting screening mammograms. An expert reader searches the image initially unaided and then once again with the aid of CAD, which prompts automatically detected suspicious regions. This could lead to a "safety-net" effect, where the initial unaided search of the image is adversely affected by the fact that it is preliminary to an additional search with CAD and may, therefore, be less thorough. To investigate the existence of such an effect, we created a visual search experiment for nonexpert observers mirroring breast screening with CAD. Each observer searched 100 images for microcalcification clusters within synthetic images in both prompted (CAD) and unprompted (no-CAD) conditions. Fifty-two participants were recruited for the study, 48 of whom had their eye movements tracked in real-time; the other 4 participants could not be accurately calibrated, so only behavioral data were collected. In the CAD condition, before prompts were displayed, image coverage was significantly lower than coverage in the no-CAD condition ( t 47 = 5.29 , p < 0.0001 ). Observer sensitivity was significantly greater for targets marked by CAD than the same targets in the no-CAD condition ( t 51 = 6.56 , p < 0.001 ). For targets not marked by CAD, there was no significant difference in observer sensitivity in the CAD condition compared with the same targets in the no-CAD condition ( t 51 = 0.54 , p = 0.59 ). These results suggest that the initial search may be influenced by the subsequent availability of CAD; if so, cross-sectional CAD efficacy studies should account for the effect when estimating benefit.

Medium versus difficult visual search: How a quantitative change in the functional visual field leads to a qualitative difference in performance.

The dominant theories of visual search assume that search is a process involving comparisons of individual items against a target description that is based on the properties of the target in isolation. Here, we present four experiments that demonstrate that this holds true only in difficult search. In medium search it seems that the relation between the target and neighbouring items is also part of the target description. We used two sets of oriented lines to construct the search items. The cardinal set contained horizontal and vertical lines, the diagonal set contained left diagonal and right diagonal lines. In all experiments, participants knew the identity of the target and the line set used to construct it. In difficult search this knowledge allowed performance to improve in displays where only half of the search items came from the same line set as the target (50% eligibility), relative to displays where all items did (100% eligibility). However, in medium search, performance was actually poorer for 50% eligibility, especially on target-absent trials. This opposite effect of ineligible items in medium search and difficult search is hard to reconcile with theories based on individual items. It is more in line with theories that conceive search as a sequence of fixations where the number of items processed during a fixation depends on the difficulty of the search task: When search is medium, multiple items are processed per fixation. But when search is difficult, only a single item is processed.

Prediction of reader estimates of mammographic density using convolutional neural networks.

Mammographic density is an important risk factor for breast cancer. In recent research, percentage density assessed visually using visual analogue scales (VAS) showed stronger risk prediction than existing automated density measures, suggesting readers may recognize relevant image features not yet captured by hand-crafted algorithms. With deep learning, it may be possible to encapsulate this knowledge in an automatic method. We have built convolutional neural networks (CNN) to predict density VAS scores from full-field digital mammograms. The CNNs are trained using whole-image mammograms, each labeled with the average VAS score of two independent readers. Each CNN learns a mapping between mammographic appearance and VAS score so that at test time, they can predict VAS score for an unseen image. Networks were trained using 67,520 mammographic images from 16,968 women and for model selection we used a dataset of 73,128 images. Two case-control sets of contralateral mammograms of screen detected cancers and prior images of women with cancers detected subsequently, matched to controls on age, menopausal status, parity, HRT and BMI, were used for evaluating performance on breast cancer prediction. In the case-control sets, odd ratios of cancer in the highest versus lowest quintile of percentage density were 2.49 (95% CI: 1.59 to 3.96) for screen-detected cancers and 4.16 (2.53 to 6.82) for priors, with matched concordance indices of 0.587 (0.542 to 0.627) and 0.616 (0.578 to 0.655), respectively. There was no significant difference between reader VAS and predicted VAS for the prior test set (likelihood ratio chi square, p = 0.134 ). Our fully automated method shows promising results for cancer risk prediction and is comparable with human performance.

Seeing the World as it is: Mimicking Veridical Motion Perception in Schizophrenia Using Non-invasive Brain Stimulation in Healthy Participants.

Schizophrenia (Sz) is a mental health disorder characterized by severe cognitive, emotional, social, and perceptual deficits. Visual deficits are found in tasks relying on the magnocellular/dorsal stream. In our first experiment we established deficits in global motion processing in Sz patients compared to healthy controls. We used a novel task in which background optic flow produces a distortion of the apparent trajectory of a moving stimulus, leading control participants to provide biased estimates of the true motion trajectory under conditions of global stimulation. Sz patients were significantly less affected by the global background motion, and reported trajectories that were more veridically accurate than those of controls. In order to study the mechanism of this effect, we performed a second experiment where we applied transcranial electrical stimulation over area MT+ to selectively modify global motion processing of optic flow displays in healthy participants. Cathodal and high frequency random noise stimulation had opposite effects on trajectory perception in optic flow. The brain stimulation over a control site and in a control task revealed that the effect of stimulation was specific for global motion processing in area MT+. These findings both support prior studies of impaired early visual processing in Sz and provide novel approaches for measurement and manipulation of the underlying circuits.

On the brink: The demise of the item in visual search moves closer.

We proposed to abandon the item as conceptual unit in visual search and adopt a fixation-based framework instead. We treat various themes raised by our commentators, including the nature of the Functional Visual Field and existing similar ideas, alongside the importance of items, covert attention, and top-down/contextual influences. We reflect on the current state of, and future directions for, visual search.

The impending demise of the item in visual search.

The way the cognitive system scans the visual environment for relevant information - visual search in short - has been a long-standing central topic in vision science. From its inception as a research topic, and despite a number of promising alternative perspectives, the study of visual search has been governed by the assumption that a search proceeds on the basis of individual items (whether processed in parallel or not). This has led to the additional assumptions that shallow search slopes (at most a few tens of milliseconds per item for target-present trials) are most informative about the underlying process, and that eye movements are an epiphenomenon that can be safely ignored. We argue that the evidence now overwhelmingly favours an approach that takes fixations, not individual items, as its central unit. Within fixations, items are processed in parallel, and the functional field of view determines how many fixations are needed. In this type of theoretical framework, there is a direct connection between target discrimination difficulty, fixations, and reaction time (RT) measures. It therefore promises a more fundamental understanding of visual search by offering a unified account of both eye movement and manual response behaviour across the entire range of observed search efficiency, and provides new directions for research. A high-level conceptual simulation with just one free and four fixed parameters shows the viability of this approach.

Spatial clustering and its effect on perceived clustering, numerosity, and dispersion.

Human observers are able to estimate the numerosity of large sets of visual elements. The occupancy model of perceived numerosity in intermediate numerical ranges is based on overlapping regions of influence. The key idea is that items within a certain range count for less than their actual numerical value and more so the closer they are to their neighbours. Therefore occupancy is sensitive to the grouping of elements, but there are other spatial properties of  configurations that could also influence perceived numerosity, such as: area of convex hull, occupancy area, total degree of connectivity, and local clustering For all indices apart from convex hull, we varied the radius of the area that defined neighbours. We tested perceived numerosity using a fixed number of elements placed at random within a circular region. Observers compared two patterns (presented in two intervals) and chose the one that appeared more numerous. The same observers performed two other separate tasks in which they judged which pattern appeared more dispersed or more clustered. In each pair of images, the number was always the same (22, 28, 34, or 40 items), because we were interested in which "appeared" more numerous on the basis of spatial configuration. The results suggest that estimates of numerosity, dispersion, and clustering are based on different spatial information, that there are alternative approaches to quantifying clustering, and that in all cases clustering is linked to a decrease in perceived numerosity. The alternative measures have different properties and different practical and computational advantages.

Targeted memory reactivation during slow wave sleep facilitates emotional memory consolidation.

STUDY OBJECTIVES: To investigate the mechanisms by which auditory targeted memory reactivation (TMR) during slow wave sleep (SWS) influences the consolidation of emotionally negative and neutral memories. DESIGN: Each of 72 (36 negative, 36 neutral) picture-location associations were encoded with a semantically related sound. During a subsequent nap, half of the sounds were replayed in SWS, before picture-location recall was examined in a final test. SETTING: Manchester Sleep Laboratory, University of Manchester. PARTICIPANTS: 15 adults (3 male) mean age = 20.40 (standard deviation ± 3.07). INTERVENTIONS: TMR with auditory cues during SWS. MEASUREMENTS AND RESULTS: Performance was assessed by memory accuracy and recall response times (RTs). Data were analyzed with a 2 (sound: replayed/not replayed) × 2 (emotion: negative/neutral) repeated measures analysis of covariance with SWS duration, and then SWS spindles, as the mean-centered covariate. Both analyses revealed a significant three-way interaction for RTs but not memory accuracy. Critically, SWS duration and SWS spindles predicted faster memory judgments for negative, relative to neutral, picture locations that were cued with TMR. CONCLUSIONS: TMR initiates an enhanced consolidation process during subsequent SWS, wherein sleep spindles mediate the selective enhancement of reactivated emotional memories.

Looming motion primes the visuomotor system.

A wealth of evidence now shows that human and animal observers display greater sensitivity to objects that move toward them than to objects that remain static or move away. Increased sensitivity in humans is often evidenced by reaction times that increase in rank order from looming, to receding, to static targets. However, it is not clear whether the processing advantage enjoyed by looming motion is mediated by the attention system or the motor system. The present study investigated this by first examining whether sensitivity is to looming motion per se or to certain monocular or binocular cues that constitute stereoscopic motion in depth. None of the cues accounted for the looming advantage. A perceptual measure was then used to examine performance with minimal involvement of the motor system. Results showed that looming and receding motion were equivalent in attracting attention, suggesting that the looming advantage is indeed mediated by the motor system. These findings suggest that although motion itself is sufficient for attentional capture, motion direction can prime motor responses.

Search through complex motion displays does not break down under spatial memory load.

Recent work has shown that observers are remarkably effective in searching displays of randomly moving items. In two experiments, we combined working memory tasks with visual search, to test whether search through such complex motion displays, as compared with search through static items, places an extra burden on spatial working memory. In our first experiment, we show that the dual-task interference observed for motion search is specific to spatial working memory, in line with earlier work for static search. In our second experiment, we found dual-task interference for both static and motion search, but no difference between them. The results support the suggestion that the same search process is active during search among static and search among moving items.

The role of convexity in perception of symmetry and in visual short-term memory.

Visual perception of shape is affected by coding of local convexities and concavities. For instance, a recent study reported that deviations from symmetry carried by convexities were easier to detect than deviations carried by concavities. We removed some confounds and extended this work from a detection of reflection of a contour (i.e., bilateral symmetry), to a detection of repetition of a contour (i.e., translational symmetry). We tested whether any convexity advantage is specific to bilateral symmetry in a two-interval (Experiment 1) and a single-interval (Experiment 2) detection task. In both, we found a convexity advantage only for repetition. When we removed the need to choose which region of the contour to monitor (Experiment 3) the effect disappeared. In a second series of studies, we again used shapes with multiple convex or concave features. Participants performed a change detection task in which only one of the features could change. We did not find any evidence that convexities are special in visual short-term memory, when the to-be-remembered features only changed shape (Experiment 4), when they changed shape and changed from concave to convex and vice versa (Experiment 5), or when these conditions were mixed (Experiment 6). We did find a small advantage for coding convexity as well as concavity over an isolated (and thus ambiguous) contour. The latter is consistent with the known effect of closure on processing of shape. We conclude that convexity plays a role in many perceptual tasks but that it does not have a basic encoding advantage over concavity.

Eye movements reveal how task difficulty moulds visual search.

In two experiments we investigated the relationship between eye movements and performance in visual search tasks of varying difficulty. Experiment 1 provided evidence that a single process is used for search among static and moving items. Moreover, we estimated the functional visual field (FVF) from the gaze coordinates and found that its size during visual search shrinks with increasing task difficulty. In Experiment 2, we used a gaze-contingent window and confirmed the validity of the size estimates. The experiment also revealed that breakdown in robustness against item motion is related to item-by-item search, rather than search difficulty per se. We argue that visual search is an eye-movement-based process that works on a continuum, from almost parallel (where many items can be processed within a fixation) to completely serial (where only one item can be processed within a fixation).

Proneness to decreased negative emotions in major depressive disorder when blaming others rather than oneself.

BACKGROUND: One widespread view holds that vulnerability to major depressive disorder (MDD) is linked to overall increases in negative emotionality. In contrast, cognitive attribution theories emphasize the importance of blaming oneself rather than others for negative events. Thus far, the contrasting predictions of these models have not been directly compared. Following the attributional perspective, we tested the hypothesis that people with remitted MDD show no overall bias towards negative emotions, but a selective bias towards self-blaming emotions relative to those emotions associated with blaming others. SAMPLING AND METHODS: We compared a remitted MDD and a control group on a novel experimental test that allowed us to directly compare proneness to specific emotions associated with different types of self-blame (guilt, shame, self-contempt/disgust) and blame of others (other-indignation/anger, other-contempt/disgust) whilst controlling for negative valence and medication status, and excluding comorbidity. RESULTS: In agreement with our hypothesis, individuals with remitted MDD exhibited an increased self-contempt bias (difference between contempt/disgust towards self and others) but no increased proneness to any other negative emotion or overall increases in perceived negative valence of stimuli. Moreover, the remitted MDD group exhibited reduced contempt/disgust towards others. CONCLUSIONS: Our results corroborate the prediction that vulnerability to MDD is associated with an imbalance of specific self- and other-blaming emotions rather than a general increase in negative emotions. Based on the composition of our sample, we speculate that self-contempt bias may be particularly characteristic of melancholic MDD subtypes and could be useful for stratification of depression in the future.

Modulating wheelchair navigation in patients with spatial neglect.

Patients who have had a stroke resulting in the deficit of visuo-spatial neglect are normally not provided with a powered wheelchair, as they are either considered or found to be unsafe navigating about their environment. As these patients are relatively unlikely to regain functional mobility by walking, the denial of alternative forms of mobility is of particular concern. Modest progress has been made over the past two decades with regards to the rehabilitation of neglect but there have been calls for further research which addresses "real world" measures of independence such as wheelchair navigation. In this study, we investigated the ability of patients with neglect to improve their performance when navigating a powered wheelchair by using theoretically-driven strategies that have shown promise in previous studies (spatial cueing and limb activation). Strategies were applied and tested in the most realistic and practical manner for each individual, based on their abilities and concurrent deficits. Performance was improved by the experimental strategies. The data suggest it is possible to apply theoretically-driven strategies to improve wheelchair navigation in patients with neglect and are supportive of further studies that could lead to improved access to powered mobility by this population in the future.

How does our search engine "see" the world? The case of amodal completion.

This article illustrates a dissociation between the perceived attributes of an object and the ability of those attributes to guide the deployment of attention in visual search. Orientation is an attribute that guides search. Thus, a vertical line will "pop out" amid horizontal distractors. Amodal completion can create perceptually convincing oriented stimuli when two elements appear to form a complete object partially hidden behind an occluder. Previous work (e.g., Rensink & Enns, Vision Research, 38, 2489-2505, 1998) has shown a preattentive role for amodal completion in search tasks. Here, we show that orientation based on perceptually compelling amodal completion may fail to guide attention. The broader conclusion is that introspection is a poor guide to the capabilities of our internal search engine.

The sparing is far from spurious: reevaluating within-trial contingency effects in the attentional blink.

The attentional blink is the marked deficit in awareness of a 2nd target (T2) when it is presented shortly after the 1st target (T1) in a stream of distractors. When the distractors between T1 and T2 are replaced by even more targets, the attentional blink is reduced or absent, indicating that the attentional blink results from online selection mechanisms that act in response to distracting input rather than being the result of T1-induced cognitive resource depletion. However, Dell'Acqua, Jolicoeur, Luria, and Pluchino (2009) recently contended that an attentional blink is found in the multiple-target case as long as the appropriate trial context and analyses are used, thus reinstating resource-based explanations of the attentional blink and challenging the selection account. Specifically, an attentional blink reemerges when target performance is analyzed contingent on previous target accuracy. We argue on theoretical and empirical grounds that neither the trial context nor the type of analysis poses a serious problem for selection accounts. We show that the attentional blink and previous target contingency effects can be dissociated, with the latter depending more on low-level, short-range competition. We conclude that selection mechanisms involved in filtering for targets still provide a strong and coherent explanation of the attentional blink.