Search

Purpose: Digital whole slide imaging allows pathologists to view slides on a computer screen instead of under a microscope. Digital viewing allows for real-time monitoring of pathologists search behavior and neurophysiological responses during the diagnostic process. One particular neurophysiological measure, pupil diameter, could provide a basis for evaluating clinical competence during training or developing tools that support the diagnostic process. Prior research shows that pupil diameter is sensitive to cognitive load and arousal, and it switches between exploration and exploitation of a visual image. Different categories of lesions in pathology pose different levels of challenge, as indicated by diagnostic disagreement among pathologists. If pupil diameter is sensitive to the perceived difficulty in diagnosing biopsies, eye-tracking could potentially be used to identify biopsies that may benefit from a second opinion. Approach: We measured case onset baseline-corrected (phasic) and uncorrected (tonic) pupil diameter in 90 pathologists who each viewed and diagnosed 14 digital breast biopsy cases that cover the diagnostic spectrum from benign to invasive breast cancer. Pupil data were extracted from the beginning of viewing and interpreting of each individual case. After removing 122 trials ( < 10 % ) with poor eye-tracking quality, 1138 trials remained. We used multiple linear regression with robust standard error estimates to account for dependent observations within pathologists. Results: We found a positive association between the magnitude of phasic dilation and subject-centered difficulty ratings and between the magnitude of tonic dilation and untransformed difficulty ratings. When controlling for case diagnostic category, only the tonic-difficulty relationship persisted. Conclusions: Results suggest that tonic pupil dilation may indicate overall arousal differences between pathologists as they interpret biopsy cases and could signal a need for additional training, experience, or automated decision aids. Phasic dilation is sensitive to characteristics of biopsies that tend to elicit higher difficulty ratings and could indicate a need for a second opinion.

Abstract Extraction of global structural regularities provides general ‘gist’ of our everyday visual environment as it does the gist of abnormality for medical experts reviewing medical images. We investigated whether naïve observers could learn this gist of medical abnormality. Fifteen participants completed nine adaptive training sessions viewing four categories of unilateral mammograms: normal, obvious-abnormal, subtle-abnormal, and global signals of abnormality (mammograms with no visible lesions but from breasts contralateral to or years prior to the development of cancer) and receiving only categorical feedback. Performance was tested pre-training, post-training, and after a week’s retention on 200 mammograms viewed for 500 ms without feedback. Performance measured as d’ was modulated by mammogram category, with the highest performance for mammograms with visible lesions. Post-training, twelve observed showed increased d’ for all mammogram categories but a subset of nine, labelled learners also showed a positive correlation of d’ across training. Critically, learners learned to detect abnormality in mammograms with only the global signals, but improvements were poorly retained. A state-of-the-art breast cancer classifier detected mammograms with lesions but struggled to detect cancer in mammograms with the global signal of abnormality. The gist of abnormality can be learned through perceptual/incidental learning in mammograms both with and without visible lesions, subject to individual differences. Poor retention suggests perceptual tuning to gist needs maintenance, converging with findings that radiologists’ gist performance correlates with the number of cases reviewed per year, not years of experience. The human visual system can tune itself to complex global signals not easily captured by current deep neural networks.

Abstract Despite numerous investigations of the prevalence effect on medical image perception, little research has been done to examine the effect of expertise, and its possible interaction with prevalence. In this study, medical practitioners were instructed to detect the presence of hip fracture in 50 X-ray images with either high prevalence (N signal = 40) or low prevalence (N signal = 10). Results showed that compared to novices (e.g., pediatricians, dentists, neurologists), the manipulation of prevalence shifted participant’s criteria in a different direction for experts who perform hip fracture diagnosis on a daily basis. That is, when prevalence rate is low (p fracture-present = 0.2), experts held more conservative criteria in answering “fracture-present,” whereas novices were more likely to believe there was fracture. Importantly, participants’ detection discriminability did not vary by the prevalence condition. In addition, all participants were more conservative with “fracture-present” responses when task difficulty increased. We suspect the apparent opposite criteria shift between experts and novices may have come from medical training that made novices to believe that a miss would result in larger cost compared to false positive, or because they failed to update their prior belief about the signal prevalence in the task, both would suggest that novices and experts may have different beliefs in placing the optimal strategy in the hip fracture diagnosis. Our work can contribute to medical education training as well as other applied clinical diagnosis that aims to mitigate the prevalence effect.

Serial Dependence is a ubiquitous visual phenomenon in which sequentially viewed images appear more similar than they actually are, thus facilitating an efficient and stable perceptual experience in human observers. Although serial dependence is adaptive and beneficial in the naturally autocorrelated visual world, a smoothing perceptual experience, it might turn maladaptive in artificial circumstances, such as medical image perception tasks, where visual stimuli are randomly sequenced. Here, we analyzed 758,139 skin cancer diagnostic records from an online app, and we quantified the semantic similarity between sequential dermatology images using a computer vision model as well as human raters. We then tested whether serial dependence in perception occurs in dermatological judgments as a function of image similarity. We found significant serial dependence in perceptual discrimination judgments of lesion malignancy. Moreover, the serial dependence was tuned to the similarity in the images, and it decayed over time. The results indicate that relatively realistic store-and-forward dermatology judgments may be biased by serial dependence. These findings help in understanding one potential source of systematic bias and errors in medical image perception tasks and hint at useful approaches that could alleviate the errors due to serial dependence. [ABSTRACT FROM AUTHOR]

Many applied screening tasks (e.g., medical image or baggage screening) involve challenging searches for which standard laboratory search is rarely equivalent. For example, whereas laboratory search frequently requires observers to look for precisely defined targets among isolated, non-overlapping images randomly arrayed on clean backgrounds, medical images present unspecified targets in noisy, yet spatially regular scenes. Those unspecified targets are typically oddities, elements that do not belong. To develop a closer laboratory analogue to this, we created a database of scenes containing subtle, ill-specified "oddity" targets. These scenes have similar perceptual densities and spatial regularities to those found in expert search tasks, and each includes 16 variants of the unedited scene wherein an oddity (a subtle deformation of the scene) is hidden. In Experiment 1, eight volunteers searched thousands of scene variants for an oddity. Regardless of their search accuracy, they were then shown the highlighted anomaly and rated its subtlety. Subtlety ratings reliably predicted search performance (accuracy and response times) and did so better than image statistics. In Experiment 2, we conducted a conceptual replication in which a larger group of naïve searchers scanned subsets of the scene variants. Prior subtlety ratings reliably predicted search outcomes. Whereas medical image targets are difficult for naïve searchers to detect, our database contains thousands of interior and exterior scenes that vary in difficulty, but are nevertheless searchable by novices. In this way, the stimuli will be useful for studying visual search as it typically occurs in expert domains: Ill-specified search for anomalies in noisy displays. [ABSTRACT FROM AUTHOR]

Extraction of global structural regularities provides general 'gist' of our everyday visual environment as it does the gist of abnormality for medical experts reviewing medical images. We investigated whether naïve observers could learn this gist of medical abnormality. Fifteen participants completed nine adaptive training sessions viewing four categories of unilateral mammograms: normal, obvious-abnormal, subtle-abnormal, and global signals of abnormality (mammograms with no visible lesions but from breasts contralateral to or years prior to the development of cancer) and receiving only categorical feedback. Performance was tested pre-training, post-training, and after a week's retention on 200 mammograms viewed for 500 ms without feedback. Performance measured as d' was modulated by mammogram category, with the highest performance for mammograms with visible lesions. Post-training, twelve observed showed increased d' for all mammogram categories but a subset of nine, labelled learners also showed a positive correlation of d' across training. Critically, learners learned to detect abnormality in mammograms with only the global signals, but improvements were poorly retained. A state-of-the-art breast cancer classifier detected mammograms with lesions but struggled to detect cancer in mammograms with the global signal of abnormality. The gist of abnormality can be learned through perceptual/incidental learning in mammograms both with and without visible lesions, subject to individual differences. Poor retention suggests perceptual tuning to gist needs maintenance, converging with findings that radiologists' gist performance correlates with the number of cases reviewed per year, not years of experience. The human visual system can tune itself to complex global signals not easily captured by current deep neural networks. [ABSTRACT FROM AUTHOR]

Despite numerous investigations of the prevalence effect on medical image perception, little research has been done to examine the effect of expertise, and its possible interaction with prevalence. In this study, medical practitioners were instructed to detect the presence of hip fracture in 50 X-ray images with either high prevalence (Nsignal = 40) or low prevalence (Nsignal = 10). Results showed that compared to novices (e.g., pediatricians, dentists, neurologists), the manipulation of prevalence shifted participant's criteria in a different direction for experts who perform hip fracture diagnosis on a daily basis. That is, when prevalence rate is low (pfracture-present = 0.2), experts held more conservative criteria in answering "fracture-present," whereas novices were more likely to believe there was fracture. Importantly, participants' detection discriminability did not vary by the prevalence condition. In addition, all participants were more conservative with "fracture-present" responses when task difficulty increased. We suspect the apparent opposite criteria shift between experts and novices may have come from medical training that made novices to believe that a miss would result in larger cost compared to false positive, or because they failed to update their prior belief about the signal prevalence in the task, both would suggest that novices and experts may have different beliefs in placing the optimal strategy in the hip fracture diagnosis. Our work can contribute to medical education training as well as other applied clinical diagnosis that aims to mitigate the prevalence effect. [ABSTRACT FROM AUTHOR]

Introduction: Radiologists routinely make life-altering decisions. Optimizing these decisions has been an important goal for many years and has prompted a great deal of research on the basic perceptual mechanisms that underlie radiologists’ decisions. Previous studies have found that there are substantial individual differences in radiologists’ diagnostic performance (e.g., sensitivity) due to experience, training, or search strategies. In addition to variations in sensitivity, however, another possibility is that radiologists might have perceptual biases—systematic misperceptions of visual stimuli. Although a great deal of research has investigated radiologist sensitivity, very little has explored the presence of perceptual biases or the individual differences in these. Methods: Here, we test whether radiologists’ have perceptual biases using controlled artificial and Generative Adversarial Networks-generated realistic medical images. In Experiment 1, observers adjusted the appearance of simulated tumors to match the previously shown targets. In Experiment 2, observers were shown with a mix of real and GAN-generated CT lesion images and they rated the realness of each image. Results: We show that every tested individual radiologist was characterized by unique and systematic perceptual biases; these perceptual biases cannot be simply explained by attentional differences, and they can be observed in different imaging modalities and task settings, suggesting that idiosyncratic biases in medical image perception may widely exist. Discussion: Characterizing and understanding these biases could be important for many practical settings such as training, pairing readers, and career selection for radiologists. These results may have consequential implications for many other fields as well, where individual observers are the linchpins for life-altering perceptual decisions. [ABSTRACT FROM AUTHOR]

Radiologists can discriminate between normal and abnormalbreast tissue at a glance, suggesting that radiologists might beusing some “global signal” of abnormality. Our study inves-tigated whether texture descriptions can be used to character-ize the global signal of abnormality and whether radiologistsuse this information during interpretation. Synthetic imageswere generated using a texture synthesis algorithm trained ontexture descriptions extracted from sections of mammograms.Radiologists completed a task that required rating the abnor-mality of briefly presented tissue sections. When the abnormaltissue had no visible lesion, radiologists seemed to use texturedescriptions; performance was similar across real and synthe-sized tissue sections. However, when the abnormal tissue had avisible lesion, radiologists seemed to rely on additional mech-anisms beyond the texture descriptions; performance increasedfor the real tissue sections. These findings suggest that radiol-ogists can use texture descriptions as global signals of abnor-mality in interpretation of breast tissue.

Objective: This research was designed to test whether behavioral indicators of pathology-related cue utilization were associated with performance on a diagnostic task. Background: Across many domains, including pathology, successful diagnosis depends on pattern recognition that is supported by associations in memory in the form of cues. Previous studies have focused on the specific information or knowledge on which medical image expertise relies. The target in this study is the more general ability to identify and interpret relevant information. Method: Data were collected from 54 histopathologists in both conference and online settings. The participants completed a pathology edition of the Expert Intensive Skills Evaluation 2.0 (EXPERTise 2.0) to establish behavioral indicators of context-related cue utilization. They also completed a separate diagnostic task designed to examine related diagnostic skills. Results: Behavioral indicators of higher or lower cue utilization were based on the participants' performance across five tasks. Accounting for the number of cases reported per year, higher cue utilization was associated with greater accuracy on the diagnostic task. A post hoc analysis suggested that higher cue utilization may be associated with a greater capacity to recognize low prevalence cases. Conclusion: This study provides support for the role of cue utilization in the development and maintenance of skilled diagnosis amongst pathologists. Application: Pathologist training needs to be structured to ensure that learners have the opportunity to form cue-based strategies and associations in memory, especially for less commonly seen diseases. [ABSTRACT FROM AUTHOR]

Abstract Prior research has shown that interruptions lead to a variety of performance costs. However, these costs are heterogenous and poorly understood. Under some circumstances, interruptions lead to large decreases in accuracy on the primary task, whereas in others task duration increases, but task accuracy is unaffected. Presently, the underlying cause of these costs is unclear. The Memory for Goals model suggests that interruptions interfere with the ability to represent the current goal of the primary task. Here, we test the idea that working memory (WM) may play a critical role in representing the current goal and thus may underlie the observed costs associated with interruption. In two experiments, we utilized laboratory-based visual search tasks, which differed in their WM demands, in order to assess how this difference influenced the observed interruption costs. Interruptions led to more severe performance costs when the target of the search changed on each trial. When the search target was consistent across trials, the cost of interruption was greatly reduced. This suggests that the WM demands associated with the primary task play an important role in determining the performance costs of interruption. Our findings suggest that it is important for research to consider the cognitive processes a task engages in order to predict the nature of the adverse effects of interruption in applied settings such as radiology.

Rationale and objectivesRadiology expertise is dependent on the use of efficient search strategies. The aim of this study is to investigate the effect of teaching search strategies on trainee's accuracy in detecting lung nodules at computed tomography.Materials and methodsTwo search strategies, "scanning" and "drilling," were tested with a randomized crossover design. Nineteen junior radiology residents were randomized into two groups. Both groups first completed a baseline lung nodule detection test allowing a free search strategy, followed by a test after scanning instruction and drilling instruction or vice versa. True positive (TP) and false positive (FP) scores and scroll behavior were registered. A mixed-design analysis of variance was applied to compare the three search conditions.ResultsSearch strategy instruction had a significant effect on scroll behavior, F(1.3) = 54.2, P

Serial Dependence is a ubiquitous visual phenomenon in which sequentially viewed images appear more similar than they actually are, thus facilitating an efficient and stable perceptual experience in human observers. Although serial dependence is adaptive and beneficial in the naturally autocorrelated visual world, a smoothing perceptual experience, it might turn maladaptive in artificial circumstances, such as medical image perception tasks, where visual stimuli are randomly sequenced. Here, we analyzed 758,139 skin cancer diagnostic records from an online app, and we quantified the semantic similarity between sequential dermatology images using a computer vision model as well as human raters. We then tested whether serial dependence in perception occurs in dermatological judgments as a function of image similarity. We found significant serial dependence in perceptual discrimination judgments of lesion malignancy. Moreover, the serial dependence was tuned to the similarity in the images, and it decayed over time. The results indicate that relatively realistic store-and-forward dermatology judgments may be biased by serial dependence. These findings help in understanding one potential source of systematic bias and errors in medical image perception tasks and hint at useful approaches that could alleviate the errors due to serial dependence.

Radiologists face the visually challenging task of detecting suspicious features within the complex and noisy backgrounds characteristic of medical images. We used a search task to examine whether the salience of target features in x-ray mammograms could be enhanced by prior adaptation to the spatial structure of the images. The observers were not radiologists, and thus had no diagnostic training with the images. The stimuli were randomly selected sections from normal mammograms previously classified with BIRADS Density scores of "fatty" versus "dense," corresponding to differences in the relative quantities of fat versus fibroglandular tissue. These categories reflect conspicuous differences in visual texture, with dense tissue being more likely to obscure lesion detection. The targets were simulated masses corresponding to bright Gaussian spots, superimposed by adding the luminance to the background. A single target was randomly added to each image, with contrast varied over five levels so that they varied from difficult to easy to detect. Reaction times were measured for detecting the target location, before or after adapting to a gray field or to random sequences of a different set of dense or fatty images. Observers were faster at detecting the targets in either dense or fatty images after adapting to the specific background type (dense or fatty) that they were searching within. Thus, the adaptation led to a facilitation of search performance that was selective for the background texture. Our results are consistent with the hypothesis that adaptation allows observers to more effectively suppress the specific structure of the background, thereby heightening visual salience and search efficiency.

Errors in radiologic interpretation are largely the result of failures of perception. This remains true despite the increasing use of computer-aided detection and diagnosis. We surveyed the literature on visual illusions during the viewing of radiologic images. Misperception of anatomical structures is a potential cause of error that can lead to patient harm if disease is seen when none is present. However, visual illusions can also help enhance the ability of radiologists to detect and characterize abnormalities. Indeed, radiologists have learned to exploit certain perceptual biases in diagnostic findings and as training tools. We propose that further detailed study of radiologic illusions would help clarify the mechanisms underlying radiologic performance and provide additional heuristics to improve radiologist training and reduce medical error.

Prior research has shown that interruptions lead to a variety of performance costs. However, these costs are heterogenous and poorly understood. Under some circumstances, interruptions lead to large decreases in accuracy on the primary task, whereas in others task duration increases, but task accuracy is unaffected. Presently, the underlying cause of these costs is unclear. The Memory for Goals model suggests that interruptions interfere with the ability to represent the current goal of the primary task. Here, we test the idea that working memory (WM) may play a critical role in representing the current goal and thus may underlie the observed costs associated with interruption. In two experiments, we utilized laboratory-based visual search tasks, which differed in their WM demands, in order to assess how this difference influenced the observed interruption costs. Interruptions led to more severe performance costs when the target of the search changed on each trial. When the search target was consistent across trials, the cost of interruption was greatly reduced. This suggests that the WM demands associated with the primary task play an important role in determining the performance costs of interruption. Our findings suggest that it is important for research to consider the cognitive processes a task engages in order to predict the nature of the adverse effects of interruption in applied settings such as radiology. [ABSTRACT FROM AUTHOR]

Errors in radiologic interpretation are largely the result of failures of perception. This remains true despite the increasing use of computer-aided detection and diagnosis. We surveyed the literature on visual illusions during the viewing of radiologic images. Misperception of anatomical structures is a potential cause of error that can lead to patient harm if disease is seen when none is present. However, visual illusions can also help enhance the ability of radiologists to detect and characterize abnormalities. Indeed, radiologists have learned to exploit certain perceptual biases in diagnostic findings and as training tools. We propose that further detailed study of radiologic illusions would help clarify the mechanisms underlying radiologic performance and provide additional heuristics to improve radiologist training and reduce medical error. [ABSTRACT FROM AUTHOR]

Abstract Interpretation of volumetric medical images represents a rapidly growing proportion of the workload in radiology. However, relatively little is known about the strategies that best guide search behavior when looking for abnormalities in volumetric images. Although there is extensive literature on two-dimensional medical image perception, it is an open question whether the conclusions drawn from these images can be generalized to volumetric images. Importantly, volumetric images have distinct characteristics (e.g., scrolling through depth, smooth-pursuit eye-movements, motion onset cues, etc.) that should be considered in future research. In this manuscript, we will review the literature on medical image perception and discuss relevant findings from basic science that can be used to generate predictions about expertise in volumetric image interpretation. By better understanding search through volumetric images, we may be able to identify common sources of error, characterize the optimal strategies for searching through depth, or develop new training and assessment techniques for radiology residents.

Sensory systems continuously recalibrate their responses according to the current stimulus environment. As a result, perception is strongly affected by the current and recent context. These adaptative changes affect both sensitivity (e.g., habituating to noise, seeing better in the dark) and appearance (e.g. how things look, what catches attention) and adjust to many perceptual properties (e.g. from light level to the characteristics of someone's face). They therefore have a profound effect on most perceptual experiences, and on how and how well the senses work in different settings. Characterizing the properties of adaptation, how it manifests, and when it influences perception in modern environments can provide insights into the diversity of human experience . Adaptation could also be leveraged both to optimize perceptual abilities (e.g. in visual inspection tasks like radiology) and to mitigate unwanted consequences (e.g. exposure to potentially unhealthy stimulus environments).

Abstract Brain tumour detection and diagnosis requires clinicians to inspect and analyse brain magnetic resonance images. Eye-tracking is commonly used to examine observers’ gaze behaviour during such medical image interpretation tasks, but analysis of eye movement sequences is limited. We therefore used ScanMatch, a novel technique that compares saccadic eye movement sequences, to examine the effect of expertise and diagnosis on the similarity of scanning patterns. Diagnostic accuracy was also recorded. Thirty-five participants were classified as Novices, Medics and Experts based on their level of expertise. Participants completed two brain tumour detection tasks. The first was a whole-brain task, which consisted of 60 consecutively presented slices from one patient; the second was an independent-slice detection task, which consisted of 32 independent slices from five different patients. Experts displayed the highest accuracy and sensitivity followed by Medics and then Novices in the independent-slice task. Experts showed the highest level of scanning pattern similarity, with medics engaging in the least similar scanning patterns, for both the whole-brain and independent-slice task. In the independent-slice task, scanning patterns were the least similar for false negatives across all expertise levels and most similar for experts when they responded correctly. These results demonstrate the value of using ScanMatch in the medical image perception literature. Future research adopting this tool could, for example, identify cases that yield low scanning similarity and so provide insight into why diagnostic errors occur and ultimately help in training radiologists.

Summary: Echocardiographers are highly specialised, skilled practitioners who play a critical role in medical imaging diagnostics. Yet, little is known about the cognitive and perceptual attributes of experts within this domain. This study was designed to examine the role of individual differences in expertise. Specifically, the contribution of a domain general visual expertise and pattern recognition, or cue utilization. Data were collected from 42 echocardiographers and 43 naïve participants. All of the participants competed the Novel Object Memory Test (NOMT). When compared, the echocardiographers were more accurate than the naïve participants. The echocardiographers also completed an echocardiography edition of the Expert Intensive Skills Evaluation 2.0, to establish behavioral indicators of context‐related cue utilization. Those with relatively higher cue utilization performed more accurately on the NOMT, controlling for exposure. These results suggest that a general perceptual ability contributes to echocardiography potential and a sensitivity to cue‐based learning may contribute to expertise. [ABSTRACT FROM AUTHOR]

Radiologists make critical decisions based on searching and interpreting medical images. The probability of a lung nodule differs across anatomical regions within the chest, raising the possibility that radiologists might have a prior expectation that creates an attentional bias. The development of expertise is also thought to cause "tuning" to relevant features, allowing radiologists to become faster and more accurate at detecting potential masses within their domain of expertise. Here, we tested both radiologists and control participants with a novel attentional-cueing paradigm to investigate whether the deployment of attention was affected (1) by a context that might invoke prior knowledge for experts, (2) by a nodule localized either on the same or on opposite sides as a subsequent target, and (3) by inversion of the nodule-present chest radiographs, to assess the orientation specificity of any effects. The participants also performed a nodule detection task to verify that our presentation duration was sufficient to extract diagnostic information. We saw no evidence of priors triggered by a normal chest radiograph cue affecting attention. When the cue was an upright abnormal chest radiograph, radiologists were faster when the lateralised nodule and the subsequent target appeared at the same rather than at opposite locations, suggesting attention was captured by the nodule. The opposite pattern was present for inverted images. We saw no evidence of cueing for control participants in any condition, which suggests that radiologists are indeed more sensitive to visual features that are not perceived as salient by naïve observers. [ABSTRACT FROM AUTHOR]

We present a structured approach to combine explainability of artificial intelligence (AI) with the scientific method for scientific discovery. We demonstrate the utility of this approach in a proof-of-concept study where we uncover biomarkers from a convolutional neural network (CNN) model trained to classify patient sex in retinal images. This is a trait that is not currently recognized by diagnosticians in retinal images, yet, one successfully classified by CNNs. Our methodology consists of four phases: In Phase 1, CNN development , we train a visual geometry group (VGG) model to recognize patient sex in retinal images. In Phase 2, Inspiration , we review visualizations obtained from post hoc interpretability tools to make observations, and articulate exploratory hypotheses. Here, we listed 14 hypotheses retinal sex differences. In Phase 3, Exploration , we test all exploratory hypotheses on an independent dataset. Out of 14 exploratory hypotheses, nine revealed significant differences. In Phase 4, Verification , we re-tested the nine flagged hypotheses on a new dataset. Five were verified, revealing (i) significantly greater length, (ii) more nodes, and (iii) more branches of retinal vasculature, (iv) greater retinal area covered by the vessels in the superior temporal quadrant, and (v) darker peripapillary region in male eyes. Finally, we trained a group of ophthalmologists ( N = 26 ) to recognize the novel retinal features for sex classification. While their pretraining performance was not different from chance level or the performance of a nonexpert group ( N = 31 ), after training, their performance increased significantly ( p < 0.001 , d = 2.63 ). These findings showcase the potential for retinal biomarker discovery through CNN applications, with the added utility of empowering medical practitioners with new diagnostic capabilities to enhance their clinical toolkit., (© The Author(s) 2023. Published by Oxford University Press on behalf of National Academy of Sciences.)

Magnetic resonance imaging (MRI) is a crucial tool for clinical brain tumor detection and delineation. Since the process of gross tumor volume delineation resides with clinicians, a better understanding of how they perform this task is required if improvements in life expectancy are to be made. Novice-expert comparison studies have been used to examine the effect of expertise on abnormality detection, but little research has investigated expertise-related differences in brain tumor delineation. In this study, undergraduate students (novices) and radiologists (experts) inspected a combination of T1 and T2 single and whole brain MRI scans, each containing a tumor. Using a tablet and stylus to provide an interactive environment, participants had an unlimited amount of time to scroll freely through the MRI slices and were instructed to delineate (i.e., draw a boundary) around any tumorous tissue. There was no reliable evidence for a difference in the gross tumor volume or total number of slices delineated between experts and novices. Agreement was low across both expertise groups and significantly lower at peripheral locations within a tumor than central locations. There was an interaction between expertise level and location within a tumor with experts displaying higher agreement at the peripheral slices than novices. An effect of brain image set on the order in which participants inspected the slices was also observed. The implications of these results for the training undertaken by early career radiologists and current practices in hospitals are discussed.

In the field of medical image perception, the holistic processing perspective contends that experts can rapidly extract global information about the image, which can be used to guide their subsequent search of the image (Swensson, 1980; Nodine and Kundel, 1987; Kundel et al., 2007). In this review, we discuss the empirical evidence supporting three different predictions that can be derived from the holistic processing perspective: Expertise in medical image perception is domain-specific, experts use parafoveal and/or peripheral vision to process large regions of the image in parallel, and experts benefit from a rapid initial glimpse of an image. In addition, we discuss a pivotal recent study (Litchfield and Donovan, 2016) that seems to contradict the assumption that experts benefit from a rapid initial glimpse of the image. To reconcile this finding with the existing literature, we suggest that global processing may serve multiple functions that extend beyond the initial glimpse of the image. Finally, we discuss future research directions, and we highlight the connections between the holistic processing account and similar theoretical perspectives and findings from other domains of visual expertise.

How we make sense of what we see and where best to look is shaped by our experience, our current task goals and how we first perceive our environment. An established way of demonstrating these factors work together is to study how eye movement patterns change as a function expertise and to observe how experts can solve complex tasks after only very brief glances at a domain-specific image. The primary focus of this paper is to introduce an innovative gaze-contingent method called the ‘Flash-Preview Moving Window’ (FPMW) paradigm (Castelhano & Henderson, 2007), which was recently developed to understand our shared expertise in scene perception and how our first glimpse of a scene is used to guide our eye movement behaviour. In keeping with this special issue on visual expertise and medicine, this paper will highlight how the FPMW paradigm has the potential to resolve long-standing theoretical issues as to how, right from the very first glance, experts are able to process domain-specific images and guide their eye movements better than novices. Since FPMW is a gaze-contingent eye-tracking method, the paper will first outline the current methodological and theoretical frontier, and how the FPMW paradigm bridges established methods used to investigate visual expertise. The paper will discuss a recent example in which the FPMW was employed to investigate medical image perception expertise for first the time (Litchfield & Donovan, 2016), and by confronting the insights and challenges this method offers, this should ultimately deepen our understanding of visual expertise.

The eye movements of experts, reading medical images, have been studied for many years. Unlike topics such as face perception, medical image perception research needs to cope with substantial, qualitative changes in the stimuli under study due to dramatic advances in medical imaging technology. For example, little is known about how radiologists search through 3D volumes of image data because they simply did not exist when earlier eye tracking studies were performed. Moreover, improvements in the affordability and portability of modern eye trackers make other, new studies practical. Here, we review some uses of eye movements in the study of medical image perception with an emphasis on newer work. We ask how basic research on scene perception relates to studies of medical ‘scenes’ and we discuss how tracking experts’ eyes may provide useful insights for medical education and screening efficiency.

One of the most important applications of visual search is in the interpretation of medical images. Like many applications of visual search, medical image interpretation is typically characterized as a low 'prevalence' context, meaning that targets are relatively rare. Breast cancer screening, for example, has a prevalence of around 0.5%. In this review, I present a survey of research on the effects of prevalence on visual search from both the medical image perception and basic psychological science perspectives. In order to provide some quantitative rigor, I include a meta-analysis of the effects of low prevalence on visual search performance. I synthesize findings from both fields, and identify both the practical suggestions for clinical practice and theoretical insights for cognitive psychologists. [ABSTRACT FROM AUTHOR]

In socially important visual search tasks, such as baggage screening and diagnostic radiology, experts miss more targets than is desirable. Computer-aided detection (CAD) programs have been developed specifically to improve performance in these professional search tasks. For example, in breast cancer screening, many CAD systems are capable of detecting approximately 90% of breast cancer, with approximately 0.5 false-positive detections per image. Nevertheless, benefits of CAD in clinical settings tend to be small (Birdwell, 2009) or even absent (Meziane et al., 2011; Philpotts, 2009). The marks made by a CAD system can be 'binary,' giving the same signal to any location where the signal is above some threshold. Alternatively, a CAD system presents an analog signal that reflects strength of the signal at a location. In the experiments reported, we compare analog and binary CAD presentations using nonexpert observers and artificial stimuli defined by two noisy signals: a visible color signal and an 'invisible' signal that informed our simulated CAD system. We found that analog CAD generally yielded better overall performance than binary CAD. The analog benefit is similar at high and low target prevalence. Our data suggest that the form of the CAD signal can directly influence performance. Analog CAD may allow the computer to be more helpful to the searcher. [ABSTRACT FROM AUTHOR]

Combining perceptual learning techniques with adaptive learning algorithms has been shown to accelerate the development of expertise in medical and STEM learning domains (Kellman & Massey, 2013; Kellman, Jacoby, Massey & Krasne, 2022). Virtually all adaptive learning systems have relied on simple accuracy data that does not take into account response bias, a problem that may be especially consequential in multi-category perceptual classifications. We investigated whether adaptive perceptual learning in skin cancer screening can be enhanced by incorporating signal detection theory (SDT) methods that separate sensitivity from criterion. SDT-style concepts were used to alter sequencing, and separately to define mastery (category retirement). SDT retirement used a running d' estimate calculated from a recent window of trials based on hit and false alarm rates. Undergraduate participants used a Skin Cancer PALM (perceptual adaptive learning module) to learn classification of 10 cancerous and readily-confused non-cancerous skin lesion types. Four adaptive conditions varied either the type of adaptive sequencing (standard vs. SDT) or retirement criteria (standard vs. SDT). A non-adaptive control condition presented didactic instruction on dermatologic screening in video form, including images, classification schemes, and detailed explanations. All adaptive conditions robustly outperformed the non-adaptive control in both learning efficiency and fluency (large effect sizes). Between adaptive conditions, SDT retirement criteria produced greater learning efficiency than standard, accuracy-based mastery criteria at both immediate and delayed posttests (medium effect sizes). SDT sequencing and standard adaptive sequencing did not differ. SDT enhancements to adaptive perceptual learning procedures have potential to enhance learning efficiency.

Humans are very adept at extracting the "gist" of a scene in a fraction of a second. We have found that radiologists can discriminate normal from abnormal mammograms at above-chance levels after a half-second viewing (d' ~ 1) but are at chance in localizing the abnormality. This pattern of results suggests that they are detecting a global signal of abnormality. What are the stimulus properties that might support this ability? We investigated the nature of the gist signal in four experiments by asking radiologists to make detection and localization responses about briefly presented mammograms in which the spatial frequency, symmetry, and/or size of the images was manipulated. We show that the signal is stronger in the higher spatial frequencies. Performance does not depend on detection of breaks in the normal symmetry of left and right breasts. Moreover, above-chance classification is possible using images from the normal breast of a patient with overt signs of cancer only in the other breast. Some signal is present in the portions of the parenchyma (breast tissue) that do not contain a lesion or that are in the contralateral breast. This signal does not appear to be a simple assessment of breast density but rather the detection of the abnormal gist may be based on a widely distributed image statistic, learned by experts. The finding that a global signal, related to disease, can be detected in parenchyma that does not contain a lesion has implications for improving breast cancer detection. [ABSTRACT FROM AUTHOR]

Introduction: Radiologists routinely make life-altering decisions. Optimizing these decisions has been an important goal for many years and has prompted a great deal of research on the basic perceptual mechanisms that underlie radiologists' decisions. Previous studies have found that there are substantial individual differences in radiologists' diagnostic performance (e.g., sensitivity) due to experience, training, or search strategies. In addition to variations in sensitivity, however, another possibility is that radiologists might have perceptual biases-systematic misperceptions of visual stimuli. Although a great deal of research has investigated radiologist sensitivity, very little has explored the presence of perceptual biases or the individual differences in these., Methods: Here, we test whether radiologists' have perceptual biases using controlled artificial and Generative Adversarial Networks-generated realistic medical images. In Experiment 1, observers adjusted the appearance of simulated tumors to match the previously shown targets. In Experiment 2, observers were shown with a mix of real and GAN-generated CT lesion images and they rated the realness of each image., Results: We show that every tested individual radiologist was characterized by unique and systematic perceptual biases; these perceptual biases cannot be simply explained by attentional differences, and they can be observed in different imaging modalities and task settings, suggesting that idiosyncratic biases in medical image perception may widely exist., Discussion: Characterizing and understanding these biases could be important for many practical settings such as training, pairing readers, and career selection for radiologists. These results may have consequential implications for many other fields as well, where individual observers are the linchpins for life-altering perceptual decisions., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2022 Wang, Manassi, Ren, Ghirardo, Canas-Bajo, Murai, Zhou and Whitney.)

The objective of the present study is to explore prioritization methods in diagnostic imaging modalities to automatically determine the contents of medical images. In this paper, we propose an efficient prioritization of brain MRI. First, the visual perception of the radiologists is adapted to identify salient regions. Then this saliency information is used as an automatic label for accurate segmentation of brain lesion to determine the scientific value of that image. The qualitative and quantitative results prove that the rankings generated by the proposed method are closer to the rankings created by radiologists. [ABSTRACT FROM AUTHOR]

Rationale and Objectives: Image perception studies have been difficult to perform using clinical images because of the problems associated with obtaining proven abnormalities and appropriate normal controls. The objective of this research was to develop and evaluate interactive software that allows the seamless removal, archiving and insertion of abnormal areas from computed tomography (CT) lung image sets for use in image perception research. Materials and Methods: The software tools for removing, archiving, and adding lesions are described in detail. The efficacy of the software to remove abnormal areas of lung CT studies was evaluated by having radiologists select the one altered image from a display of four. The software for adding lesions was evaluated by having radiologists classify displayed CT slices with lesions as real or artificial along with their confidence level. Results: Observers could not reliably detect when images had been altered by the software. In the lesion-removal experiment, the observers correctly identified the altered display in only 15.8 +/- 2.8 of 56 sets. In the lesion-add experiment, the observers correctly identified the artificially placed lesions in 38.2 +/- 3.9 of 77 sets. The frequency distribution of the correct responses did not differ from that expected from chance selection. Conclusions: The results from both of these experiments demonstrate that radiologists could not distinguish between original and altered images. We conclude that this software can be used with volumetric CT lung images for creating normal control and target data sets for medical image perception research. [ABSTRACT FROM AUTHOR]

Many important real-world decision tasks involve the detection of rarely occurring targets (e.g., weapons in luggage, potentially cancerous abnormalities in radiographs). Over the past decade, it has been repeatedly demonstrated that extreme prevalence (both high and low) leads to an increase in errors. While this "prevalence effect" is well established, the cognitive and/or perceptual mechanisms responsible for it are not. One reason for this is that the most common tool for analyzing prevalence effects, Signal Detection Theory, cannot distinguish between different biases that might be present. Through an application to pathology image-based decision-making, we illustrate that an evidence accumulation modeling framework can be used to disentangle different types of biases. Importantly, our results show that prevalence influences both response expectancy and stimulus evaluation biases, with novices (students, N = 96) showing a more pronounced response expectancy bias and experts (medical laboratory professionals, N = 19) showing a more pronounced stimulus evaluation bias. [ABSTRACT FROM AUTHOR]