Search

Abstract Pupillary contagion occurs when one’s pupil size unconsciously adapts to the pupil size of an observed individual and is presumed to reflect the transfer of arousal. Importantly, when estimating pupil contagion, low level stimuli properties need to be controlled for, to ensure that observations of pupillary changes are due to internal change in arousal rather than the external differences between stimuli. Here, naturalistic images of children’s faces depicting either small or large pupils were presented to a group of children and adolescents with a wide range of autistic traits, a third of whom had been diagnosed with autism. We examined the extent to which pupillary contagion reflects autonomic nervous system reaction through pupil size change, heart rate and skin conductance response. Our second aim was to determine the association between arousal reaction to stimuli and degree of autistic traits. Results show that pupil contagion and concomitant heart rate change, but not skin conductance change, was evident when gaze was restricted to the eye region of face stimuli. A positive association was also observed between pupillary contagion and autistic traits when participants’ gaze was constrained to the eye region. Findings add to a broader understanding of the mechanisms underlying pupillary contagion and its association with autism.

Abstract Williams syndrome (WS) is associated with atypical social communication and cognition reminiscent of the behaviours observed in autism. Nonetheless, WS also differs significantly from autism, such as regarding social motivation, which is typically enhanced in WS and reduced in autism. This study sought to examine the conditions’ transdiagnostic similarities and differences for autistic symptoms and social functioning, and their developmental trajectories, by comparing individuals with WS (n = 24) and those diagnosed with idiopathic autism (n = 24) and attention deficit hyperactivity disorder (ADHD; n = 24), aged 9 to 53 years, on measures of autism, social functioning, IQ and cooccurring psychiatric conditions. Although only 12.5% in the WS group met the criteria for an autism diagnosis, a majority exhibited distinct difficulties within social communication, social cognition, repetitive behaviours, and atypical sensory reactivity resembling autism. Conversely, elevated social motivation and a high number of social initiatives accompany these characteristics. No group differences in the developmental trajectories of autism symptoms were found. Our results demonstrate that autistic behaviours are more frequent in individuals with WS, than in individuals with idiopathic ADHD, and emphasize the need for clinical management of these behaviours.

IntroductionAccess to evidence-based psychological therapy for posttraumatic stress disorder (PTSD) is limited. Blended Trauma-focused Cognitive Therapy (bTF-CT), merging internet-modules with a few therapy sessions, may be a pathway to enhance treatment access while maintaining the benefits of face-to-face therapy.ObjectivesThis study aimed to evaluate the feasibility, acceptability, and preliminary effectiveness of bTF-CT for PTSD in routine care, as well as the feasibility of assessments and data collection.MethodA single-arm design was adopted. bTF-CT was provided to 17 participants across two psychiatric and one primary care clinic. Assessments were conducted pre, during, post and 6-months following treatment. We assessed feasibility and acceptability via self-report questionnaires, retention, and attrition rates. To estimate preliminary treatment effectiveness the PTSD Symptom Checklist (PCL-5) was used to assess PTSD symptom severity.ResultsTreatment satisfaction was high with a mean score of 28.7 out of 32 on the Client Satisfaction Questionnaire (SD = 3.5). The dropout rate was low, with 88% treatment retention. Program adherence was satisfactory, with scores ranging from 2.13 to 3.13 out of 4 on the internet intervention patient adherence scale. On the PCL-5, 88% made a reliable change, 64% demonstrated a clinically significant change, and the mean change from pre to post was 24 points (d = 2.13). Some negative effects were reported, such as unpleasant memories, feelings, and disrupted sleep, but these were temporary according to symptom scales.ConclusionsbTF-CT appears to be acceptable, feasible, and potentially effective when delivered in routine care. A large-scale non-inferiority trial to assess effectiveness compared to a gold-standard treatment is warranted.Clinical Trial RegistrationClinicaltrials.gov, identifier NCT04881643.

The tall cell subtype (TC-PTC) is an aggressive subtype of papillary thyroid carcinoma (PTC). The TC-PTC is defined as a PTC comprising at least 30% epithelial cells that are three times as tall as they are wide. In practice, this definition is difficult to adhere to, resulting in high inter-observer variability. In this multicenter study, we validated a previously trained deep learning (DL)-based algorithm for detection of tall cells on 160 externally collected hematoxylin and eosin (HE)-stained PTC whole-slide images. In a test set of 360 manual annotations of regions of interest from 18 separate tissue sections in the external dataset, the DL-based algorithm detected TCs with a sensitivity of 90.6% and a specificity of 88.5%. The DL algorithm detected non-TC areas with a sensitivity of 81.6% and a specificity of 92.9%. In the validation datasets, 20% and 30% TC thresholds correlated with a significantly shorter relapse-free survival. In conclusion, the DL algorithm detected TCs in unseen, external scanned HE tissue slides with high sensitivity and specificity without any retraining.

BackgroundInfections caused by soil-transmitted helminths (STHs) are the most prevalent neglected tropical diseases and result in a major disease burden in low- and middle-income countries, especially in school-aged children. Improved diagnostic methods, especially for light intensity infections, are needed for efficient, control and elimination of STHs as a public health problem, as well as STH management. Image-based artificial intelligence (AI) has shown promise for STH detection in digitized stool samples. However, the diagnostic accuracy of AI-based analysis of entire microscope slides, so called whole-slide images (WSI), has previously not been evaluated on a sample-level in primary healthcare settings in STH endemic countries.Methodology/principal findingsStool samples (n = 1,335) were collected during 2020 from children attending primary schools in Kwale County, Kenya, prepared according to the Kato-Katz method at a local primary healthcare laboratory and digitized with a portable whole-slide microscopy scanner and uploaded via mobile networks to a cloud environment. The digital samples of adequate quality (n = 1,180) were split into a training (n = 388) and test set (n = 792) and a deep-learning system (DLS) developed for detection of STHs. The DLS findings were compared with expert manual microscopy and additional visual assessment of the digital samples in slides with discordant results between the methods. Manual microscopy detected 15 (1.9%) Ascaris lumbricoides, 172 (21.7%) Tricuris trichiura and 140 (17.7%) hookworm (Ancylostoma duodenale or Necator americanus) infections in the test set. Importantly, more than 90% of all STH positive cases represented light intensity infections. With manual microscopy as the reference standard, the sensitivity of the DLS as the index test for detection of A. lumbricoides, T. trichiura and hookworm was 80%, 92% and 76%, respectively. The corresponding specificity was 98%, 90% and 95%. Notably, in 79 samples (10%) classified as negative by manual microscopy for a specific species, STH eggs were detected by the DLS and confirmed correct by visual inspection of the digital samples.Conclusions/significanceAnalysis of digitally scanned stool samples with the DLS provided high diagnostic accuracy for detection of STHs. Importantly, a substantial number of light intensity infections were missed by manual microscopy but detected by the DLS. Thus, analysis of WSIs with image-based AI may provide a future tool for improved detection of STHs in a primary healthcare setting, which in turn could facilitate monitoring and evaluation of control programs.

This paper addresses concerns related to the ethical implications of artificial intelligence (AI) and its impact on human values, with a particular focus on fair outcomes. Existing design frameworks and regulations for ensuring fairness in AI are too general and impractical. Instead, we advocate for understanding fairness as situated in practice, shaped by practitioners’ values, allowing stakeholders control in the situation. To accomplish this, the paper contributes by conceptually exploring a potential synergy by combining Cultural-Historical Activity Theory (CHAT) and Meta-Design. By doing so, human activities can be transformed to deal with challenges, in this case, those emerging from adaptive AI tools. While professional software developers are essential for making significant changes to the tool and providing solutions, users’ involvement is equally important. Users are domain experts when it comes to determining practical solutions and aligning structures with their work practices. CHAT contributes through its emphasis on context, history, and mediation by tools. This enables a critical analysis of activity systems, helping to reveal underlying contradictions and identify areas where improvements or innovations are necessary. Meta-Design provides design concepts and perspectives that aim to empower participants, allowing them to actively shape the processes of tool design to align with their specific local needs and evolving conceptions of fairness in use-time. This offers an approach to empowering people and promoting more fair AI design.

Abstract This eye-tracking study investigated the effect of sound-induced arousal on social orienting under different auditory cue conditions in 5-month-old (n = 25; n = 13 males) and 10-month-old infants (n = 21; n = 14 males) participating in a spontaneous visual search task. Results showed: (1) larger pupil dilation discriminating between high and low volume (b = 0.02, p = 0.007), but not between social and non-social sounds (b = 0.004, p = 0.64); (2) faster visual orienting (b = − 0.09, p

Abstract Typically developing humans automatically synchronize their arousal levels, resulting in pupillary contagion, or spontaneous adaptation of pupil size to that of others. This phenomenon emerges in infancy and is believed to facilitate social interaction. Williams syndrome (WS) is a genetic condition characterized by a hyper-social personality and social interaction challenges. Pupillary contagion was examined in individuals with WS (n = 44), age-parallel-matched typically developing children and adults (n = 65), and infants (n = 79). Bayesian statistics were used. As a group, people with WS did not show pupillary contagion (Bayes factors supporting the null: 25–50) whereas control groups did. This suggests a very early emerging atypical developmental trajectory. In WS, higher pupillary contagion was associated with lower autistic symptoms of social communication. Diminished synchronization of arousal may explain why individuals with WS have social challenges, whereas synchronization of arousal is not a necessary correlate of high social motivation.

Abstract Turner syndrome is a genetic condition caused by a complete or partial loss of one of the X chromosomes. Previous studies indicate that Turner syndrome is associated with challenges in social skills, but the underlying mechanisms remain largely unexplored. A possible mechanism is a reduced social influence on learning. The current study examined the impact of social and non-social feedback on learning in women with Turner syndrome (n = 35) and a sex- and age-matched control group (n = 37). Participants were instructed to earn points by repeatedly choosing between two stimuli with unequal probabilities of resulting in a reward. Mastering the task therefore required participants to learn through feedback which of the two stimuli was more likely to be rewarded. Data were analyzed using computational modeling and analyses of choice behavior. Social feedback led to a more explorative choice behavior in the control group, resulting in reduced learning compared to non-social feedback. No effects of social feedback on learning were found in Turner syndrome. The current study thus indicates that women with Turner syndrome may be less sensitive to social influences on reinforcement learning, than the general population.

Abstract Background Children with brain tumors are at high risk of neurocognitive decline after radiotherapy (RT). However, there is a lack of studies on how RT doses to organs at risk (OARs) impacts neurocognition. The aim of this study was to examine dose-risk relationships for mean RT dose to different brain structures important for neurocognitive networks. We explored previously established OARs and potentially new OARs. Methods A sample of 44 pediatric brain tumor survivors who had received proton and/or photon RT were included. Correlations between mean RT doses to OARs and IQ were analyzed. Previously established OARs were cochleae, optic chiasm, optic nerve, pituitary gland, hypothalamus, hippocampus and pons. Potential new OARs for RT-induced neurocognitive decline were cerebellum, vermis and thalamus. Results Mean RT dose to different OARs correlated with several IQ subtests. Higher mean RT dose to cochleae, optic nerve, cerebellum, vermis and pons was correlated with lower performance on particularly full-scale IQ (FIQ), Perceptual Reasoning (PRI), Working Memory (WMI) and Processing Speed Index (PSI). Higher mean RT dose to hippocampus correlated with lower performance on processing speed and working memory. For those receiving whole brain RT (WBRT), higher mean RT dose to the pituitary gland correlated with lower performance on working memory. Conclusion A high dose-risk correlation was found between IQ subtests and mean RT dose in established and potential new OARs. Thus, in the lack of validated dose constraints for vulnerable brain structures, a parsimonious approach in RT planning should be considered to preserve neurocognitive networks.

AbstractSpeaking in a foreign language is considered challenging to both teach and learn. Virtual humans (VHs), as conversational agents (CAs), provide opportunities to practise speaking skills. Lower secondary school students (N = 25) engaged in an AI-based spoken dialogue system (SDS) and interacted verbally with VHs in simulated everyday-life scenarios to solve given tasks. Our analysis is based on system-generated metrics and self-reported experiences collected through questionnaires, logbooks, and interviews. Thematic analysis resulted in seven themes, revolving around the speaking practice method, scenarios and technology, which, in combination with descriptive statistics, enabled a deeper understanding of the students’ experiences. The results indicate that, on average, they found it easy, fun, and safe, but sometimes frustrating in scenarios not always relevant to their everyday lives. Factors suggested as underlying the levels of experienced frustration include technical issues and constraints with the system, such as not being understood or heard as expected. The findings suggest that lower secondary school students conversing with VHs in the SDS in an institutional educational context facilitated a beneficial opportunity for practising speaking skills, especially pronunciation and interaction in dialogues, aligning with the key principles of second language acquisition (SLA) for language development.

Abstract Attention-deficit/hyperactivity disorder (ADHD) follows a variable course across childhood. Disrupted arousal has been hypothesized to underlie core symptoms as well as comorbid internalizing and externalizing conditions. The current study examined eye-movement and pupil-dilation metrics indexing arousal as longitudinal predictors of ADHD, externalizing, and internalizing symptoms over a 2-year period. Participants aged 8–13 years (N = 54, 30% with a diagnosis of ADHD) completed a modified version of the gap-overlap task including arousal-inducing auditory warning signals. Parents rated symptoms at the time of testing and at 2 years follow-up. Phasic alerting (reaction-time reduction after alerting cues) is an index of arousal. Here, larger phasic alerting effects predicted higher ADHD-symptom levels 2 years later. Blunted pupil-dilation responses predicted externalizing symptoms at T2, controlling for ADHD and externalizing at T1. Our results support the theory that ADHD is associated with altered arousal. Blunted arousal reactivity may be a longitudinal risk factor for externalizing problems in children with ADHD symptoms.

Abstract Assessment of burn extent and depth are critical and require very specialized diagnosis. Automated image-based algorithms could assist in performing wound detection and classification. We aimed to develop two deep-learning algorithms that respectively identify burns, and classify whether they require surgery. An additional aim assessed the performances in different Fitzpatrick skin types. Annotated burn (n = 1105) and background (n = 536) images were collected. Using a commercially available platform for deep learning algorithms, two models were trained and validated on 70% of the images and tested on the remaining 30%. Accuracy was measured for each image using the percentage of wound area correctly identified and F1 scores for the wound identifier; and area under the receiver operating characteristic (AUC) curve, sensitivity, and specificity for the wound classifier. The wound identifier algorithm detected an average of 87.2% of the wound areas accurately in the test set. For the wound classifier algorithm, the AUC was 0.885. The wound identifier algorithm was more accurate in patients with darker skin types; the wound classifier was more accurate in patients with lighter skin types. To conclude, image-based algorithms can support the assessment of acute burns with relatively good accuracy although larger and different datasets are needed.

Abstract Williams syndrome (WS) is a rare genetic condition characterized by high social interest and approach motivation as well as intellectual disability and anxiety. Despite the fact that social stimuli are believed to have an increased intrinsic reward value in WS, it is not known whether this translates to learning and decision making. Genes homozygously deleted in WS are linked to sociability in the general population, making it a potential model condition for understanding the social brain. Probabilistic reinforcement learning was studied with either social or non-social rewards for correct choices. Social feedback improved learning in individuals with Williams syndrome but not in typically developing controls or individuals with other intellectual disabilities. Computational modeling indicated that these effects on social feedback were mediated by a shift towards higher weight given to rewards relative to punishments and increased choice consistency. We conclude that reward learning in WS is characterized by high volatility and a tendency to learn how to avoid punishment rather than how to gain rewards. Social feedback can partly normalize this pattern and promote adaptive reward learning.

Abstract Background Clinical implications of different types of vascular calcification are poorly understood. The two most abundant forms of calcification, nodular and sheet calcification, have not been quantitatively analyzed in relation to the clinical presentation of lower extremity arterial disease (LEAD). Methods The study analyzed 51 femoral artery plaques collected during femoral endarterectomy, characterized by the presence of > 90% stenosis. Comprehensive clinical data was obtained from patient records, including magnetic resonance angiography (MRA) images, toe pressure and ankle brachial index measurements and laboratory values. The plaques were longitudinally sectioned, stained with Hematoxylin and Eosin and digitized in a deep learning platform for quantification of the relative area of nodular and sheet calcification to the plaque section area. A deep learning artificial intelligence algorithm was designed and independently validated to reliably quantify nodular calcification and sheet calcification. Vessel measurements and quantity of each calcification category was compared to the risk factors and clinical presentation. Results On average, > 90% stenosed vessels contained 22.4 ± 12.3% of nodular and 14.5 ± 11.8% of sheet calcification. Nodular calcification area proportion in lesions with > 90% stenosis is associated with reduced risk of critically low toe pressure (

Abstract Background Individuals with Williams syndrome (WS) have an elevated risk for anxiety disorders throughout the life span, making it a research priority to identify the individual factors associated with anxiety. Most of the existing literature is based on questionnaire data and suggests that impaired executive functions (EF) increase the risk for anxiety in WS. The aim of this study was to use direct measures by trained clinicians to investigate the effects of general intelligence, inhibition, sustained attention, and working memory on anxiety in WS, to further elucidate potential underlying mechanisms. Method Twenty-four individuals with WS participated in the study (mean age: 29 years, range: 9–53 years), together with at least one of their parents. The MINI international neuropsychiatric interview for DSM-5 was completed to establish clinical diagnosis of anxiety, and the Clinical Global Impression Scale – Severity was used for an expert rating of symptom severity. Intellectual abilities were measured using the Wechsler scales, and attention and inhibition using the Conner’s Continuous Performance Test. In addition, a parent-report questionnaire measuring EF, learning and memory was collected. Results In contrast to the apriori hypothesis, there was no significant association between anxiety and core elements of EF such as working memory, sustained attention, and inhibition (i.e. the process of restraining one’s impulses or behaviour). Using ordinal logistic regression analyses, we showed that decreasing intelligence quotient (IQ) and age are associated with elevated anxiety. We confirmed these results in between-groups analyses (anxiety disorder vs no current anxiety disorder), and low IQ was associated with higher risk of having an anxiety diagnosis. In addition, Bayesian statistics gave substantial evidence for no significant association between anxiety and inhibition. Conclusion By using direct measures of psychological pathology and functioning, the current results provide a deeper characterisation of the WS phenotype and provide novel insights into the potential mechanisms underpinning anxiety.

When the COVID-19 pandemic started spreading globally, there was a fear that addictive behaviors would increase due to changes in everyday life caused by restrictions due to COVID-19. Studies were carried out to explore if this was true for gambling, typically revealing no overall increase in gambling behavior, although individuals who had previous experience with gambling problems were more likely to increase gambling during the pandemic. However, these studies only included individuals with previous gambling problems. It remains unknown whether other vulnerable groups, such as individuals with common mental disorders increased their gambling. This study aimed to explore the level of gambling problems among individuals with a history of mental disorders, namely, (i) pre-pandemic gamblers and (ii) pandemic-onset gamblers. Furthermore, we explored if worry and isolation mediate gambling and problem gambling. The data were analyzed using descriptive statistics and a structural equation model to investigate mediation. The results showed a high prevalence of at-risk and problem gambling in both groups. The pre-pandemic gamblers had a high level of at-risk and problem gambling. Furthermore, the individuals that started to gamble during the pandemic had an even higher degree of at-risk and problem gambling. The mediation showed that the onset of gambling was linked with the worry of COVID-infection and that worry predicted the level of gambling problems. This study highlights that vulnerability factors, isolation, and worry can be triggers for individuals with common mental disorders to engage in gambling as well as the importance of screening this population for gambling problems.

ObjectiveIndividuals with psychiatric disorders may be both vulnerable and sensitive to rapid societal changes that have occurred during the COVID-19 pandemic. To fully understand these impacts, repeated measurements of these individuals are warranted. The current longitudinal study set out to perform monthly assessment of individuals with common psychiatric disorders using established questionnaires with a possibility for them to self- rate their symptoms, over time.MethodsRecruitment of individuals who identified themselves as struggling with mental health problems, living in Sweden between July 2020 and June 2021 using an online survey. The individuals answered questions on demographics, psychiatric history, current psychiatric symptoms (e.g., Patient Health Questionnaire, PHQ-9; General Anxiety Disorder, GAD-7), somatic health, health-care contacts and any changes therein during the pandemic. Monthly, longitudinal assessments are still ongoing (consenting participants provide data for 1 year), and here we present descriptive statistics from the baseline measurement. All measurements from baseline (>400 items), and follow-ups are presented in detail.ResultsA total of 6.095 participants (average age 35 years) submitted complete baseline data. Marital status (43% single) and number of years of education (48% highest degree being high school) were evenly distributed in this population. The most common lifetime psychiatric disorder in the sample was depressive disorder (80.5%) and generalized anxiety disorder (45.9%), with a substantial proportion having severe symptoms of depression. (30.5%) and anxiety (37.1%). Lifetime suicidal ideation (75.0%) and non-suicidal self-harm (57.7%) were prevalent in the group and 14.5% reported drug use during the pandemic. Allergies (36.8%) were the most common somatic condition, followed by irritable bowel syndrome (18.7%). For those having experienced a traumatic event, 39% showed symptoms during the pandemic indicating PTSD. Regarding contact with mental health services during the pandemic, 22% had established a new contact, and 20% reported to have increased their psychiatric medication compared to before the pandemic.ConclusionBaseline data collected during the pandemic from individuals in Sweden with pre-existing psychiatric disorders demonstrate that this sample represents a population suitable for an investigation on the long-term impact of the pandemic, as intended by the longitudinal investigation that is ongoing. Follow-up questionnaires over a 12-month period are being collected and will indicate how the health and well-being of this population was impacted during the changes and uncertainties that have been characteristic of the past 2 years.

Abstract The treatment of patients with ERBB2 (HER2)-positive breast cancer with anti-ERBB2 therapy is based on the detection of ERBB2 gene amplification or protein overexpression. Machine learning (ML) algorithms can predict the amplification of ERBB2 based on tumor morphological features, but it is not known whether ML-derived features can predict survival and efficacy of anti-ERBB2 treatment. In this study, we trained a deep learning model with digital images of hematoxylin–eosin (H&E)-stained formalin-fixed primary breast tumor tissue sections, weakly supervised by ERBB2 gene amplification status. The gene amplification was determined by chromogenic in situ hybridization (CISH). The training data comprised digitized tissue microarray (TMA) samples from 1,047 patients. The correlation between the deep learning–predicted ERBB2 status, which we call H&E-ERBB2 score, and distant disease-free survival (DDFS) was investigated on a fully independent test set, which included whole-slide tumor images from 712 patients with trastuzumab treatment status available. The area under the receiver operating characteristic curve (AUC) in predicting gene amplification in the test sets was 0.70 (95% CI, 0.63–0.77) on 354 TMA samples and 0.67 (95% CI, 0.62–0.71) on 712 whole-slide images. Among patients with ERBB2-positive cancer treated with trastuzumab, those with a higher than the median morphology–based H&E-ERBB2 score derived from machine learning had more favorable DDFS than those with a lower score (hazard ratio [HR] 0.37; 95% CI, 0.15–0.93; P = 0.034). A high H&E-ERBB2 score was associated with unfavorable survival in patients with ERBB2-negative cancer as determined by CISH. ERBB2-associated morphology correlated with the efficacy of adjuvant anti-ERBB2 treatment and can contribute to treatment-predictive information in breast cancer.

Introduction Burn injuries are a common traumatic injury. Large burns have high mortality requiring intensive care and accurate mortality predictions. To assess if machine learning (ML) could improve predictions, ML algorithms were tested and compared with the original and revised Baux score. Methods Admission data and mortality outcomes were collected from patients at Uppsala University Hospital Burn Centre from 2002 to 2019. Prognostic variables were selected, ML algorithms trained and predictions assessed by analysis of the area under the receiver operating characteristic curve (AUC). Comparison was made with Baux scores using DeLong test Results A total of 17 prognostic variables were selected from 92 patients. AUCs in leave-one-out cross-validation for a decision tree model, an extreme boosting model, a random forest model, a support-vector machine (SVM) model and a generalised linear regression model (GLM) were 0.83 (95% confidence interval [CI] = 0.72–0.94), 0.92 (95% CI = 0.84–1), 0.92 (95% CI = 0.84–1), 0.92 (95% CI = 0.84–1) and 0.84 (95% CI = 0.74–0.94), respectively. AUCs for the Baux score and revised Baux score were 0.85 (95% CI = 0.75–0.95) and 0.84 (95% CI = 0.74–0.94). No significant differences were observed when comparing ML algorithms with Baux score and revised Baux score. Secondary variable selection was made to analyse model performance. Conclusion This proof-of-concept study showed initial credibility in using ML algorithms to predict mortality in burn patients. The sample size was small and future studies are needed with larger sample sizes, further variable selections and prospective testing of the algorithms. Lay Summary Burn injuries are one of the most common traumatic injuries especially in countries with limited prevention and healthcare resources. To treat a patient with large burns who has been admitted to an intensive care unit, it is often necessary to assess the risk of a fatal outcome. Physicians traditionally use simplified scores to calculate risks. One commonly used score, the Baux score, uses age of the patient and the size of the burn to predict the risk of death. Adding the factor of inhalation injury, the score is then called the revised Baux score. However, there are a number of additional causes that can influence the risk of fatal outcomes that Baux scores do not take into account. Machine learning is a method of data modelling where the system learns to predict outcomes based on previous cases and is a branch of artificial intelligence. In this study we evaluated several machine learning methods for outcome prediction in patients admitted for burn injury. We gathered data on 93 patients at admission to the intensive care unit and our experiments show that machine learning methods can reach an accuracy comparable with Baux scores in calculating the risk of fatal outcomes. This study represents a proof of principle and future studies on larger patient series are required to verify our results as well as to evaluate the methods on patients in real-life situations.

Background: Prediction of clinical outcomes for individual cancer patients is an important step in the disease diagnosis and subsequently guides the treatment and patient counseling. In this work, we develop and evaluate a joint outcome and biomarker supervised (estrogen receptor expression and ERBB2 expression and gene amplification) multitask deep learning model for prediction of outcome in breast cancer patients in two nation-wide multicenter studies in Finland (the FinProg and FinHer studies). Our approach combines deep learning with expert knowledge to provide more accurate, robust, and integrated prediction of breast cancer outcomes. Materials and methods: Using deep learning, we trained convolutional neural networks (CNNs) with digitized tissue microarray (TMA) samples of primary hematoxylin-eosin-stained breast cancer specimens from 693 patients in the FinProg series as input and breast cancer-specific survival as the endpoint. The trained algorithms were tested on 354 TMA patient samples in the same series. An independent set of whole-slide (WS) tumor samples from 674 patients in another multicenter study (FinHer) was used to validate and verify the generalization of the outcome prediction based on CNN models by Cox survival regression and concordance index (c-index). Visual cancer tissue characterization, i.e., number of mitoses, tubules, nuclear pleomorphism, tumor-infiltrating lymphocytes, and necrosis was performed on TMA samples in the FinProg test set by a pathologist and combined with deep learning-based outcome prediction in a multitask algorithm. Results: The multitask algorithm achieved a hazard ratio (HR) of 2.0 (95% confidence interval [CI] 1.30–3.00), P

IntroductionAccording to the World Health Organization, the tall cell variant (TCV) is an aggressive subtype of papillary thyroid carcinoma (PTC) comprising at least 30% epithelial cells two to three times as tall as they are wide. In practice, applying this definition is difficult causing substantial interobserver variability. We aimed to train a deep learning algorithm to detect and quantify the proportion of tall cells (TCs) in PTC.MethodsWe trained the deep learning algorithm using supervised learning, testing it on an independent dataset, and further validating it on an independent set of 90 PTC samples from patients treated at the Hospital District of Helsinki and Uusimaa between 2003 and 2013. We compared the algorithm-based TC percentage to the independent scoring by a human investigator and how those scorings associated with disease outcomes. Additionally, we assessed the TC score in 71 local and distant tumor relapse samples from patients with aggressive disease.ResultsIn the test set, the deep learning algorithm detected TCs with a sensitivity of 93.7% and a specificity of 94.5%, whereas the sensitivity fell to 90.9% and specificity to 94.1% for non-TC areas. In the validation set, the deep learning algorithm TC scores correlated with a diminished relapse-free survival using cutoff points of 10% (p = 0.044), 20% (p < 0.01), and 30% (p = 0.036). The visually assessed TC score did not statistically significantly predict survival at any of the analyzed cutoff points. We observed no statistically significant difference in the TC score between primary tumors and relapse tumors determined by the deep learning algorithm or visually.ConclusionsWe present a novel deep learning-based algorithm to detect tall cells, showing that a high deep learning-based TC score represents a statistically significant predictor of less favorable relapse-free survival in PTC.

Abstract Background Recent theories have linked autism to challenges in prediction learning and social cognition. It is unknown, however, how autism affects learning about threats from others “demonstrators” through observation, which contains predictive learning based on social information. The aims of this study are therefore to investigate social fear learning in individual with autism spectrum disorder (ASD) and to examine whether typically developing social cognition is necessary for successful observational learning. Methods Adults with ASD (n = 23) and neurotypical controls (n = 25) completed a social fear learning (SFL) procedure in which participants watched a “demonstrator” receiving electrical shocks in conjunction with a previously neutral conditioned stimulus (CS+), but never with a safe control stimulus (CS−). Skin conductance was used to measure autonomic responses of learned threat responses to the CS+ versus CS−. Visual attention was measured during learning using eye tracking. To establish a non-social learning baseline, each participant also underwent a test of Pavlovian conditioning. Results During learning, individuals with ASD attended less to the demonstrator’s face, and when later tested, displayed stronger observational, but not Pavlovian, autonomic indices of learning (skin conductance) compared to controls. In controls, both higher levels of attention to the demonstrator’s face and trait empathy predicted diminished expressions of learning during test. Limitations The relatively small sample size of this study and the typical IQ range of the ASD group limit the generalizability of our findings to individuals with ASD in the average intellectual ability range. Conclusions The enhanced social threat learning in individuals with ASD may be linked to difficulties using visual attention and mental state attributions to downregulate their emotion.

Background: Social anxiety disorder (SAD) has its typical onset in childhood and adolescence. Maladaptive processing of social information may contribute to the etiology and maintenance of SAD. During face perception, individuals execute a succession of visual fixations known as a scanpath which facilitates information processing. Atypically long scanpaths have been reported in adults with SAD, but no data exists from pediatric samples. SAD has also been linked to atypical arousal during face perception. Both metrics were examined in one of the largest eye-tracking studies of pediatric SAD to date.Methods: Participants were children and adolescents with SAD (n = 61) and healthy controls (n = 39) with a mean age of 14 years (range 10–17) who completed an emotion recognition task. The visual scanpath and pupil dilation (an indirect index of arousal) were examined using eye tracking.Results: Scanpaths of youth with SAD were shorter, less distributed, and consisted of a smaller number of fixations than those of healthy controls. These findings were supported by both frequentist and Bayesian statistics. Higher pupil dilation was also observed in the SAD group, but despite a statistically significant group difference, this result was not supported by the Bayesian analysis.Conclusions: The results were contrary to findings from adult studies, but similar to what has been reported in neurodevelopmental conditions associated with social interaction impairments. Restricted scanpaths may disrupt holistic representation of faces known to favor adaptive social understanding.

Lamellar metaplastic bone, osteoid metaplasia (OM), is found in atherosclerotic plaques, especially in the femoral arteries. In the carotid arteries, OM has been documented to be associated with plaque stability. This study investigated the clinical impact of OM load in femoral artery plaques of patients with lower extremity artery disease (LEAD) by using a deep learning-based image analysis algorithm. Plaques from 90 patients undergoing endarterectomy of the common femoral artery were collected and analyzed. After decalcification and fixation, 4-μm-thick longitudinal sections were stained with hematoxylin and eosin, digitized, and uploaded as whole-slide images on a cloud-based platform. A deep learning-based image analysis algorithm was trained to analyze the area percentage of OM in whole-slide images. Clinical data were extracted from electronic patient records, and the association with OM was analyzed. Fifty-one (56.7%) sections had OM. Females with diabetes had a higher area percentage of OM than females without diabetes. In male patients, the area percentage of OM inversely correlated with toe pressure and was significantly associated with severe symptoms of LEAD including rest pain, ulcer, or gangrene. According to our results, OM is a typical feature of femoral artery plaques and can be quantified using a deep learning-based image analysis method. The association of OM load with clinical features of LEAD appears to differ between male and female patients, highlighting the need for a gender-specific approach in the study of the mechanisms of atherosclerotic disease. In addition, the role of plaque characteristics in the treatment of atherosclerotic lesions warrants further consideration in the future.

BackgroundMalaria remains a major global health problem with a need for improved field-usable diagnostic tests. We have developed a portable, low-cost digital microscope scanner, capable of both brightfield and fluorescence imaging. Here, we used the instrument to digitize blood smears, and applied deep learning (DL) algorithms to detect Plasmodium falciparum parasites.MethodsThin blood smears (n = 125) were collected from patients with microscopy-confirmed P. falciparum infections in rural Tanzania, prior to and after initiation of artemisinin-based combination therapy. The samples were stained using the 4',6-diamidino-2-phenylindole fluorogen and digitized using the prototype microscope scanner. Two DL algorithms were trained to detect malaria parasites in the samples, and results compared to the visual assessment of both the digitized samples, and the Giemsa-stained thick smears.ResultsDetection of P. falciparum parasites in the digitized thin blood smears was possible both by visual assessment and by DL-based analysis with a strong correlation in results (r = 0.99, p < 0.01). A moderately strong correlation was observed between the DL-based thin smear analysis and the visual thick smear-analysis (r = 0.74, p < 0.01). Low levels of parasites were detected by DL-based analysis on day three following treatment initiation, but a small number of fluorescent signals were detected also in microscopy-negative samples.ConclusionQuantification of P. falciparum parasites in DAPI-stained thin smears is feasible using DL-supported, point-of-care digital microscopy, with a high correlation to visual assessment of samples. Fluorescent signals from artefacts in samples with low infection levels represented the main challenge for the digital analysis, thus highlighting the importance of minimizing sample contaminations. The proposed method could support malaria diagnostics and monitoring of treatment response through automated quantification of parasitaemia and is likely to be applicable also for diagnostics of other Plasmodium species and other infectious diseases.

Social anxiety disorder (SAD) is a psychiatric condition that often onsets in childhood. Cognitive models underline the role of attention in the maintenance of SAD, but studies on youth populations are few, particularly those using eye tracking to measure attention. Cognitive behavioral therapy (CBT) for SAD includes interventions targeting attention, like exposure to eye contact, but the link between CBT and attention bias is largely unexplored. This study investigated attention bias in youth with SAD and the association with outcome from CBT. Latency to attend to pictures of faces with different emotions (vigilance) and latency to disengage from social stimuli (avoidance) was examined in N = 25 adolescents (aged 13-17) with SAD in relation to treatment outcome. Vigilance was operationalized as the time it took to relocate the gaze from a central position to a peripherally appearing social stimulus. The latency to disengage from a centrally located social stimulus, when a non-social stimulus appeared in the periphery, was used as a proxy for avoidance. Attention characteristics in the SAD group were compared to non-anxious (NA) controls (N = 22). Visual attention was measured using eye tracking. Participants in both the SAD and NA groups were vigilant towards angry faces, compared to neutral and happy faces. Similarly, both groups disengaged attention faster from angry faces. Adolescents with SAD who disengaged faster from social stimuli had less social anxiety after CBT. The results indicate that anxious youth display a vigilant-avoidant attention pattern to threat. However, partly inconsistent with previous research, the same pattern was observed in the NA group.

BackgroundDetection of lymph node metastases is essential in breast cancer diagnostics and staging, affecting treatment and prognosis. Intraoperative microscopy analysis of sentinel lymph node frozen sections is standard for detection of axillary metastases but requires access to a pathologist for sample analysis. Remote analysis of digitized samples is an alternative solution but is limited by the requirement for high-end slide scanning equipment.ObjectiveTo determine whether the image quality achievable with a low-cost, miniature digital microscope scanner is sufficient for detection of metastases in breast cancer lymph node frozen sections.MethodsLymph node frozen sections from 79 breast cancer patients were digitized using a prototype miniature microscope scanner and a high-end slide scanner. Images were independently reviewed by two pathologists and results compared between devices with conventional light microscopy analysis as ground truth.ResultsDetection of metastases in the images acquired with the miniature scanner yielded an overall sensitivity of 91% and specificity of 99% and showed strong agreement when compared to light microscopy (k = 0.91). Strong agreement was also observed when results were compared to results from the high-end slide scanner (k = 0.94). A majority of discrepant cases were micrometastases and sections of which no anticytokeratin staining was available.ConclusionAccuracy of detection of metastatic cells in breast cancer sentinel lymph node frozen sections by visual analysis of samples digitized using low-cost, point-of-care microscopy is comparable to analysis of digital samples scanned using a high-end, whole slide scanner. This technique could potentially provide a workflow for digital diagnostics in resource-limited settings, facilitate sample analysis at the point-of-care and reduce the need for trained experts on-site during surgical procedures.

The subtype of the papillary thyroid carcinoma tall-cell variant has a worse prognosis than does the conventional papillary type (papillary thyroid carcinoma). The new World Health Organization 2017 classification defines a tall-cell variant as a tumour consisting of over 30% of cells that are two or three times as tall as they are wide. However, thresholds have differed. Our aim was to study how tall cells affect the prognosis of papillary thyroid carcinoma patients and to determine, for such cells, a cut-off percentage. Our cohort included 65 papillary thyroid carcinoma patients who underwent surgery at Helsinki University Hospital between 1973 and 1996: originally, 36 otherwise-matched patient pairs, eventually comprising 34 patients with an adverse outcome plus 31 who had recovered. All samples were digitally scanned and scored by two investigators based on tall cell composition. The cohort was analysed with four tall cell thresholds: 10%, 30%, 50% and 70% with a median follow-up of 22 years. In survival analysis, only the 70% threshold showed a correlation with reduced overall survival, disease-specific survival and relapse-free survival. A correlation also emerged with death from papillary thyroid carcinoma. In multivariate analysis, a 70% cut-off and age at diagnosis significantly affected DSS. Increasing tall cell score correlated with increasing age and extrathyroidal extensions. A tall cell composition of 10%, 30% or 50% showed no correlation with adverse outcome and suggests that the choice of pathologists reporting tall-cell variant should be a 70% threshold.

A review of flywheel energy storage technology was made, with a special focus on the progress in automotive applications. We found that there are at least 26 university research groups and 27 companies contributing to flywheel technology development. Flywheels are seen to excel in high-power applications, placing them closer in functionality to supercapacitors than to batteries. Examples of flywheels optimized for vehicular applications were found with a specific power of 5.5 kW/kg and a specific energy of 3.5 Wh/kg. Another flywheel system had 3.15 kW/kg and 6.4 Wh/kg, which can be compared to a state-of-the-art supercapacitor vehicular system with 1.7 kW/kg and 2.3 Wh/kg, respectively. Flywheel energy storage is reaching maturity, with 500 flywheel power buffer systems being deployed for London buses (resulting in fuel savings of over 20%), 400 flywheels in operation for grid frequency regulation and many hundreds more installed for uninterruptible power supply (UPS) applications. The industry estimates the mass-production cost of a specific consumer-car flywheel system to be 2000 USD. For regular cars, this system has been shown to save 35% fuel in the U.S. Federal Test Procedure (FTP) drive cycle.

Background: Diagnostic support for clinicians is a domain of application of mHealth technologies with a slow uptake despite promising opportunities, such as image-based clinical support. The absence of a roadmap for the adoption and implementation of these types of applications is a further obstacle. Objectives: This article provides the groundwork for a roadmap to implement image-based support for clinicians, focusing on how to overcome potential barriers affecting front-line users, the health-care organization and the technical system. Methods: A consensual approach was used during a two-day roundtable meeting gathering a convenience sample of stakeholders (n = 50) from clinical, research, policymaking and business fields and from different countries. A series of sessions was held including small group discussions followed by reports to the plenary. Session moderators synthesized the reports in a number of theme-specific strategies that were presented to the participants again at the end of the meeting for them to determine their individual priority. Results: There were four to seven strategies derived from the thematic sessions. Once reviewed and prioritized by the participants some received greater priorities than others. As an example, of the seven strategies related to the front-line users, three received greater priority: the need for any system to significantly add value to the users; the usability of mHealth apps; and the goodness-of-fit into the work flow. Further, three aspects cut across the themes: ease of integration of the mHealth applications; solid ICT infrastructure and support network; and interoperability. Conclusions: Research and development in image-based diagnostic pave the way to making health care more accessible and more equitable. The successful implementation of those solutions will necessitate a seamless introduction into routines, adequate technical support and significant added value.

Background: Microscopy remains the gold standard in the diagnosis of neglected tropical diseases. As resource limited, rural areas often lack laboratory equipment and trained personnel, new diagnostic techniques are needed. Low-cost, point-of-care imaging devices show potential in the diagnosis of these diseases. Novel, digital image analysis algorithms can be utilized to automate sample analysis. Objective: Evaluation of the imaging performance of a miniature digital microscopy scanner for the diagnosis of soil-transmitted helminths and Schistosoma haematobium, and training of a deep learning-based image analysis algorithm for automated detection of soil-transmitted helminths in the captured images. Methods: A total of 13 iodine-stained stool samples containing Ascaris lumbricoides, Trichuris trichiura and hookworm eggs and 4 urine samples containing Schistosoma haematobium were digitized using a reference whole slide-scanner and the mobile microscopy scanner. Parasites in the images were identified by visual examination and by analysis with a deep learning-based image analysis algorithm in the stool samples. Results were compared between the digital and visual analysis of the images showing helminth eggs. Results: Parasite identification by visual analysis of digital slides captured with the mobile microscope was feasible for all analyzed parasites. Although the spatial resolution of the reference slide-scanner is higher, the resolution of the mobile microscope is sufficient for reliable identification and classification of all parasites studied. Digital image analysis of stool sample images captured with the mobile microscope showed high sensitivity for detection of all helminths studied (range of sensitivity = 83.3–100%) in the test set (n = 217) of manually labeled helminth eggs. Conclusions: In this proof-of-concept study, the imaging performance of a mobile, digital microscope was sufficient for visual detection of soil-transmitted helminths and Schistosoma haematobium. Furthermore, we show that deep learning-based image analysis can be utilized for the automated detection and classification of helminths in the captured images.

Current advances within medical technology show great potential from a global health perspective. Inexpensive, effective solutions to common problems within diagnostics, medical procedures and access to medical information are emerging within almost all fields of medicine. The innovations can benefit health care both in resource-limited and in resource-rich settings. However, there is a big gap between the proof-of-concept stage and implementation. This article will give examples of promising solutions, with special focus on mobile image- and sensor-based diagnostics. We also discuss how technology and frugal innovations could be made sustainable and widely available. Finally, a list of critical factors for success is presented, based on both our own experiences and the literature.

Williams syndrome (WS) is a rare genetic condition associated with high sociability, intellectual disability, and social cognitive challenges. Attention to others' eyes is crucial for social understanding. Orienting to, and from other's eyes was studied in WS (n = 37, mean age = 23, age range 9-53). The WS group was compared to a typically developing comparison participants (n = 167) in stratified age groups from infancy to adulthood. Typically developing children and adults were quicker and more likely to orient to eyes than the mouth. This bias was absent in WS. The WS group had reduced peak saccadic velocities, indicating hypo-arousal. The current study indicates reduced orienting to others' eyes in WS, which may affect social interaction skills.

BACKGROUND:Each year more than 10 million people worldwide are burned severely enough to require medical attention, with clinical outcomes noticeably worse in resource poor settings. Expert clinical advice on acute injuries can play a determinant role and there is a need for novel approaches that allow for timely access to advice. We developed an interactive mobile phone application that enables transfer of both patient data and pictures of a wound from the point-of-care to a remote burns expert who, in turn, provides advice back. METHODS AND RESULTS:The application is an integrated clinical decision support system that includes a mobile phone application and server software running in a cloud environment. The client application is installed on a smartphone and structured patient data and photographs can be captured in a protocol driven manner. The user can indicate the specific injured body surface(s) through a touchscreen interface and an integrated calculator estimates the total body surface area that the burn injury affects. Predefined standardised care advice including total fluid requirement is provided immediately by the software and the case data are relayed to a cloud server. A text message is automatically sent to a burn expert on call who then can access the cloud server with the smartphone app or a web browser, review the case and pictures, and respond with both structured and personalized advice to the health care professional at the point-of-care. CONCLUSIONS:In this article, we present the design of the smartphone and the server application alongside the type of structured patient data collected together with the pictures taken at point-of-care. We report on how the application will be introduced at point-of-care and how its clinical impact will be evaluated prior to roll out. Challenges, strengths and limitations of the system are identified that may help materialising or hinder the expected outcome to provide a solution for remote consultation on burns that can be integrated into routine acute clinical care and thereby promote equity in injury emergency care, a growing public health burden.

This paper presents experimental results on an innovative electric driveline employing a kinetic energy storage device as energy buffer. A conceptual division of losses in the system was created, separating the complete system into three parts according to their function. This conceptualization of the system yielded a meaningful definition of the concept of efficiency. Additionally, a thorough theoretical framework for the prediction of losses associated with energy storage and transfer in the system was developed. A large number of spin-down tests at varying pressure levels were performed. A separation of the measured data into the different physical processes responsible for power loss was achieved from the corresponding dependence on rotational velocity. This comparison yielded an estimate of the perpendicular resistivity of the stranded copper conductor of 2.5 × 10−8 ± 3.5 × 10−9. Further, power and energy were measured system-wide during operation, and an analysis of the losses was performed. The analytical solution was able to reproduce the measured distribution of losses in the system to an accuracy of 4.7% (95% CI). It was found that the losses attributed to the function of kinetic energy storage in the system amounted to between 45% and 65%, depending on usage.

INTRODUCTION:A significant barrier to medical diagnostics in low-resource environments is the lack of medical care and equipment. Here we present a low-cost, cloud-connected digital microscope for applications at the point-of-care. We evaluate the performance of the device in the digital assessment of estrogen receptor-alpha (ER) expression in breast cancer samples. Studies suggest computer-assisted analysis of tumor samples digitized with whole slide-scanners may be comparable to manual scoring, here we study whether similar results can be obtained with the device presented. MATERIALS AND METHODS:A total of 170 samples of human breast carcinoma, immunostained for ER expression, were digitized with a high-end slide-scanner and the point-of-care microscope. Corresponding regions from the samples were extracted, and ER status was determined visually and digitally. Samples were classified as ER negative (

IntroductionMicroscopy is the gold standard for diagnosis of malaria, however, manual evaluation of blood films is highly dependent on skilled personnel in a time-consuming, error-prone and repetitive process. In this study we propose a method using computer vision detection and visualization of only the diagnostically most relevant sample regions in digitized blood smears.MethodsGiemsa-stained thin blood films with P. falciparum ring-stage trophozoites (n = 27) and uninfected controls (n = 20) were digitally scanned with an oil immersion objective (0.1 µm/pixel) to capture approximately 50,000 erythrocytes per sample. Parasite candidate regions were identified based on color and object size, followed by extraction of image features (local binary patterns, local contrast and Scale-invariant feature transform descriptors) used as input to a support vector machine classifier. The classifier was trained on digital slides from ten patients and validated on six samples.ResultsThe diagnostic accuracy was tested on 31 samples (19 infected and 12 controls). From each digitized area of a blood smear, a panel with the 128 most probable parasite candidate regions was generated. Two expert microscopists were asked to visually inspect the panel on a tablet computer and to judge whether the patient was infected with P. falciparum. The method achieved a diagnostic sensitivity and specificity of 95% and 100% as well as 90% and 100% for the two readers respectively using the diagnostic tool. Parasitemia was separately calculated by the automated system and the correlation coefficient between manual and automated parasitemia counts was 0.97.ConclusionWe developed a decision support system for detecting malaria parasites using a computer vision algorithm combined with visualization of sample areas with the highest probability of malaria infection. The system provides a novel method for blood smear screening with a significantly reduced need for visual examination and has a potential to increase the throughput in malaria diagnostics.

Studies have supported two different hypotheses of reduced eye gaze in people with ASD; gaze "avoidance" and gaze "indifference," while less is known about the role of anxiety. We tested these hypotheses using an eye-tracking paradigm that cued the eyes or mouth of emotional faces. Autistic children (n = 12, mean age 7 years) looked faster away from both eyes and mouths than controls (n = 22). This effect was not explained by anxiety symptoms. No difference was found in latency towards either area. These results indicate that attentional avoidance of autistic children is not specific to eyes, and that they do not show attentional indifference to eyes compared to controls. Atypicalities in visual scanning in ASD are possibly unrelated to specific facial areas.

BACKGROUND: Microscopy, being relatively easy to perform at low cost, is the universal diagnostic method for detection of most globally important parasitic infections. As quality control is hard to maintain, misdiagnosis is common, which affects both estimates of parasite burdens and patient care. Novel techniques for high-resolution imaging and image transfer over data networks may offer solutions to these problems through provision of education, quality assurance and diagnostics. Imaging can be done directly on image sensor chips, a technique possible to exploit commercially for the development of inexpensive "mini-microscopes". Images can be transferred for analysis both visually and by computer vision both at point-of-care and at remote locations. METHODS/PRINCIPAL FINDINGS: Here we describe imaging of helminth eggs using mini-microscopes constructed from webcams and mobile phone cameras. The results show that an inexpensive webcam, stripped off its optics to allow direct application of the test sample on the exposed surface of the sensor, yields images of Schistosoma haematobium eggs, which can be identified visually. Using a highly specific image pattern recognition algorithm, 4 out of 5 eggs observed visually could be identified. CONCLUSIONS/SIGNIFICANCE: As proof of concept we show that an inexpensive imaging device, such as a webcam, may be easily modified into a microscope, for the detection of helminth eggs based on on-chip imaging. Furthermore, algorithms for helminth egg detection by machine vision can be generated for automated diagnostics. The results can be exploited for constructing simple imaging devices for low-cost diagnostics of urogenital schistosomiasis and other neglected tropical infectious diseases.

Background: Immune cell infiltration in tumor is an emerging prognostic biomarker in breast cancer. The gold standard for quantification of immune cells in tissue sections is visual assessment through a microscope, which is subjective and semi-quantitative. In this study, we propose and evaluate an approach based on antibody-guided annotation and deep learning to quantify immune cell-rich areas in hematoxylin and eosin (H&E) stained samples. Methods: Consecutive sections of formalin-fixed parafin-embedded samples obtained from the primary tumor of twenty breast cancer patients were cut and stained with H&E and the pan-leukocyte CD45 antibody. The stained slides were digitally scanned, and a training set of immune cell-rich and cell-poor tissue regions was annotated in H&E whole-slide images using the CD45-expression as a guide. In analysis, the images were divided into small homogenous regions, superpixels, from which features were extracted using a pretrained convolutional neural network (CNN) and classified with a support of vector machine. The CNN approach was compared to texture-based classification and to visual assessments performed by two pathologists. Results: In a set of 123,442 labeled superpixels, the CNN approach achieved an F-score of 0.94 (range: 0.92-0.94) in discrimination of immune cell-rich and cell-poor regions, as compared to an F-score of 0.88 (range: 0.87-0.89) obtained with the texture-based classification. When compared to visual assessment of 200 images, an agreement of 90% (k = 0.79) to quantify immune infiltration with the CNN approach was achieved while the inter-observer agreement between pathologists was 90% (k = 0.78). Conclusions: Our findings indicate that deep learning can be applied to quantify immune cell infiltration in breast cancer samples using a basic morphology staining only. A good discrimination of immune cell-rich areas was achieved, well in concordance with both leukocyte antigen expression and pathologists′ visual assessment.

BACKGROUND: The basis for correctly assessing the burden of parasitic infections and the effects of interventions relies on a somewhat shaky foundation as long as we do not know how reliable the reported laboratory findings are. Thus virtual microscopy, successfully introduced as a histopathology tool, has been adapted for medical parasitology. METHODOLOGY/PRINCIPAL FINDINGS: Specimens containing parasites in tissues, stools, and blood have been digitized and made accessible as a "webmicroscope for parasitology" (WMP) on the Internet (http://www.webmicroscope.net/parasitology).These digitized specimens can be viewed ("navigated" both in the x-axis and the y-axis) at the desired magnification by an unrestricted number of individuals simultaneously. For virtual microscopy of specimens containing stool parasites, it was necessary to develop the technique further in order to enable navigation in the z plane (i.e., "focusing"). Specimens were therefore scanned and photographed in two or more focal planes. The resulting digitized specimens consist of stacks of laterally "stiched" individual images covering the entire area of the sample photographed at high magnification. The digitized image information (approximately 10 GB uncompressed data per specimen) is accessible at data transfer speeds from 2 to 10 Mb/s via a network of five image servers located in different parts of Europe. Image streaming and rapid data transfer to an ordinary personal computer makes web-based virtual microscopy similar to conventional microscopy. CONCLUSION/SIGNIFICANCE: The potential of this novel technique in the field of medical parasitology to share identical parasitological specimens means that we can provide a "gold standard", which can overcome several problems encountered in quality control of diagnostic parasitology. Thus, the WMP may have an impact on the reliability of data, which constitute the basis for our understanding of the vast problem of neglected tropical diseases. The WMP can be used also in the absence of a fast Internet communication. An ordinary PC, or even a laptop, may function as a local image server, e.g., in health centers in tropical endemic areas.