Next generation mycological diagnosis: Artificial intelligence-based classifier of the presence of Malassezia yeasts in tape strip samples.

BACKGROUND: Malassezia yeasts are almost universally present on human skin worldwide. While they can cause diseases such as pityriasis versicolor, their implication in skin homeostasis and pathophysiology of other dermatoses is still unclear. Their analysis using native microscopy of skin tape strips is operator dependent and requires skill, training and significant amounts of hands-on time. OBJECTIVES AND METHODS: To standardise and improve the speed and quality of diagnosis of Malassezia in skin tape strip samples, we sought to create an artificial intelligence-based algorithm for this image classification task. Three algorithms, each using different internal architectures, were trained and validated on a manually annotated dataset of 1113 images from 22 samples. RESULTS: The Vision Transformer-based algorithm performed the best with a validation accuracy of 94%, sensitivity of 94.0% and specificity of 93.5%. Visualisations providing insight into the reasoning of the algorithm were presented and discussed. CONCLUSION: Our image classifier achieved very good performance in the diagnosis of the presence of Malassezia yeasts in tape strip samples of human skin and can therefore improve the speed and quality of, and access to this diagnostic test. By expanding data sources and explainability, the algorithm could also provide teaching points for more novice operators in future.

Burden of atopic dermatitis in Europe: A population-centred approach leveraging web search data in 21 European countries.

BACKGROUND: The complexity, high prevalence, and substantial personal and socioeconomic burden collectively render atopic dermatitis (AD) a major public health concern. Using crowdsourced Internet data has the potential to provide unique insights into this concern, as demonstrated by several previous studies. However, a comprehensive comparison across European countries remains lacking. OBJECTIVES: The study aimed to investigate AD-related web searches across Europe to assess spatiotemporal variations and associations between disease-related and external factors. METHODS: AD-related web search data were extracted for 21 European countries between February 2019 and January 2023. Descriptive analysis and autocorrelation functions were performed to examine spatiotemporal patterns. Correlations (r) were used to evaluate the associations between web searches and disease-related, socioeconomic and meteorological data. RESULTS: Over 241 million AD-related web searches were identified, with search volume varying substantially among European countries (p < 0.001) and correlating with AD prevalence and disease burden (both r = 0.51, p = 0.019). Search volume increased between 2019 and 2023 in all countries and seasonally peaked in January and March. Negative correlations with median population age (r = -0.46, p = 0.039), number of general practitioners (r = -0.29, p = 0.226) and specialists (r = -0.27, p = 0.270) were observed. Moderate to strong correlations were found between search volume and cold, humid and windy weather with fewer sunshine hours, while higher online interest typically occurred 1-3 months after such weather conditions. CONCLUSION: The study highlights the great potential of online crowdsourced data analysis, for example, to investigate the impact of climate change or to identify unmet needs at a population level. Furthermore, the growing online interest in AD and the corresponding seasonal peaks emphasize the necessity of adapting treatment plans, intensifying public health campaigns, and disseminating reliable online information by governments and healthcare providers, especially during these periods.

Automatic body part identification in real-world clinical dermatological images using machine learning.

BACKGROUND: Dermatological conditions are prevalent across all population sub-groups. The affected body part is of importance to their diagnosis, therapy, and research. The automatic identification of body parts in dermatological clinical pictures could therefore improve clinical care by providing additional information for clinical decision-making algorithms, discovering hard-to-treat areas, and research by identifying new patterns of disease. PATIENTS AND METHODS: In this study, we used 6,219 labelled dermatological images from our clinical database, which were used to train and validate a convolutional neural network. As a use case, qualitative heatmaps for the body part distribution in common dermatological conditions was generated using this system. RESULTS: The algorithm reached a mean balanced accuracy of 89% (range 74.8%-96.5%). Non-melanoma skin cancer photos were mostly of the face and torso, while hotspots of eczema and psoriasis image distribution included the torso, legs, and hands. CONCLUSIONS: The accuracy of this system is comparable to the best to-date published algorithms for image classification challenges, suggesting this algorithm could boost diagnosis, therapy, and research of dermatological conditions.

Therapeutic inhibition of CXCR1/2: where do we stand?

Mounting experimental evidence from in vitro and in vivo animal studies points to an essential role of the CXCL8-CXCR1/2 axis in neutrophils in the pathophysiology of inflammatory and autoimmune diseases. In addition, the pathogenetic involvement of neutrophils and the CXCL8-CXCR1/2 axis in cancer progression and metastasis is increasingly recognized. Consequently, therapeutic targeting of CXCR1/2 or CXCL8 has been intensively investigated in recent years using a wide array of in vitro and animal disease models. While a significant benefit for patients with unwanted neutrophil-mediated inflammatory conditions may be expected from a potential clinical use of inhibitors, their use in severe infections or sepsis might be problematic and should be carefully and thoroughly evaluated in animal models and clinical trials. Translating the approaches using inhibitors of the CXCL8-CXCR1/2 axis to cancer therapy is definitively a new and promising research avenue, which parallels the ongoing efforts to clearly define the involvement of neutrophils and the CXCL8-CXCR1/2 axis in neoplastic diseases. Our narrative review summarizes the current literature on the activation and inhibition of these receptors in neutrophils, key inhibitor classes for CXCR2 and the therapeutic relevance of CXCR2 inhibition focusing here on gastrointestinal diseases.

Assessing the national burden of allergic asthma by web-search data, pollen counts, and drug prescriptions in Germany and Sweden.

Background: Asthma and its main phenotype allergic asthma are prevalent, chronic, and complex diseases affecting 4% of the population. One main trigger for allergic asthma exacerbations is pollen. Online health information search behavior by people is increasing, and analysis of web-search data can provide valuable insight into disease burden and risk factors of a population. Objectives: We sought to perform a web-search data analysis and correlation to climate factors and pollen in 2 European countries. Methods: We analyzed the national web-search volume for allergic asthma-related keywords in Germany and Sweden from 2018 to 2021 and correlated it to local pollen counts, climatic factors, and drug prescription rates. Results: Per capita, more searches were conducted in Sweden than in Germany. A complex geographic stratification within the countries was observed. Search results were seasonal with a peak in spring and correlated with pollen counts in both countries. However, anti-asthmatic drug prescription rates in Sweden, as well as temperature and precipitation in both countries, did not correlate with search volume. Conclusion: Our analysis offers population-level insights about this complex disease by reporting its needs and establishing the correlation to pollen counts, which enables a targeted approach in the public health management of allergic asthma. Local pollen counts, as opposed to temperature or precipitation, might be good predictors of allergic asthma disease burden.

Outlier detection in dermatology: Performance of different convolutional neural networks for binary classification of inflammatory skin diseases.

BACKGROUND: Artificial intelligence (AI) and convolutional neural networks (CNNs) represent rising trends in modern medicine. However, comprehensive data on the performance of AI practices in clinical dermatologic images are non-existent. Furthermore, the role of professional data selection for training remains unknown. OBJECTIVES: The aims of this study were to develop AI applications for outlier detection of dermatological pathologies, to evaluate CNN architectures' performance on dermatological images and to investigate the role of professional pre-processing of the training data, serving as one of the first anchor points regarding data selection criteria in dermatological AI-based binary classification tasks of non-melanoma pathologies. METHODS: Six state-of-the-art CNN architectures were evaluated for their accuracy, sensitivity and specificity for five dermatological diseases and using five data subsets, including data selected by two dermatologists, one with 5 and the other with 11 years of clinical experience. RESULTS: Overall, 150 CNNs were evaluated on up to 4051 clinical images. The best accuracy was reached for onychomycosis (accuracy = 1.000), followed by bullous pemphigoid (accuracy = 0.951) and lupus erythematosus (accuracy = 0.912). The CNNs InceptionV3, Xception and ResNet50 achieved the best accuracy in 9, 8 and 6 out of 25 data sets, respectively (36.0%, 32.0% and 24.0%). On average, the data set provided by the senior physician and the data set provided in accordance with both dermatologists performed the best (accuracy = 0.910). CONCLUSIONS: This AI approach for the detection of outliers in dermatological diagnoses represents one of the first studies to evaluate the performance of different CNNs for binary decisions in clinical non-dermatoscopic images of a variety of dermatological diseases other than melanoma. The selection of images by an experienced dermatologist during pre-processing had substantial benefits for the performance of the CNNs. These comparative results might guide future AI approaches to dermatology diagnostics, and the evaluated CNNs might be applicable for the future training of dermatology residents.

Munich atopy prediction study (MAPS): protocol for a prospective birth cohort addressing clinical and molecular risk factors for atopic dermatitis in early childhood.

INTRODUCTION: The pathogenesis of atopic diseases is highly complex, and the exact mechanisms leading to atopic dermatitis (AD) onset in infants remain mostly enigmatic. In addition to an interdependent network of components of skin development in young age and skin barrier dysfunction underlying AD development that is only partially understood, a complex interplay between environmental factors and lifestyle habits with skin barrier and immune dysregulation is suspected to contribute to AD onset. This study aims to comprehensively evaluate individual microbiome and immune responses in the context of environmental determinants related the risk of developing AD in the first 4 years of a child's life. METHODS AND ANALYSES: The 'Munich Atopic Prediction Study' is a comprehensive clinical and biological investigation of a prospective birth cohort from Munich, Germany. Information on pregnancy, child development, environmental factors, parental exposures to potential allergens and acute or chronic diseases of children and parents are collected by questionnaires together with a meticulous clinical examination by trained dermatologists focusing on allergies, skin health, and in particular signs of AD at 2 months after birth and then every 6 months. In addition, skin barrier functions are assessed through cutometry, corneometry and transepidermal water loss at every visit. These measurements are completed with allergy diagnostics and extensive microbiome analyses from stool and skin swabs as well as transcriptome analyses using skin microbiopsies.The aim is to assess the relevance of different known and yet unknown risk factors of AD onset and exacerbations in infants and to identify possible accessible and robust biomarkers. ETHICS AND DISSEMINATION: The study is approved by the Ethical Committee of the Medical Faculty of the Technical University of Munich (reference 334/16S). All relevant study results will be presented at national and international conferences and in peer-reviewed journals.

Granulocyte-dependent autoantibody-induced skin blistering.

Autoimmune phenomena occur in healthy individuals, but when self-tolerance fails, the autoimmune response may result in specific pathology. According to Witebsky's postulates, one of the criteria in diagnosing a disease as autoimmune is the reproduction of the disease in experimental animals by the passive transfer of autoantibodies. For epidermolysis bullosa acquisita (EBA), a prototypic organ-specific autoimmune disease of skin and mucous membranes, several experimental models were recently established. In the animal model described in our present work, purified IgG antibodies against a stretch of 200 amino acids (aa 757-967) of collagen VII are injected repeatedly into mice reproducing the blistering phenotype as well as the histo- and immunopathological features characteristic to human EBA (1). Full-blown widespread disease is usually seen 5-6 days after the first injection and the extent of the disease correlates with the dose of the administered collagen VII-specific IgG. The tissue damage (blister formation) in the experimental EBA is depending on the recruitment and activation of granulocytes by tissue-bound autoantibodies (2,-4). Therefore, this model allows for the dissection of the granulocyte-dependent inflammatory pathway involved in the autoimmune tissue damage, as the model reproduces only the T cell-independent phase of the efferent autoimmune response. Furthermore, its value is underlined by a number of studies demonstrating the blister-inducing potential of autoantibodies in vivo and investigating the mechanism of the blister formation in EBA (1,3,-6). Finally, this model will greatly facilitate the development of new anti-inflammatory therapies in autoantibody-induced diseases. Overall, the passive transfer animal model of EBA is an accessible and instructive disease model and will help researchers to analyze not only EBA pathogenesis but to answer fundamental biologically and clinically essential autoimmunity questions.