Annotation protocol and crowdsourcing multiple instance learning classification of skin histological images: The CR-AI4SkIN dataset.

Digital Pathology (DP) has experienced a significant growth in recent years and has become an essential tool for diagnosing and prognosis of tumors. The availability of Whole Slide Images (WSIs) and the implementation of Deep Learning (DL) algorithms have paved the way for the appearance of Artificial Intelligence (AI) systems that support the diagnosis process. These systems require extensive and varied data for their training to be successful. However, creating labeled datasets in histopathology is laborious and time-consuming. We have developed a crowdsourcing-multiple instance labeling/learning protocol that is applied to the creation and use of the CR-AI4SkIN dataset.2 CR-AI4SkIN contains 271 WSIs of 7 Cutaneous Spindle Cell (CSC) neoplasms with expert and non-expert labels at region and WSI levels. It is the first dataset of these types of neoplasms made available. The regions selected by the experts are used to learn an automatic extractor of Regions of Interest (ROIs) from WSIs. To produce the embedding of each WSI, the representations of patches within the ROIs are obtained using a contrastive learning method, and then combined. Finally, they are fed to a Gaussian process-based crowdsourcing classifier, which utilizes the noisy non-expert WSI labels. We validate our crowdsourcing-multiple instance learning method in the CR-AI4SkIN dataset, addressing a binary classification problem (malign vs. benign). The proposed method obtains an F1 score of 0.7911 on the test set, outperforming three widely used aggregation methods for crowdsourcing tasks. Furthermore, our crowdsourcing method also outperforms the supervised model with expert labels on the test set (F1-score = 0.6035). The promising results support the proposed crowdsourcing multiple instance learning annotation protocol. It also validates the automatic extraction of interest regions and the use of contrastive embedding and Gaussian process classification to perform crowdsourcing classification tasks.

Computational pathology in 2030: a Delphi study forecasting the role of AI in pathology within the next decade.

BACKGROUND: Artificial intelligence (AI) is rapidly fuelling a fundamental transformation in the practice of pathology. However, clinical integration remains challenging, with no AI algorithms to date in routine adoption within typical anatomic pathology (AP) laboratories. This survey gathered current expert perspectives and expectations regarding the role of AI in AP from those with first-hand computational pathology and AI experience. METHODS: Perspectives were solicited using the Delphi method from 24 subject matter experts between December 2020 and February 2021 regarding the anticipated role of AI in pathology by the year 2030. The study consisted of three consecutive rounds: 1) an open-ended, free response questionnaire generating a list of survey items; 2) a Likert-scale survey scored by experts and analysed for consensus; and 3) a repeat survey of items not reaching consensus to obtain further expert consensus. FINDINGS: Consensus opinions were reached on 141 of 180 survey items (78.3%). Experts agreed that AI would be routinely and impactfully used within AP laboratory and pathologist clinical workflows by 2030. High consensus was reached on 100 items across nine categories encompassing the impact of AI on (1) pathology key performance indicators (KPIs) and (2) the pathology workforce and specific tasks performed by (3) pathologists and (4) AP lab technicians, as well as (5) specific AI applications and their likelihood of routine use by 2030, (6) AI's role in integrated diagnostics, (7) pathology tasks likely to be fully automated using AI, and (8) regulatory/legal and (9) ethical aspects of AI integration in pathology. INTERPRETATION: This systematic consensus study details the expected short-to-mid-term impact of AI on pathology practice. These findings provide timely and relevant information regarding future care delivery in pathology and raise key practical, ethical, and legal challenges that must be addressed prior to AI's successful clinical implementation. FUNDING: No specific funding was provided for this study.

Natural Language Processing in Pathology: Current Trends and Future Insights.

Natural language processing (NLP) plays a key role in advancing health care, being key to extracting structured information from electronic health reports. In the last decade, several advances in the field of pathology have been derived from the application of NLP to pathology reports. Herein, a comprehensive review of the most used NLP methods for extracting, coding, and organizing information from pathology reports is presented, including how the development of tools is used to improve workflow. In addition, this article discusses, from a practical point of view, the steps necessary to extract data and encode natural language information for its analytical processing, ranging from preprocessing of text to its inclusion in complex algorithms. Finally, the potential of NLP-based automatic solutions for improving workflow in pathology and their further applications in the near future is highlighted.

Clinical and pathological features of Merkel cell carcinoma: A 4-year follow-up observational retrospective study in Spain.

BACKGROUND: Merkel cell carcinoma (MCC) is a malignant skin cancer with a 5-year survival rate of approximately 50%. Knowledge of MCC has increased in recent years mostly due to improved diagnosis techniques. In Spain there is lack of information regarding the incidence and tumour characteristics, and the treatment approaches are not standardised. The objective of this study was to provide information of the clinical and epidemiological characteristics of MCC patients in Spain. METHODS: Retrospective, observational study involving 192 patients from 25 Spanish hospitals. Evaluated variables included overall survival and incidence rate of Merkel cell polyomavirus, in patients diagnosed from 2012 to 2016. RESULTS: The Spanish incidence rate was estimated 0.32/100,000 inhabitants/year, with variations according to geographical regions, being slightly higher in areas with greater sunlight exposure. In total, 61.5% of tumours showed expansive growth (progressive growth of the tumour), 78.6% showed localisation in UV-exposed skin. 97.4% of patients were diagnosed by excisional biopsy. Surgery was the first line treatment in 96.6% of patients, radiotherapy in 24.6%, and chemotherapy in 6.3%. These treatments were not mutually exclusive. Median overall survival was 38.3 months (78.4% at 12 months and 60% at 24 months). MCPyV was present in 33.8% of patients. CONCLUSION: The incidence of MCC in Spain is one of the highest in Europe, with a slight predominance in men. The sample has shown that a biopsy is available for diagnosis in most cases. Moreover, the treatment is surgical when the tumour is localized and is associated with lymphadenectomy, and/or it is radiotherapy if widespread.

Blind color deconvolution, normalization, and classification of histological images using general super Gaussian priors and Bayesian inference.

BACKGROUND AND OBJECTIVE: Color variations in digital histopathology severely impact the performance of computer-aided diagnosis systems. They are due to differences in the staining process and acquisition system, among other reasons. Blind color deconvolution techniques separate multi-stained images into single stained bands which, once normalized, can be used to eliminate these negative color variations and improve the performance of machine learning tasks. METHODS: In this work, we decompose the observed RGB image in its hematoxylin and eosin components. We apply Bayesian modeling and inference based on the use of Super Gaussian sparse priors for each stain together with prior closeness to a given reference color-vector matrix. The hematoxylin and eosin components are then used for image normalization and classification of histological images. The proposed framework is tested on stain separation, image normalization, and cancer classification problems. The results are measured using the peak signal to noise ratio, normalized median intensity and the area under ROC curve on five different databases. RESULTS: The obtained results show the superiority of our approach to current state-of-the-art blind color deconvolution techniques. In particular, the fidelity to the tissue improves 1,27 dB in mean PSNR. The normalized median intensity shows a good normalization quality of the proposed approach on the tested datasets. Finally, in cancer classification experiments the area under the ROC curve improves from 0.9491 to 0.9656 and from 0.9279 to 0.9541 on Camelyon-16 and Camelyon-17, respectively, when the original and processed images are used. Furthermore, these figures of merits are better than those obtained by the methods compared with. CONCLUSIONS: The proposed framework for blind color deconvolution, normalization and classification of images guarantees fidelity to the tissue structure and can be used both for normalization and classification. In addition, color deconvolution enables the use of the optical density space for classification, which improves the classification performance.

Role of artificial intelligence in multidisciplinary imaging diagnosis of gastrointestinal diseases.

The use of artificial intelligence-based tools is regarded as a promising approach to increase clinical efficiency in diagnostic imaging, improve the interpretability of results, and support decision-making for the detection and prevention of diseases. Radiology, endoscopy and pathology images are suitable for deep-learning analysis, potentially changing the way care is delivered in gastroenterology. The aim of this review is to examine the key aspects of different neural network architectures used for the evaluation of gastrointestinal conditions, by discussing how different models behave in critical tasks, such as lesion detection or characterization (i.e. the distinction between benign and malignant lesions of the esophagus, the stomach and the colon). To this end, we provide an overview on recent achievements and future prospects in deep learning methods applied to the analysis of radiology, endoscopy and histologic whole-slide images of the gastrointestinal tract.

Histological Changes Related to Symptomatic Improvement of Spontaneous Keloids Treated with a Low-Dosage Regimen of UVA-1 Phototherapy.

Keloids are a difficult-to-treat disease characterized by an imbalance in mechanisms of tissue reparation. We present the case of a middle-aged woman with spontaneous keloids which histologically and clinically improved after UVA-1 phototherapy treatment. There are few reported cases of keloids treated with high doses of UVA-1 phototherapy. We used a low-dosage regimen with a good response in only one cycle, which could diminish the risk of skin cancer development.

Leishmaniasis due to Leishmania infantum: Integration of human, animal and environmental data through a One Health approach.

The aim of this study was to explore Leishmania infantum epidemiology through a One Health approach that promotes a better estimation of leishmaniasis burden and a deeper understanding of the spatial distribution of the key actors of the parasite life cycle (vectors, reservoirs and humans). We conducted a 14-year mixed retrospective and prospective study of leishmaniasis cases in an endemic area in southern Spain (Granada province), to estimate the human incidence and its association with the vector presence, cryptic leishmaniasis rates and canine leishmaniasis prevalence. We found an annual linear increase in the incidence that cannot be fully explained by active case surveillance and the improvement of PCR diagnostic techniques. 49.4% of cases were not reported to the surveillance system. Approximately half of the human cases correspond to the visceral form that occurred more frequently in men; cutaneous, mucosal and cryptic forms were also detected. Leishmaniasis is no longer a disease of young children, accounting for a quarter of immunocompetent patients and most infected people remained asymptomatic. Human and canine leishmaniasis, cryptic or symptomatic, are present in the whole province, where there is a medium/high risk of the presence of Phlebotomus perniciosus, the main vector. We found association between the incidence of human leishmaniasis and the presence of the vector, but not with the prevalence of canine leishmaniasis and cryptic human leishmaniasis. A potential hot spot was also found, where high leishmaniasis incidence may be associated to the involvement of host species other than dogs.

Complete Digital Pathology for Routine Histopathology Diagnosis in a Multicenter Hospital Network.

CONTEXT.­: Complete digital pathology and whole slide imaging for routine histopathology diagnosis is currently in use in few laboratories worldwide. Granada University Hospitals, Spain, which comprises 4 hospitals, adopted full digital pathology for primary histopathology diagnosis in 2016. OBJECTIVE.­: To describe the methodology adopted and the resulting experience at Granada University Hospitals in transitioning to full digital diagnosis. DESIGN.­: All histopathology glass slides generated for routine diagnosis were digitized at ×40 using the Philips IntelliSite Pathology Solution, which includes an ultrafast scanner and an image management system. All hematoxylin-eosin-stained preparations and immunohistochemistry and histochemistry slides were digitized. The existing sample-tracking software and image management system were integrated to allow data interchange through the Health Level 7 protocol. RESULTS.­: Circa 160 000 specimens have been signed out using digital pathology for primary diagnosis. This comprises more than 800 000 digitized glass slides. The scanning error rate during the implementation phase was below 1.5%, and subsequent workflow optimization rendered this rate negligible. Since implementation, Granada University Hospitals pathologists have signed out 21% more cases per year on average. CONCLUSIONS.­: Digital pathology is an adequate medium for primary histopathology diagnosis. Successful digitization relies on existing sample tracking and integration of the information technology infrastructure. Rapid and reliable scanning at ×40 equivalent was key to the transition to a fully digital workflow. Digital pathology resulted in efficiency gains in the preanalytical and analytical phases, and created the basis for computational pathology: the use of computer-assisted tools to aid diagnosis.

Study of Human Leukocyte Antigen (HLA) in 13 cases of familial frontal fibrosing alopecia: CYP21A2 gene p.V281L mutation from congenital adrenal hyperplasia linked to HLA class I haplotype HLA-A*33:01; B*14:02; C*08:02 as a genetic marker.

BACKGROUND/OBJECTIVES: The aetiology of frontal fibrosing alopecia is unknown, and its genetic aspect remains uncharacterised. The aim of this report is to elucidate if major histocompatibility complex is associated with familial frontal fibrosing alopecia. METHODS: A case-control study was performed of 13 patients with frontal fibrosing alopecia belonging to six families. Their human leukocyte antigen profiles were compared to the data of 636 healthy controls without frontal fibrosing alopecia. Patients underwent high-resolution genomic typing for human leukocyte antigen class I and II loci by PCR-SSO for Luminex. In addition, CYP21A2 gene (major histocompatibility complex class III) mutations were detected by PCR-SSO on strips. RESULTS: 61.5% of patients shared CYP21A2 gene p.V281L linked to the F16A human leukocyte antigen class I haplotype (HLA-A*33:01; B*14:02; C*08:02; Pc < 0.000001). The patients F16A-negative shared other human leukocyte antigen class I haplotypes: Y16A (3/13) and S26 (2/13). CONCLUSION: CYP21A2 gene p.V281L mutation can be used as a genetic marker for susceptibility to familial frontal fibrosing alopecia. Both the linkage of the mutation to F16A and the fact that F16A-negative patients share other human leukocyte antigen class I haplotype, point to an antigen-driven mechanism in susceptible patients with these haplotypes.

Quiste broncogénico gástrico: una localización singular / Gastric bronchogenic cyst: a singular presentation

Los quistes broncogénicos son formaciones quísticas originadas por una anomalía de desarrollo del eje traqueobronquial durante la embriogénesis a partir de la pared ventral del intestino anterior. Se define al quiste broncogénico como una yema pulmonar ectópica cuya localización más frecuente es en la carina, intraparenquimatoso y en el mediastino; presenta otras localizaciones atípicas y menos comunes, como las regiones cervical, supraclavicular, esofágica, retroperitoneal y cutánea. Se presenta el caso de una mujer joven que consultó al servicio de urgencias con síntomas de dolor abdominal, a quien se le practicaron múltiples pruebas y se le diagnosticó un quiste broncogénico gástrico.

Bronchogenic cysts are cystic formations originated by a tracheobronchial axis developmental anomaly during embryogenesis from the ventral wall of the anterior intestine. A bronchogenic cyst is defined as an ectopic pulmonary bud with most frequent location in the carina, intraparenchymal and mediastinal, presenting atypical and less common locations such as the cervical, supraclavicular, esophageal, retroperitoneal, and cutaneous regions. We present the case of a young woman, who consults the emergency department with abdominal pain, who is subjected to multiple tests being diagnosed as gastric bronchogenic cyst

Familial frontal fibrosing alopecia: A cross-sectional study of 20 cases from nine families.

BACKGROUND/OBJECTIVES: Frontal fibrosing alopecia (FFA) is a scarring alopecia whose prevalence is increasing. The pathogenesis of this disease is not well known. Genetic, environmental, hormonal and autoimmunity related factors have been considered; however, only a few cases of familial frontal fibrosing alopecia have been reported. MATERIAL AND METHODS: A cross-sectional study was performed at University Hospital in Granada (Spain). Twenty patients with frontal fibrosing alopecia belonging to nine different families were included, and clinical and dermoscopic features were analysed. RESULTS: Overall, 90% of the patients studied were women (mean age 61.4 years). About 50% of the patients had grade II frontal fibrosing alopecia at the time of diagnosis, whilst 35% had grades III or V. Mean recession was 2.83 cm in the frontal area and 1.99 cm in the temporo-parietal area. Daughters presented a shorter recession area and earlier debut of the disease than mothers. Androgenetic alopecia was found in only two patients (10%). The dermoscopic signs most commonly found were perifollicular erythema (85%), hyperkeratosis (85%), and absence of vellus hair in the hairline (78.9%). CONCLUSION: This study adds to the growing evidence that there is a genetic component to frontal fibrosing alopecia. The clinical pattern of frontal fibrosing alopecia was not different from that found in non-familial cases, but the debut of the disease in daughters of mothers with frontal fibrosing alopecia may be earlier.

A Case of Intertriginous and Flexural Exanthema caused by Amoxicillin.

We report a 50-year-old male patient who presented to the Dermatological Outpatient Clinic at the Hospital Universitario San Cecilio, Granada, Spain, in 2017 with symmetrical inguinal eruption and eruption on the dorsum of both feet four hours after the intake of amoxicillin. Physical examination showed confluent non-palpable purpuric macules covering the dorsum of both feet and medial malleolus, giving rise to dusky erythema in some areas. Only oral antihistamines were prescribed and cutaneous exanthema resolved within three weeks. Symmetric drug-related intertriginous and flexural exanthema (SDRIFE) is a sub-type of systemic allergic contact dermatitis, where previous sensitisation can only be demonstrated in up to 50% of patients by skin patch testing. Therefore, a provocation test was performed with amoxicillin without prior skin patch testing. As drug provocation produced the same reaction, the patient was diagnosed with SDRIFE. A parvovirus B19 infection was ruled out by negative serology. SDRIFE is challenging to distinguish from other skin rashes with similar features and distribution; it is important to be aware of these characteristics and their possible causes.

Pseudoangiosarcomatous squamous cell carcinoma of the skin: A need for a more rigorous nomenclature for histopathological variants of squamous cell carcinoma.

Over the years, squamous cell carcinomas (SCC) that mimicked vascular lesions have been encompassed within different classifications and the underlying etiopathogenic mechanisms have been interpreted in different ways by different authors. Here, we present a case of SCC with pseudovascular areas in the right leg of a 96-year-old woman with chronic venous insufficiency. Histopathological examination closely resembled an angiosarcoma, but the immunohistochemical negativity for endothelial markers and the strong positivity for the pancytokeratin marker AE1/AE3 revealed the epithelial nature of the neoplasm. After a comprehensive review of all similar previously published cases, we believe that it is necessary to separate SCC with pseudoluminal structures composed of glandular-like areas (pseudoglandular or adenoid SCC) from those mimicking vascular lumina (pseudovascular and pseudoangiosarcomatous SCC). We would like to emphasize that acantholytic SCC, a definitive variant of SCC, can be further classified into the common or ordinary subtype of acantholytic SCC, that shows solid nests containing numerous acantholytic atypical keratinocytes without any mimickers for specific structures, and pseudoglandular, pseudovascular and pseudoangiosarcomatous subtypes when glandular or vascular structures are mimicked, respectively.

Comparative Study of Shrinkage in Human Skin, Artificial Human Skin, and Mouse Skin.

INTRODUCTION: The shrinkage of surgical specimens (SS) is known in human skin (HS) but has not been studied in an artificial skin (AS) or mouse skin (MS). OBJECTIVES: To quantify the degree of shrinkage of SS and establish its timing in HS and an in vitro and animal model to explore the possible causes of this phenomenon. METHODOLOGY: We collected 100 SS of HS, 50 SS of AS synthesized with fibrin-agarose biomaterials and 21 SS of MS. The width and length of specimens were measured before the surgical excision (pre-SE), at 5 minutes postsurgery (ex vivo), and after 24 hours of fixation in formalin (postfixation). Histological staining was performed to analyze the differences between HS, AS, and MS that may explain the differences in shrinkage. RESULTS: Between pre-SE and postfixation, the width and length shrank by 16.1% and 17.1% in HS, 14.5% and 8.5% in AS, and 26.5% and 23.1% in MS (P < 0.01), respectively. Shrinkage largely occurred between pre-SE and ex vivo. Cells and interstitial fibers were scant in AS and abundant in MS. CONCLUSIONS: Almost all of the shrinkage occurred during the first 5 minutes postsurgery. According to the AS model findings, 53.6% of SS shrinkage would be explained by the action of dermal fibers and other cellular components of the dermis.