Community-developed checklists for publishing images and image analyses.

Images document scientific discoveries and are prevalent in modern biomedical research. Microscopy imaging in particular is currently undergoing rapid technological advancements. However, for scientists wishing to publish obtained images and image-analysis results, there are currently no unified guidelines for best practices. Consequently, microscopy images and image data in publications may be unclear or difficult to interpret. Here, we present community-developed checklists for preparing light microscopy images and describing image analyses for publications. These checklists offer authors, readers and publishers key recommendations for image formatting and annotation, color selection, data availability and reporting image-analysis workflows. The goal of our guidelines is to increase the clarity and reproducibility of image figures and thereby to heighten the quality and explanatory power of microscopy data.

Oncogenic role and target properties of the lysine-specific demethylase KDM1A in chronic lymphocytic leukemia.

In chronic lymphocytic leukemia (CLL), epigenetic alterations are considered to centrally shape the transcriptional signatures that drive disease evolution and underlie its biological and clinical subsets. Characterizations of epigenetic regulators, particularly histone-modifying enzymes, are very rudimentary in CLL. In efforts to establish effectors of the CLL-associated oncogene T-cell leukemia 1A (TCL1A), we identified here the lysine-specific histone demethylase KDM1A to interact with the TCL1A protein in B cells in conjunction with an increased catalytic activity of KDM1A. We demonstrate that KDM1A is upregulated in malignant B cells. Elevated KDM1A and associated gene expression signatures correlated with aggressive disease features and adverse clinical outcomes in a large prospective CLL trial cohort. Genetic Kdm1a knockdown in Eµ-TCL1A mice reduced leukemic burden and prolonged animal survival, accompanied by upregulated p53 and proapoptotic pathways. Genetic KDM1A depletion also affected milieu components (T, stromal, and monocytic cells), resulting in significant reductions in their capacity to support CLL-cell survival and proliferation. Integrated analyses of differential global transcriptomes (RNA sequencing) and H3K4me3 marks (chromatin immunoprecipitation sequencing) in Eµ-TCL1A vs iKdm1aKD;Eµ-TCL1A mice (confirmed in human CLL) implicate KDM1A as an oncogenic transcriptional repressor in CLL which alters histone methylation patterns with pronounced effects on defined cell death and motility pathways. Finally, pharmacologic KDM1A inhibition altered H3K4/9 target methylation and revealed marked anti-B-cell leukemic synergisms. Overall, we established the pathogenic role and effector networks of KDM1A in CLL via tumor-cell intrinsic mechanisms and its impacts in cells of the microenvironment. Our data also provide rationales to further investigate therapeutic KDM1A targeting in CLL.

Genomic and Phenotypic Biomarkers for Precision Medicine Guidance in Advanced Prostate Cancer.

OPINION STATEMENT: Prostate cancer (PCa) is the second most diagnosed malignant neoplasm and is one of the leading causes of cancer-related death in men worldwide. Despite significant advances in screening and treatment of PCa, given the heterogeneity of this disease, optimal personalized therapeutic strategies remain limited. However, emerging predictive and prognostic biomarkers based on individual patient profiles in combination with computer-assisted diagnostics have the potential to guide precision medicine, where patients may benefit from therapeutic approaches optimally suited to their disease. Also, the integration of genotypic and phenotypic diagnostic methods is supporting better informed treatment decisions. Focusing on advanced PCa, this review discusses polygenic risk scores for screening of PCa and common genomic aberrations in androgen receptor (AR), PTEN-PI3K-AKT, and DNA damage response (DDR) pathways, considering clinical implications for diagnosis, prognosis, and treatment prediction. Furthermore, we evaluate liquid biopsy, protein biomarkers such as serum testosterone levels, SLFN11 expression, total alkaline phosphatase (tALP), neutrophil-to-lymphocyte ratio (NLR), tissue biopsy, and advanced imaging tools, summarizing current phenotypic biomarkers and envisaging more effective utilization of diagnostic and prognostic biomarkers in advanced PCa. We conclude that prognostic and treatment predictive biomarker discovery can improve the management of patients, especially in metastatic stages of advanced PCa. This will result in decreased mortality and enhanced quality of life and help design a personalized treatment regimen.

AC-265347 Inhibits Neuroblastoma Tumor Growth by Induction of Differentiation without Causing Hypocalcemia.

Neuroblastoma is the most common extracranial solid tumor of childhood, with heterogeneous clinical manifestations ranging from spontaneous regression to aggressive metastatic disease. The calcium-sensing receptor (CaSR) is a G protein-coupled receptor (GPCR) that senses plasmatic fluctuation in the extracellular concentration of calcium and plays a key role in maintaining calcium homeostasis. We have previously reported that this receptor exhibits tumor suppressor properties in neuroblastoma. The activation of CaSR with cinacalcet, a positive allosteric modulator of CaSR, reduces neuroblastoma tumor growth by promoting differentiation, endoplasmic reticulum (ER) stress and apoptosis. However, cinacalcet treatment results in unmanageable hypocalcemia in patients. Based on the bias signaling shown by calcimimetics, we aimed to identify a new drug that might exert tumor-growth inhibition similar to cinacalcet, without affecting plasma calcium levels. We identified a structurally different calcimimetic, AC-265347, as a promising therapeutic agent for neuroblastoma, since it reduced tumor growth by induction of differentiation, without affecting plasma calcium levels. Microarray analysis suggested biased allosteric modulation of the CaSR signaling by AC-265347 and cinacalcet towards distinct intracellular pathways. No upregulation of genes involved in calcium signaling and ER stress were observed in patient-derived xenografts (PDX) models exposed to AC-265347. Moreover, the most significant upregulated biological pathways promoted by AC-265347 were linked to RHO GTPases signaling. AC-265347 upregulated cancer testis antigens (CTAs), providing new opportunities for CTA-based immunotherapies. Taken together, this study highlights the importance of the biased allosteric modulation when targeting GPCRs in cancer. More importantly, the capacity of AC-265347 to promote differentiation of malignant neuroblastoma cells provides new opportunities, alone or in combination with other drugs, to treat high-risk neuroblastoma patients.

Characterization of Negative Allosteric Modulators of the Calcium-Sensing Receptor for Repurposing as a Treatment of Asthma.

Asthma is still an incurable disease, and there is a recognized need for novel small-molecule therapies for people with asthma, especially those poorly controlled by current treatments. We previously demonstrated that calcium-sensing receptor (CaSR) negative allosteric modulators (NAMs), calcilytics, uniquely suppress both airway hyperresponsiveness (AHR) and inflammation in human cells and murine asthma surrogates. Here we assess the feasibility of repurposing four CaSR NAMs, which were originally developed for oral therapy for osteoporosis and previously tested in the clinic as a novel, single, and comprehensive topical antiasthma therapy. We address the hypotheses, using murine asthma surrogates, that topically delivered CaSR NAMs 1) abolish AHR; 2) are unlikely to cause unwanted systemic effects; 3) are suitable for topical application; and 4) inhibit airway inflammation to the same degree as the current standard of care, inhaled corticosteroids, and, furthermore, inhibit airway remodeling. All four CaSR NAMs inhibited poly-L-arginine-induced AHR in naïve mice and suppressed both AHR and airway inflammation in a murine surrogate of acute asthma, confirming class specificity. Repeated exposure to inhaled CaSR NAMs did not alter blood pressure, heart rate, or serum calcium concentrations. Optimal candidates for repurposing were identified based on anti-AHR/inflammatory activities, pharmacokinetics/pharmacodynamics, formulation, and micronization studies. Whereas both inhaled CaSR NAMs and inhaled corticosteroids reduced airways inflammation, only the former prevented goblet cell hyperplasia in a chronic asthma model. We conclude that inhaled CaSR NAMs are likely a single, safe, and effective topical therapy for human asthma, abolishing AHR, suppressing airways inflammation, and abrogating some features of airway remodeling. SIGNIFICANCE STATEMENT: Calcium-sensing receptor (CaSR) negative allosteric modulators (NAMs) reduce airway smooth muscle hyperresponsiveness, reverse airway inflammation as efficiently as topical corticosteroids, and suppress airway remodeling in asthma surrogates. CaSR NAMs, which were initially developed for oral therapy of osteoporosis proved inefficacious for this indication despite being safe and well tolerated. Here we show that structurally unrelated CaSR NAMs are suitable for inhaled delivery and represent a one-stop, steroid-free approach to asthma control and prophylaxis.

CD10 inhibits cell motility but expression is associated with advanced stage disease in colorectal cancer.

INTRODUCTION: CD10 is a cell membrane-bound endopeptidase which is expressed in normal small bowel but not in normal colon. It is aberrantly expressed in a small proportion of colorectal cancers (CRC) and this has been associated with liver metastasis and poor prognosis. We sought to investigate the mechanism of CD10 activity and its association with clinicopathological features. MATERIAL AND METHODS: CD10 was stably knocked down by lentiviral shRNA transduction in the CRC cell lines SW480 and SW620 which are derived from a primary tumour and its corresponding metastasis respectively. Expression of epithelial - mesenchymal transition (EMT) markers was tested as well as the effect of knockdown on cell viability, migration and invasion assays. In addition, immunohistochemical expression of CD10 in primary colorectal tumours (N = 84) in a tissue microarray was digitally quantified and analysed for associations with clinicopathological variables. RESULTS: Knockdown of CD10 did not alter cell viability in SW480, but migration and invasion levels increased (P < 0.001 for each) and this was associated with a cadherin switch. In SW620, CD10 knockdown caused a reduction in cell viability after 72 h (P = 0.0018) but it had no effect on cell migration and invasion. Expression of epithelial CD10 in primary tumours was associated with presence of lymph node invasion (P = 0.001) and advanced Duke's stage (P = 0.001). CONCLUSIONS: Our results suggest that the function of CD10 may change during tumour evolution. It may inhibit cell motility in early-stage disease whilst promoting cell viability in late-stage disease. It has a complex role and further studies are needed to elucidate the suitability of CD10 as a prognostic marker or therapeutic target.

Community-developed checklists for publishing images and image analyses.

Images document scientific discoveries and are prevalent in modern biomedical research. Microscopy imaging in particular is currently undergoing rapid technological advancements. However for scientists wishing to publish the obtained images and image analyses results, there are to date no unified guidelines. Consequently, microscopy images and image data in publications may be unclear or difficult to interpret. Here we present community-developed checklists for preparing light microscopy images and image analysis for publications. These checklists offer authors, readers, and publishers key recommendations for image formatting and annotation, color selection, data availability, and for reporting image analysis workflows. The goal of our guidelines is to increase the clarity and reproducibility of image figures and thereby heighten the quality and explanatory power of microscopy data is in publications.

A dual decoder U-Net-based model for nuclei instance segmentation in hematoxylin and eosin-stained histological images.

Even in the era of precision medicine, with various molecular tests based on omics technologies available to improve the diagnosis process, microscopic analysis of images derived from stained tissue sections remains crucial for diagnostic and treatment decisions. Among other cellular features, both nuclei number and shape provide essential diagnostic information. With the advent of digital pathology and emerging computerized methods to analyze the digitized images, nuclei detection, their instance segmentation and classification can be performed automatically. These computerized methods support human experts and allow for faster and more objective image analysis. While methods ranging from conventional image processing techniques to machine learning-based algorithms have been proposed, supervised convolutional neural network (CNN)-based techniques have delivered the best results. In this paper, we propose a CNN-based dual decoder U-Net-based model to perform nuclei instance segmentation in hematoxylin and eosin (H&E)-stained histological images. While the encoder path of the model is developed to perform standard feature extraction, the two decoder heads are designed to predict the foreground and distance maps of all nuclei. The outputs of the two decoder branches are then merged through a watershed algorithm, followed by post-processing refinements to generate the final instance segmentation results. Moreover, to additionally perform nuclei classification, we develop an independent U-Net-based model to classify the nuclei predicted by the dual decoder model. When applied to three publicly available datasets, our method achieves excellent segmentation performance, leading to average panoptic quality values of 50.8%, 51.3%, and 62.1% for the CryoNuSeg, NuInsSeg, and MoNuSAC datasets, respectively. Moreover, our model is the top-ranked method in the MoNuSAC post-challenge leaderboard.

Disturbances in microbial skin recolonization and cutaneous immune response following allogeneic stem cell transfer.

The composition of the gut microbiome influences the clinical course after allogeneic hematopoietic stem cell transplantation (HSCT), but little is known about the relevance of skin microorganisms. In a single-center, observational study, we recruited a cohort of 50 patients before undergoing conditioning treatment and took both stool and skin samples up to one year after HSCT. We could confirm intestinal dysbiosis following HSCT and report that the skin microbiome is likewise perturbed in HSCT-recipients. Overall bacterial colonization of the skin was decreased after conditioning. Particularly patients that developed acute skin graft-versus-host disease (aGVHD) presented with an overabundance of Staphylococcus spp. In addition, a loss in alpha diversity was indicative of aGVHD development already before disease onset and correlated with disease severity. Further, co-localization of CD45+ leukocytes and staphylococci was observed in the skin of aGVHD patients even before disease development and paralleled with upregulated genes required for antigen-presentation in mononuclear phagocytes. Overall, our data reveal disturbances of the skin microbiome as well as cutaneous immune response in HSCT recipients with changes associated with cutaneous aGVHD.

Clinical and cytological effects of pimecrolimus cream 1% after resolution of active atopic dermatitis lesions by topical corticosteroids: a randomized controlled trial.

BACKGROUND: Topical pimecrolimus may maintain remissions of atopic dermatitis (AD) by inhibiting subclinical inflammation. OBJECTIVE: To evaluate clinical and cytological effects of pimecrolimus in topical corticosteroid-treated and resolved AD lesions. METHODS: Patients (n=67) with resolved AD lesions were randomized to 3-week double-blind treatment with either pimecrolimus cream 1% or vehicle cream. Outcome measures were reduction in Eczema Area and Severity Index (EASI) and number of leukocytes in skin biopsies in all randomized patients who were evaluable at the end of study. RESULTS: The proportion of patients with a localized EASI<2 at the end of study was higher with pimecrolimus cream 1% than with vehicle cream (73.5 vs. 39.4%, respectively). There was a significant decrease in the number of infiltrating CD45+ cells in pimecrolimus cream 1% compared with placebo cream (-88.2 vs. 43.2 cells/mm(2), respectively, p=0.047) and a slight but nonsignificant reduction in the number of dermal dendritic cells, Langerhans cells, T cells and macrophages with pimecrolimus versus vehicle cream. LIMITATIONS: This was an exploratory study. CONCLUSION: Topical pimecrolimus was effective at maintaining betamethasone-17α-valerate-induced AD remission by inhibiting recurrences of the inflammatory infiltrate in the skin.

CryoNuSeg: A dataset for nuclei instance segmentation of cryosectioned H&E-stained histological images.

Nuclei instance segmentation plays an important role in the analysis of hematoxylin and eosin (H&E)-stained images. While supervised deep learning (DL)-based approaches represent the state-of-the-art in automatic nuclei instance segmentation, annotated datasets are required to train these models. There are two main types of tissue processing protocols resulting in formalin-fixed paraffin-embedded samples (FFPE) and frozen tissue samples (FS), respectively. Although FFPE-derived H&E stained tissue sections are the most widely used samples, H&E staining of frozen sections derived from FS samples is a relevant method in intra-operative surgical sessions as it can be performed more rapidly. Due to differences in the preparation of these two types of samples, the derived images and in particular the nuclei appearance may be different in the acquired whole slide images. Analysis of FS-derived H&E stained images can be more challenging as rapid preparation, staining, and scanning of FS sections may lead to deterioration in image quality. In this paper, we introduce CryoNuSeg, the first fully annotated FS-derived cryosectioned and H&E-stained nuclei instance segmentation dataset. The dataset contains images from 10 human organs that were not exploited in other publicly available datasets, and is provided with three manual mark-ups to allow measuring intra-observer and inter-observer variabilities. Moreover, we investigate the effects of tissue fixation/embedding protocol (i.e., FS or FFPE) on the automatic nuclei instance segmentation performance and provide a baseline segmentation benchmark for the dataset that can be used in future research. A step-by-step guide to generate the dataset as well as the full dataset and other detailed information are made available to fellow researchers at https://github.com/masih4/CryoNuSeg.

Investigating the Impact of the Bit Depth of Fluorescence-Stained Images on the Performance of Deep Learning-Based Nuclei Instance Segmentation.

Nuclei instance segmentation can be considered as a key point in the computer-mediated analysis of histological fluorescence-stained (FS) images. Many computer-assisted approaches have been proposed for this task, and among them, supervised deep learning (DL) methods deliver the best performances. An important criterion that can affect the DL-based nuclei instance segmentation performance of FS images is the utilised image bit depth, but to our knowledge, no study has been conducted so far to investigate this impact. In this work, we released a fully annotated FS histological image dataset of nuclei at different image magnifications and from five different mouse organs. Moreover, by different pre-processing techniques and using one of the state-of-the-art DL-based methods, we investigated the impact of image bit depth (i.e., eight bits vs. sixteen bits) on the nuclei instance segmentation performance. The results obtained from our dataset and another publicly available dataset showed very competitive nuclei instance segmentation performances for the models trained with 8 bit and 16 bit images. This suggested that processing 8 bit images is sufficient for nuclei instance segmentation of FS images in most cases. The dataset including the raw image patches, as well as the corresponding segmentation masks is publicly available in the published GitHub repository.

Next-Generation Digital Histopathology of the Tumor Microenvironment.

Progress in cancer research is substantially dependent on innovative technologies that permit a concerted analysis of the tumor microenvironment and the cellular phenotypes resulting from somatic mutations and post-translational modifications. In view of a large number of genes, multiplied by differential splicing as well as post-translational protein modifications, the ability to identify and quantify the actual phenotypes of individual cell populations in situ, i.e., in their tissue environment, has become a prerequisite for understanding tumorigenesis and cancer progression. The need for quantitative analyses has led to a renaissance of optical instruments and imaging techniques. With the emergence of precision medicine, automated analysis of a constantly increasing number of cellular markers and their measurement in spatial context have become increasingly necessary to understand the molecular mechanisms that lead to different pathways of disease progression in individual patients. In this review, we summarize the joint effort that academia and industry have undertaken to establish methods and protocols for molecular profiling and immunophenotyping of cancer tissues for next-generation digital histopathology-which is characterized by the use of whole-slide imaging (brightfield, widefield fluorescence, confocal, multispectral, and/or multiplexing technologies) combined with state-of-the-art image cytometry and advanced methods for machine and deep learning.

MoNuSAC2020: A Multi-Organ Nuclei Segmentation and Classification Challenge.

Detecting various types of cells in and around the tumor matrix holds a special significance in characterizing the tumor micro-environment for cancer prognostication and research. Automating the tasks of detecting, segmenting, and classifying nuclei can free up the pathologists' time for higher value tasks and reduce errors due to fatigue and subjectivity. To encourage the computer vision research community to develop and test algorithms for these tasks, we prepared a large and diverse dataset of nucleus boundary annotations and class labels. The dataset has over 46,000 nuclei from 37 hospitals, 71 patients, four organs, and four nucleus types. We also organized a challenge around this dataset as a satellite event at the International Symposium on Biomedical Imaging (ISBI) in April 2020. The challenge saw a wide participation from across the world, and the top methods were able to match inter-human concordance for the challenge metric. In this paper, we summarize the dataset and the key findings of the challenge, including the commonalities and differences between the methods developed by various participants. We have released the MoNuSAC2020 dataset to the public.

The effects of skin lesion segmentation on the performance of dermatoscopic image classification.

BACKGROUND AND OBJECTIVE: Malignant melanoma (MM) is one of the deadliest types of skin cancer. Analysing dermatoscopic images plays an important role in the early detection of MM and other pigmented skin lesions. Among different computer-based methods, deep learning-based approaches and in particular convolutional neural networks have shown excellent classification and segmentation performances for dermatoscopic skin lesion images. These models can be trained end-to-end without requiring any hand-crafted features. However, the effect of using lesion segmentation information on classification performance has remained an open question. METHODS: In this study, we explicitly investigated the impact of using skin lesion segmentation masks on the performance of dermatoscopic image classification. To do this, first, we developed a baseline classifier as the reference model without using any segmentation masks. Then, we used either manually or automatically created segmentation masks in both training and test phases in different scenarios and investigated the classification performances. The different scenarios included approaches that exploited the segmentation masks either for cropping of skin lesion images or removing the surrounding background or using the segmentation masks as an additional input channel for model training. RESULTS: Evaluated on the ISIC 2017 challenge dataset which contained two binary classification tasks (i.e. MM vs. all and seborrheic keratosis (SK) vs. all) and based on the derived area under the receiver operating characteristic curve scores, we observed four main outcomes. Our results show that 1) using segmentation masks did not significantly improve the MM classification performance in any scenario, 2) in one of the scenarios (using segmentation masks for dilated cropping), SK classification performance was significantly improved, 3) removing all background information by the segmentation masks significantly degraded the overall classification performance, and 4) in case of using the appropriate scenario (using segmentation for dilated cropping), there is no significant difference of using manually or automatically created segmentation masks. CONCLUSIONS: We systematically explored the effects of using image segmentation on the performance of dermatoscopic skin lesion classification.

Transfer learning using a multi-scale and multi-network ensemble for skin lesion classification.

BACKGROUND AND OBJECTIVE: Skin cancer is among the most common cancer types in the white population and consequently computer aided methods for skin lesion classification based on dermoscopic images are of great interest. A promising approach for this uses transfer learning to adapt pre-trained convolutional neural networks (CNNs) for skin lesion diagnosis. Since pre-training commonly occurs with natural images of a fixed image resolution and these training images are usually significantly smaller than dermoscopic images, downsampling or cropping of skin lesion images is required. This however may result in a loss of useful medical information, while the ideal resizing or cropping factor of dermoscopic images for the fine-tuning process remains unknown. METHODS: We investigate the effect of image size for skin lesion classification based on pre-trained CNNs and transfer learning. Dermoscopic images from the International Skin Imaging Collaboration (ISIC) skin lesion classification challenge datasets are either resized to or cropped at six different sizes ranging from 224 × 224 to 450 × 450. The resulting classification performance of three well established CNNs, namely EfficientNetB0, EfficientNetB1 and SeReNeXt-50 is explored. We also propose and evaluate a multi-scale multi-CNN (MSM-CNN) fusion approach based on a three-level ensemble strategy that utilises the three network architectures trained on cropped dermoscopic images of various scales. RESULTS: Our results show that image cropping is a better strategy compared to image resizing delivering superior classification performance at all explored image scales. Moreover, fusing the results of all three fine-tuned networks using cropped images at all six scales in the proposed MSM-CNN approach boosts the classification performance compared to a single network or a single image scale. On the ISIC 2018 skin lesion classification challenge test set, our MSM-CNN algorithm yields a balanced multi-class accuracy of 86.2% making it the currently second ranked algorithm on the live leaderboard. CONCLUSIONS: We confirm that the image size has an effect on skin lesion classification performance when employing transfer learning of CNNs. We also show that image cropping results in better performance compared to image resizing. Finally, a straightforward ensembling approach that fuses the results from images cropped at six scales and three fine-tuned CNNs is shown to lead to the best classification performance.

A Multi-Organ Nucleus Segmentation Challenge.

Generalized nucleus segmentation techniques can contribute greatly to reducing the time to develop and validate visual biomarkers for new digital pathology datasets. We summarize the results of MoNuSeg 2018 Challenge whose objective was to develop generalizable nuclei segmentation techniques in digital pathology. The challenge was an official satellite event of the MICCAI 2018 conference in which 32 teams with more than 80 participants from geographically diverse institutes participated. Contestants were given a training set with 30 images from seven organs with annotations of 21,623 individual nuclei. A test dataset with 14 images taken from seven organs, including two organs that did not appear in the training set was released without annotations. Entries were evaluated based on average aggregated Jaccard index (AJI) on the test set to prioritize accurate instance segmentation as opposed to mere semantic segmentation. More than half the teams that completed the challenge outperformed a previous baseline. Among the trends observed that contributed to increased accuracy were the use of color normalization as well as heavy data augmentation. Additionally, fully convolutional networks inspired by variants of U-Net, FCN, and Mask-RCNN were popularly used, typically based on ResNet or VGG base architectures. Watershed segmentation on predicted semantic segmentation maps was a popular post-processing strategy. Several of the top techniques compared favorably to an individual human annotator and can be used with confidence for nuclear morphometrics.

Monocytes/macrophages in kidney allograft intimal arteritis: no association with markers of humoral rejection or with inferior outcome.

BACKGROUND: Several studies indicate that interstitial and intracapillary monocytes/macrophages (MO) represent a significant proportion of graft-infiltrating cells in renal allografts and that their presence may unfavourably affect clinical outcome. Much less is known about the role of MO in vascular rejection of transplanted kidneys. The aim of our study was to determine the cellular composition of immune cell infiltrates in intimal arteritis and to analyse whether it is associated with features of humoral immunity and impaired graft survival. METHODS: In 34 recipients with vascular rejection, we determined the proportion of intimal and interstitial MO and T-cells (expressed as ratio of CD68- and CD3-positive cells) in immunohistochemically double-labelled slides. RESULTS: Intimal arteritis is always composed of T-cells and MO with a median CD68/CD3 ratio of 1.03. In 47% of cases, however, T-cells predominate (CD68/CD3 ratio <1). The median interstitial CD68/CD3 ratio is 0.61, with T-cells dominating in 64% of cases. There is no correlation between the cellular composition of arterial and interstitial infiltrates. The proportion of interstitial and arterial MO has no impact on graft survival, and is, in contrast to previous reports on MO in allograft glomerulitis and capillaritis, not associated with C4d staining. CONCLUSIONS: Intimal arteritis in kidney allograft rejection is composed of a mixed infiltrate of MO and T-lymphocytes. In contrast to MO in PTCitis and glomerulitis, the MO in intimal arteritis are not associated with markers of humoral immune response and there are no different allograft outcomes between MO and T-lymphocyte-dominated groups.

Fusing fine-tuned deep features for skin lesion classification.

Malignant melanoma is one of the most aggressive forms of skin cancer. Early detection is important as it significantly improves survival rates. Consequently, accurate discrimination of malignant skin lesions from benign lesions such as seborrheic keratoses or benign nevi is crucial, while accurate computerised classification of skin lesion images is of great interest to support diagnosis. In this paper, we propose a fully automatic computerised method to classify skin lesions from dermoscopic images. Our approach is based on a novel ensemble scheme for convolutional neural networks (CNNs) that combines intra-architecture and inter-architecture network fusion. The proposed method consists of multiple sets of CNNs of different architecture that represent different feature abstraction levels. Each set of CNNs consists of a number of pre-trained networks that have identical architecture but are fine-tuned on dermoscopic skin lesion images with different settings. The deep features of each network were used to train different support vector machine classifiers. Finally, the average prediction probability classification vectors from different sets are fused to provide the final prediction. Evaluated on the 600 test images of the ISIC 2017 skin lesion classification challenge, the proposed algorithm yields an area under receiver operating characteristic curve of 87.3% for melanoma classification and an area under receiver operating characteristic curve of 95.5% for seborrheic keratosis classification, outperforming the top-ranked methods of the challenge while being simpler compared to them. The obtained results convincingly demonstrate our proposed approach to represent a reliable and robust method for feature extraction, model fusion and classification of dermoscopic skin lesion images.