Combined colour deconvolution and artificial intelligence approach for region-selective immunohistochemical labelling quantification: The example of alpha smooth muscle actin in mouse kidney.

Immunohistochemical (IHC) localisation of protein expression is a widely used tool in pathology. This is semi-quantitative and exhibits substantial intra- and inter-observer variability. Digital approaches based on stain quantification applied to IHC are precise but still operator-dependent and time-consuming when regions of interest (ROIs) must be defined to quantify protein expression in a specific tissue area. This study aimed at developing an IHC quantification workflow that benefits from colour deconvolution for stain quantification and artificial intelligence for automatic ROI definition. The method was tested on 10 whole slide images (WSI) of alpha-smooth muscle actin (aSMA) stained mouse kidney sections. The task was to identify aSMA-positive areas within the glomeruli automatically. Total aSMA detection was performed using two channels (DAB, haematoxylin) colour deconvolution. Glomeruli segmentation within the same IHC WSI was performed by training a convolutional neural network with annotated examples of glomeruli. For both aSMA and glomeruli, binary masks were created. Co-localisation was performed by overlaying the masks and assigning red/green colours, with yellow indicative of a co-localised signal. The workflow described and exemplified using the case of aSMA expression in glomeruli can be applied to quantify the expression of IHC markers within different structures of immunohistochemically stained slides. The technique is objective, has a fully automated threshold approach (colour deconvolution phase) and uses AI to eliminate operator-dependent steps.

Quantifying acute kidney injury in an Ischaemia-Reperfusion Injury mouse model using deep-learning-based semantic segmentation in histology.

This study focuses on ischaemia-reperfusion injury (IRI) in kidneys, a cause of acute kidney injury (AKI) and end-stage kidney disease (ESKD). Traditional kidney damage assessment methods are semi-quantitative and subjective. This study aims to use a convolutional neural network (CNN) to segment murine kidney structures after IRI, quantify damage via CNN-generated pathological measurements, and compare this to conventional scoring. The CNN was able to accurately segment the different pathological classes, such as Intratubular casts and Tubular necrosis, with an F1 score of over 0.75. Some classes, such as Glomeruli and Proximal tubules, had even higher statistical values with F1 scores over 0.90. The scoring generated based on the segmentation approach statistically correlated with the semiquantitative assessment (Spearman's rank correlation coefficient=0.94). The heatmap approach localised the intratubular necrosis mainly in the outer stripe of the outer medulla, while the tubular casts were also present in more superficial or deeper portions of the cortex and medullary areas. This study presents a CNN model capable of segmenting multiple classes of interest, including acute IRI-specific pathological changes, in a whole mouse kidney section and can provide insights into the distribution of pathological classes within the whole mouse kidney section.

TNF-α induced extracellular release of keratinocyte high-mobility group box 1 in Stevens-Johnson syndrome/toxic epidermal necrolysis: Biomarker and putative mechanism of pathogenesis.

Decreased epidermal high-mobility group box 1 (HMGB1) expression is an early marker of epidermal injury in Stevens-Johnson syndrome/toxic epidermal necrolysis (SJS/TEN). Etanercept, an anti-tumor necrosis factor therapeutic, is effective in the treatment of SJS/TEN. The objective was to characterize antitumor necrosis factor-alpha (TNF-α)-mediated HMGB1 keratinocyte/epidermal release and etanercept modulation. HMGB1 release from TNF-α treated (± etanercept), or doxycycline-inducible RIPK3 or Bak-expressing human keratinocyte cells (HaCaTs) was determined by western blot/ELISA. Healthy skin explants were treated with TNF-α or serum (1:10 dilution) from immune checkpoint inhibitor-tolerant, lichenoid dermatitis or SJS/TEN patients ± etanercept. Histological and immunohistochemical analysis of HMGB1 was undertaken. TNF-α induced HMGB1 release in vitro via both necroptosis and apoptosis. Exposure of skin explants to TNF-α or SJS/TEN serum resulted in significant epidermal toxicity/detachment with substantial HMGB1 release which was attenuated by etanercept. Whole-slide image analysis of biopsies demonstrated significantly lower epidermal HMGB1 in pre-blistered SJS/TEN versus control (P < 0.05). Keratinocyte HMGB1 release, predominantly caused by necroptosis, can be attenuated by etanercept. Although TNF-α is a key mediator of epidermal HMGB1 release, other cytokines/cytotoxic proteins also contribute. Skin explant models represent a potential model of SJS/TEN that could be utilized for further mechanistic studies and targeted therapy screening.

Use of deep learning for the classification of hyperplastic lymph node and common subtypes of canine lymphomas: a preliminary study.

Artificial Intelligence has observed significant growth in its ability to classify different types of tumors in humans due to advancements in digital pathology technology. Among these tumors, lymphomas are quite common in dogs, despite studies on the application of AI in domestic species are scarce. This research aims to employ deep learning (DL) through convolutional neural networks (CNNs) to distinguish between normal lymph nodes and 3 WHO common subtypes of canine lymphomas. To train and validate the CNN, 1,530 high-resolution microscopic images derived from whole slide scans (WSIs) were used, including those of background areas, hyperplastic lymph nodes (n = 4), and three different lymphoma subtypes: diffuse large B cell lymphoma (DLBCL; n = 5), lymphoblastic (LBL; n = 5), and marginal zone lymphoma (MZL; n = 3). The CNN was able to correctly identify 456 images of the possible 457 test sets, achieving a maximum accuracy of 99.34%. The results of this study have demonstrated the feasibility of using deep learning to differentiate between hyperplastic lymph nodes and lymphomas, as well as to classify common WHO subtypes. Further research is required to explore the implications of these findings and validate the ability of the network to classify a broader range of lymphomas.