Histological Slides Research Articles

Preservation of Testicular Tissue and Alleviation of Oxidative Stress by Carvacrol Following Torsion/ Detorsion in Adult Male Rats

Objectives: Testicular torsion is a critical medical condition necessitating prompt diagnosis and intervention. This study aimed to explore the potential safeguarding effect of carvacrol against histological damage and oxidative stress resulting from torsion/detorsion (T/D) in rat testes. Materials and Methods: A total of 32 adult male rats were randomly divided into four groups. The sham group did not undergo any intervention. The second group received an intraperitoneal injection of 75 mg/kg carvacrol half an hour before detorsion application. The third group was administered 80 mg/kg carvacrol intraperitoneally without detorsion. The fourth group (control) experienced (T/D) through the administration of saline. Following a 5-hour reperfusion period, the left testis was excised for histological slide preparation. Blood serum was used to measure antioxidant enzyme levels. Data analysis was performed using the SPSS version 19 software and analysis of variance tests. Results: Significant histological alterations were observed between the sham and other three groups. The levels of superoxide dismutase (SOD), glutathione peroxidase (GPx), and testosterone significantly decreased in all treatment groups compared to the control group (P<0.05). Conversely, malondialdehyde (MDA) levels increased in the torsion control group in contrast to the sham group (P<0.05). Carvacrol administration mitigated MDA levels in the treatment groups. Also, there were significant differences in tissue parameters between the sham and the other groups (P<0.05). Conclusions: According to the results of this study, carvacrol possesses the potential to mitigate testicular tissue damage, enhance testicular function, and ameliorate oxidative stress consequential to testicular rotation.

Synthesis and antitumor activity of a 2-hydroxy substituted chalcone (C6)

Cancers account for approximately 13% of all deaths each year with breast cancer being one of the most common. Chemotherapeutic agents used in cancer usually have nonspecific toxicity which kills both tumor and normal cells and cause serious side effects that often limit their efficacy. This study aimed to synthesize chalcone and evaluate its antitumor activity. Chalcone was synthesized using Claisen-Schmidt condensation and characterized using spectroscopic techniques. The LD50 of the synthesized compound was estimated using OECD-425 guidelines in rats. A mammary tumor was induced with a single subcutaneous administration of 65 mg/kg of 1-methyl nitrosourea (MNU). The rats were palpated weekly to determine the size of the tumor. Eight weeks post MNU administration, the rats were divided into five groups of six rats each and treated with graded doses (12.5, 25, and 50 mg/kg) of the compound and paclitaxel (10 mg/kg) for six weeks. Before treatment, three rats were randomly selected and sacrificed, and mammary gland samples were subjected to histological assessment to confirm the induction of the tumor. At the end of the treatment, the rats were euthanized, and mammary glands were collected and subjected to histological evaluation. The possible mechanism of action of the synthesized compound was elucidated in silico using molecular docking. The compound was synthesized and named C6. C6 was found to be relatively safe with LD50 above 2000 mg/kg and exhibited remarkable antitumor activity. MNU-induced mammary tumor rats treated with C6 produced a significant decrease in tumor diameter when compared with the untreated group, histology slides displayed fewer signs of hyperplasia and small numbers of connective tissue with larger lobules when compared with the untreated group. In silico tubulin-binding interactions revealed that C6 binding to tubulin was like that of colchicine, and also has higher affinity to tubulin than colchicine. The synthesized chalcone demonstrated significant antitumor activities in MNU-induced mammary tumors in rats postulated via inhibition of tubulin polymerization.

Mixed clear cell/endometrioid and clear cell/serous carcinoma of the uterus are clinicopathologically similar to pure clear cell carcinoma: An NRG Oncology/Gynecologic Oncology Group (GOG-210) study of 311 women

ObjectivesClear cell carcinoma is a high-risk subtype of endometrial cancer. Some patients have a mixture of clear cell carcinoma with other histologic types (endometrioid or serous) or cannot be neatly assigned to one of these types. Protocol GOG-8032 within GOG-210 was designed to determine whether these tumors differ from pure clear cell carcinoma in stage at diagnosis, initial pattern of spread, or patient survival. MethodsThe term “mixed” was applied to tumors with multiple identifiable components, and “indeterminate” was applied to tumors with features intermediate between different histologic types. Three hundred eleven women with pure, mixed, or indeterminate clear cell carcinoma were identified in a larger cohort of patients undergoing hysterectomy for endometrial cancer in GOG-210. Histologic slides were centrally reviewed by expert pathologists. Baseline and follow-up data were analyzed. ResultsOne hundred thirty-six patients had pure clear cell carcinoma and 175 had a mixed or indeterminate clear cell pattern. Baseline clinicopathologic characteristics were similar except for a small difference in age at presentation. Univariate survival analysis confirmed the significance of typical endometrial cancer prognostic factors. Patients in the mixed categories had disease-free and overall survival similar to pure clear cell carcinoma, but the indeterminate clear cell/endometrioid group had longer survival. ConclusionIn clear cell endometrial cancer, the presence of a definite admixed endometrioid or serous component did not correlate with a significant difference in prognosis. Patients whose tumors had indeterminate clear cell features had better prognosis. Some of these tumors may be endometrioid tumors mimicking clear cell carcinoma.

Optimizing Storage and Computational Efficiency: An Efficient Algorithm for Whole Slide Image Size Reduction

ObjectiveTo efficiently store, transfer, and analyze whole slide imaging (WSI), we developed an image-processing algorithm to remove the unneeded background in a WSI and assemble tissue-containing parts into smaller WSIs without any change in tissue area image resolution. Patients and MethodsWe used histology slides of nondysplastic Barrett esophagus, low-grade dysplasia, and high-grade dysplasia, which were digitized using Aperio AT2 Scanner from January 1992 to September 2020. The algorithm involved converting color images to grayscale images, binarizing images by assigning zero to the background and 1 to the foreground, filling the holes and dilating the foreground masks, and extracting connected components. Using the coordinates of each component, the vertices of the smallest surrounding bounding box were calculated, and tissue-containing parts were cropped from the original slide. The smallest possible rectangle that encloses all bounding boxes containing tissue was found using the rectangle-packer package. ResultsThe algorithm resulted in a mean reduction of 7.11×. The performance of a previously developed deep learning model for the detection of Barrett esophagus dysplasia grade on the size-reduced WSIs was comparable with that on the original WSIs. ConclusionOur algorithm for WSI size reduction can assist researchers in storing, transferring, and analyzing WSIs while optimally using them in their workflow.

Histomorphometric Changes of the Masseter Muscle of Rats After a Single Injection of Botulinum Toxin Type A.

It has been reported that botulinum toxin type A (BoNT-A) produces structural changes in masticatory muscles. However, not all histomorphometric parameters affected by BoNT-A parameters have been assessed. This study investigated the histomorphometric changes in the masseter muscle of rats after a single injection of BoNT-A. Forty-four adult animals were randomly divided into control group (n = 22) and BoNT-A group (n = 22). Controls received a single dose of 0.14mL/kg of saline in masseter muscles, and the BoNT-A group received a 7U/Kg of BoNT-A. The groups received the same volume of injected substances. Animals were sacrificed on 7th (n = 5), 14th (n = 5), 21st (n = 5), 28th (n = 4) and 90th (n = 3) days post-treatment. Histological masseter tissue slides were obtained from hematoxylin-eosin treatment and analyzed in optical microscopy regarding muscle cross-sectional area, amount of connective tissue and quantity and diameter of myocytes. For statistical analysis, generalized linear models were used to compare the data (ANOVA). In all test, the significance level of 5% was set. BoNT-A values of cross-sectional area of the masseter muscle were significantly lower than controls (p < 0.01) throughout the study. Regarding myocytes quantity, BoNT-A subgroups presented higher values than controls (p < 0.0001) since the 14th day until the end of the study; however, the diameter of myocytes was smaller in all BoNT-A subgroups (p < 0.0001) in all assessment points. The amount of connective tissue was higher in BoNT-A subgroups (p < 0.0001) throughout the study. A single injection of BoNT-A altered the structure of masseter muscle of rats, regarding its histomorphometric parameters. This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .

A 3-dimensional histology computer model of malignant melanoma and its implications for digital pathology

BackgroundHistorically, cancer diagnoses have been made by pathologists using two-dimensional histological slides. However, with the advent of digital pathology and artificial intelligence, slides are being digitised, providing new opportunities to integrate their information. Since nature is 3-dimensional (3D), it seems intuitive to digitally reassemble the 3D structure for diagnosis. ObjectiveTo develop the first human-3D-melanoma-histology-model with full data and code availability. Further, to evaluate the 3D-simulation together with experienced pathologists in the field and discuss the implications of digital 3D-models for the future of digital pathology. MethodsA malignant melanoma of the skin was digitised via 3 µm cuts by a slide scanner; an open-source software was then leveraged to construct the 3D model. A total of nine pathologists from four different countries with at least 10 years of experience in the histologic diagnosis of melanoma tested the model and discussed their experiences as well as implications for future pathology. ResultsWe successfully constructed and tested the first 3D-model of human melanoma. Based on testing, 88.9% of pathologists believe that the technology is likely to enter routine pathology within the next 10 years; advantages include a better reflectance of anatomy, 3D assessment of symmetry and the opportunity to simultaneously evaluate different tissue levels at the same time; limitations include the high consumption of tissue and a yet inferior resolution due to computational limitations. Conclusions3D-histology-models are promising for digital pathology of cancer and melanoma specifically, however, there are yet limitations which need to be carefully addressed.

Mueller matrix polarization parameters correlate with local recurrence in patients with stage III colorectal cancer

The peri-tumoural stroma has been explored as a useful source of prognostic information in colorectal cancer. Using Mueller matrix (MM) polarized light microscopy for quantification of unstained histology slides, the current study assesses the prognostic potential of polarimetric characteristics of peri-tumoural collagenous stroma architecture in 38 human stage III colorectal cancer (CRC) patient samples. Specifically, Mueller matrix transformation and polar decomposition parameters were tested for association with 5-year patient local recurrence outcomes. The results show that some of these polarimetric parameters were significantly different (p value < 0.05) for the recurrence versus the no-recurrence patient cohorts (Mann–Whitney U test). MM parameters may thus be prognostically valuable towards improving clinical management/treatment stratification in CRC patients.

Privacy-preserving continual learning methods for medical image classification: a comparative analysis.

The implementation of deep learning models for medical image classification poses significant challenges, including gradual performance degradation and limited adaptability to new diseases. However, frequent retraining of models is unfeasible and raises concerns about healthcare privacy due to the retention of prior patient data. To address these issues, this study investigated privacy-preserving continual learning methods as an alternative solution. We evaluated twelve privacy-preserving non-storage continual learning algorithms based deep learning models for classifying retinal diseases from public optical coherence tomography (OCT) images, in a class-incremental learning scenario. The OCT dataset comprises 108,309 OCT images. Its classes include normal (47.21%), drusen (7.96%), choroidal neovascularization (CNV) (34.35%), and diabetic macular edema (DME) (10.48%). Each class consisted of 250 testing images. For continuous training, the first task involved CNV and normal classes, the second task focused on DME class, and the third task included drusen class. All selected algorithms were further experimented with different training sequence combinations. The final model's average class accuracy was measured. The performance of the joint model obtained through retraining and the original finetune model without continual learning algorithms were compared. Additionally, a publicly available medical dataset for colon cancer detection based on histology slides was selected as a proof of concept, while the CIFAR10 dataset was included as the continual learning benchmark. Among the continual learning algorithms, Brain-inspired-replay (BIR) outperformed the others in the continual learning-based classification of retinal diseases from OCT images, achieving an accuracy of 62.00% (95% confidence interval: 59.36-64.64%), with consistent top performance observed in different training sequences. For colon cancer histology classification, Efficient Feature Transformations (EFT) attained the highest accuracy of 66.82% (95% confidence interval: 64.23-69.42%). In comparison, the joint model achieved accuracies of 90.76% and 89.28%, respectively. The finetune model demonstrated catastrophic forgetting in both datasets. Although the joint retraining model exhibited superior performance, continual learning holds promise in mitigating catastrophic forgetting and facilitating continual model updates while preserving privacy in healthcare deep learning models. Thus, it presents a highly promising solution for the long-term clinical deployment of such models.

Prediction of lymph node metastasis in primary gastric cancer from pathological images and clinical data by multimodal multiscale deep learning

Background:Gastric cancer is the third most common cause of cancer-related death. Accurate preoperative prediction of lymph node metastasis (LNM) in primary gastric cancer strongly influences the choice of surgical approach and the prognosis of gastric cancer patients. Objective:To develop and validate a deep learning-based model to analyze routine histological slides of patients with primary gastric cancer as well as clinical data to predict the occurrence of LNM preoperatively. Patients and Methods:Radical surgery slides from 309 patients and biopsy slides from 157 patients were collected from Fujian Cancer Hospital, along with radical surgery slides from 306 patients from The Cancer Genome Atlas (TCGA). Clinical data, including age, gender, lauren classification, and tumor location, were collected. These datasets were used to develop and validate a deep learning-based model. Results:Five models were trained via cross-validation, with a mean area under the receiver operating characteristic curve (AUC) (standard deviation [SD]) of 0.877 (0.048) achieved. There was a significant difference in scores between both classes (LNM positive [N+] and LNM negative [N0]) ( p<0.001). we validated the performance of the model on biopsy slides and achieved a mean AUC (SD) of 0.725 (0.020). In the analysis of clinical data, the lauren classification was demonstrated to be an independent risk factor for predicting LNM. Conclusion:Our study confirmed that deep learning-based image analysis could preoperatively predict LNM in patients with primary gastric cancer, combining histological slides at different magnification scales and relevant clinical data, showing superiority over individual modality prediction.

Сell clusters isolation in glioblastomas and their functional and molecular characterization using new morphometric approaches

BackgroundDigital pathology has come a long way in terms of creating tools to improve existing diagnostic approaches. However, several pathology fields, such as neuropathology, are still characterized by low coverage from machine learning tools and neural network analysis, which may be due to the complexity of the internal cellular and molecular structure of the corresponding neoplasms, including glioblastomas. MethodIn the framework of this study, using advanced proprietary tools for obtaining images of histological slides and their deep morphometric analysis, we studied samples of 198 patients with glioblastoma with the selection of morphometric cell clusters. Also, cells of each cluster were isolated, and their proliferative, migratory, invasive activity, survival ability, aerobic glycolysis activity, and chemo- and radioresistance were studied. ResultsFour morphometric clusters were identified, including small-cell cluster, paracirculonuclear cluster, hypochromic cluster, and macronuclear cluster, which significantly differed in morphometric parameters and functional parameters. Hypochromic cluster cells demonstrated the highest proliferation activity; macronuclear cluster was the most active glucose consumer; paracirculonuclear cluster had the most prominent migratory and invasive activity and hypoxia resistance; small-cell cluster demonstrated predominantly average values of all parameters. Moreover, additional analysis revealed the presence of a separate subcluster of stem cell elements that correspond in their molecular properties to glioma stem cells and are present in all four clusters. It also turned out that several key molecular parameters of glioblastoma, such as mutational modifications in the EGFR, PDGFRA, and NF1 genes, along with the molecular GBM subtype, are significantly correlated with the identified cell clusters. ConclusionsThus, the results represent an up-and-coming innovation in the practical field of digital pathology and fundamental questions of glioma carcinogenesis.

Multiparameter interferometric polarization-enhanced imaging differentiates carcinoma in situ from inflammation of the bladder: an ex vivo study.

Successful differentiation of carcinoma in situ (CIS) from inflammation in the bladder is key to preventing unnecessary biopsies and enabling accurate therapeutic decisions. Current standard-of-care diagnostic imaging techniques lack the specificity needed to differentiate these states, leading to false positives. We introduce multiparameter interferometric polarization-enhanced (MultiPIPE) imaging as a promising technology to improve the specificity of detection for better biopsy guidance and clinical outcomes. In this ex vivo study, we extract tissue attenuation-coefficient-based and birefringence-based parameters from MultiPIPE imaging data, collected with a bench-top system, to develop a classifier for the differentiation of benign and CIS tissues. We also analyze morphological features from second harmonic generation imaging and histology slides and perform imaging-to-morphology correlation analysis. MultiPIPE enhances specificity to differentiate CIS from benign tissues by nearly 20% and reduces the false-positive rate by more than four-fold over clinical standards. We also show that the MultiPIPE measurements correlate well with changes in morphological features in histological assessments. The results of our study show the promise of MultiPIPE imaging to be used for better differentiation of bladder inflammation from flat tumors, leading to a fewer number of unnecessary procedures and shorter operating room (OR) time.

Histopathology of chironomids exposed to fly ash and microplastics as a new biomarker of ecotoxicological assessment

In the last few decades, industrial pollution has gained extensive attention in terms of its effect on the aquatic environment. This imposes the need to develop sensitive biomarkers for early detection of pollutant toxicity in ecotoxicological assessment. The advantages of histopathological biomarkers are many, including quick reaction to the presence of contaminants, and the small number of individuals needed for efficient analysis. The present study analyzed the negative effect of lignite coal fly ash (LCFA) and microplastic particles (MPs) on Chironomus riparius, a suggested model organism by the Organization for Economic Cooperation and Development (OECD). This study aimed to perform histological analyses of larval tissues and target potential changes in treated groups that could serve as promising histopathological biomarkers of the contaminant's negative effects. Following that, other known sensitive sub-organismal biomarkers were analyzed and paired with the histopathological ones. Histological analysis of larvae showed a significantly decreased length of microvilli in midgut regions II and III in both treatments. Treatments with MPs affected oxidative stress parameters: thiobarbituric acid reactive substances (TBARS), advanced oxidation protein products (AOPP), superoxide dismutase (SOD), and hemoglobin levels, while LCFA significantly affected all tested sub-organismal biomarkers (DNA damage, levels of AOPP, SOD, and hemoglobin), except catalase (CAT) and TBARS. When observing histological slides, a significant shortage of brush border length in the posterior parts of the midgut was detected in all treatments. In the case of LCFA, the appearance of intensive vacuolization of digestive cells with inclusions resembling apoptotic bodies, in mentioned regions was also detected. This study demonstrated high sensitivity of brush border length to the MPs and LCFA exposure, complementary to other tested sub-organismal biomarkers. Revealing the great potential of this histopathological biomarker in ecotoxicological studies contributes to the international standard ecotoxicology assessment of emerging pollutants.

A Quantitative Histologic Analysis of Oogenesis in the Flatfish Species Pleuronectes platessa as a Tool for Fisheries Management.

The following paper gives a detailed description of the oogenesis cycle for the European Plaice (Pleuronectes platessa), from oogonia to post-ovulatory follicle, including ovarian follicle and zona pellucida sizes. Noteworthy particularities were the difficulty in identifying cortical alveoli due to their very small size. Quantitative histology (stereology) on histological slides was used to determine a first size at maturity for females from the English Channel, which was found to be smaller compared to the literature (19 cm). Stereology also determined a first spawning event starting in January, with a peak in February and ongoing until March. Moreover, the use of stereology showed misclassifications for individuals categorized into a maturity phase using a macroscopic visual method. Misclassifications were found with individuals that had spawned (D) but were put under the immature (A) phase, and individuals in development (B) classified under D.

Gigapixel end-to-end training using streaming and attention.

Current hardware limitations make it impossible to train convolutional neural networks on gigapixel image inputs directly. Recent developments in weakly supervised learning, such as attention-gated multiple instance learning, have shown promising results, but often use multi-stage or patch-wise training strategies risking suboptimal feature extraction, which can negatively impact performance. In this paper, we propose to train a ResNet-34 encoder with an attention-gated classification head in an end-to-end fashion, which we call StreamingCLAM, using a streaming implementation of convolutional layers. This allows us to train end-to-end on 4-gigapixel microscopic images using only slide-level labels. We achieve a mean area under the receiver operating characteristic curve of 0.9757 for metastatic breast cancer detection (CAMELYON16), close to fully supervised approaches using pixel-level annotations. Our model can also detect MYC-gene translocation in histologic slides of diffuse large B-cell lymphoma, achieving a mean area under the ROC curve of 0.8259. Furthermore, we show that our model offers a degree of interpretability through the attention mechanism.

Deep learning predicts patients outcome and mutations from digitized histology slides in gastrointestinal stromal tumor

Risk assessment of gastrointestinal stromal tumor (GIST) according to the AFIP/Miettinen classification and mutational profiling are major tools for patient management. However, the AFIP/Miettinen classification depends heavily on mitotic counts, which is laborious and sometimes inconsistent between pathologists. It has also been shown to be imperfect in stratifying patients. Molecular testing is costly and time-consuming, therefore, not systematically performed in all countries. New methods to improve risk and molecular predictions are hence crucial to improve the tailoring of adjuvant therapy. We have built deep learning (DL) models on digitized HES-stained whole slide images (WSI) to predict patients’ outcome and mutations. Models were trained with a cohort of 1233 GIST and validated on an independent cohort of 286 GIST. DL models yielded comparable results to the Miettinen classification for relapse-free-survival prediction in localized GIST without adjuvant Imatinib (C-index=0.83 in cross-validation and 0.72 for independent testing). DL splitted Miettinen intermediate risk GIST into high/low-risk groups (p value = 0.002 in the training set and p value = 0.29 in the testing set). DL models achieved an area under the receiver operating characteristic curve (AUC) of 0.81, 0.91, and 0.71 for predicting mutations in KIT, PDGFRA and wild type, respectively, in cross-validation and 0.76, 0.90, and 0.55 in independent testing. Notably, PDGFRA exon18 D842V mutation, which is resistant to Imatinib, was predicted with an AUC of 0.87 and 0.90 in cross-validation and independent testing, respectively. Additionally, novel histological criteria predictive of patients’ outcome and mutations were identified by reviewing the tiles selected by the models. As a proof of concept, our study showed the possibility of implementing DL with digitized WSI and may represent a reproducible way to improve tailoring therapy and precision medicine for patients with GIST.

Comparative study of ovarian development in wild and captive-reared long-whiskered Sperata aor (Hamilton, 1822)

Long-whiskered catfish Sperata aor is a freshwater catfish known for its supreme flesh quality and fast growth, whose captive-reared broodstock denotes a difficult challenge for aquaculture. The reproductive dysfunctions in long-whiskered catfish raised in tank conditions were observed by comparing tissue biochemical composition and ovarian histology of wild female broodstock. Sixty (60) female broodstocks were used in the current study, consisting of 30 reared at sandy-muddy soil tank bottoms in captive conditions and 30 wild individuals collected from the haor basin during the breeding season. The fish reproductive state was investigated using the biometric and reproductive parameters, biochemical composition and levels of amino acids in the different tissues, and histological analysis of ovarian development. Results revealed that the biometrical parameters of wild and captive female broodstocks exhibited no remarkable difference (p > 0.05). Nevertheless, the wild fish had remarkably higher (p < 0.05) GSI (8.73%), oocyte weight (0.45 mg/egg), and ripeness (27.08%) in comparison with captive-reared broodstock. The total length and body weight, body weight and ovary weight, ovipositor diameter and ovary weight, and GSI and HSI displayed a positive relationship with R2 = 1, R2 = 1, R2 = 0.993, and R2 = 0.973, respectively, for wild broodstock, while R2 = 0.994, R2 = 0.806, R2 = 0.804, and R2 = 0.896, respectively, for captive broodstock. Additionally, the proximate composition in oocytes and liver tissues in both broodstocks did not differ significantly (p > 0.05). However, two essential amino acids (EAA), i.e., lysine and phenylalanine, and two non-essential amino acids, i.e., glutamic acid and glycine, were highly significant differences (p < 0.05) in the oocytes and liver of wild broodstock compared to the captive-reared broodstock. On the other hand, the EAA, e.g., isoleucine, threonine, leucine, and arginine, were highly dominated in both wild and captive female brood oocytes and liver. The ovarian histological slides from each fish group showed three oocytes developmental stages that indicated the asynchronous-reproductive ovarian oocytes of this fish. This study may be useful to fully understand the factors affecting the spawning and reproduction of S. aor broodstock, crucial for management in captive conditions as well as conservation and protection for sustainable aquaculture management of S. aor.

Malignant Mesothelioma subtyping via sampling driven multiple instance prediction on tissue image and cell morphology data

Malignant Mesothelioma is a difficult to diagnose and highly lethal cancer usually associated with asbestos exposure. It can be broadly classified into three subtypes: Epithelioid, Sarcomatoid, and a hybrid Biphasic subtype in which significant components of both of the previous subtypes are present. Early diagnosis and identification of the subtype informs treatment and can help improve patient outcome. However, the subtyping of malignant mesothelioma, and specifically the recognition of transitional features from routine histology slides has a high level of inter-observer variability.In this work, we propose an end-to-end multiple instance learning (MIL) approach for malignant mesothelioma subtyping. This uses an adaptive instance-based sampling scheme for training deep convolutional neural networks on bags of image patches that allows learning on a wider range of relevant instances compared to max or top-N based MIL approaches. We also investigate augmenting the instance representation to include aggregate cellular morphology features from cell segmentation. The proposed MIL approach enables identification of malignant mesothelial subtypes of specific tissue regions. From this a continuous characterisation of a sample according to predominance of sarcomatoid vs epithelioid regions is possible, thus avoiding the arbitrary and highly subjective categorisation by currently used subtypes. Instance scoring also enables studying tumor heterogeneity and identifying patterns associated with different subtypes. We have evaluated the proposed method on a dataset of 234 tissue micro-array cores with an AUROC of 0.89±0.05 for this task. The dataset and developed methodology is available for the community at: https://github.com/measty/PINS.

Automated Detection and Scoring of Tumor-Infiltrating Lymphocytes in Breast Cancer Histopathology Slides.

Detection of tumor-infiltrating lymphocytes (TILs) in cancer images has gained significant importance as these lymphocytes can be used as a biomarker in cancer detection and treatment procedures. Our goal was to develop and apply a TILs detection tool that utilizes deep learning models, following two sequential steps. First, based on the guidelines from the International Immuno-Oncology Biomarker Working Group (IIOBWG) on Breast Cancer, we labeled 63 large pathology imaging slides and annotated the TILs in the stroma area to create the dataset required for model development. In the second step, various machine learning models were employed and trained to detect the stroma where U-Net deep learning structure was able to achieve 98% accuracy. After detecting the stroma area, a Mask R-CNN model was employed for the TILs detection task. The R-CNN model detected the TILs in various images and was used as the backbone analysis network for the GUI development of the TILs detection tool. This is the first study to combine two deep learning models for TILs detection at the cellular level in breast tumor histopathology slides. Our novel approach can be applied to scoring TILs in large cancer slides. Statistical analysis showed that the output of the implemented approach had 95% concordance with the scores assigned by the pathologists, with a p-value of 0.045 (n = 63). This demonstrated that the results from the developed software were statistically meaningful and highly accurate. The implemented approach in analyzing whole tumor histology slides and the newly developed TILs detection tool can be used for research purposes in biomedical and pathology applications and it can provide researchers and clinicians with the TIL score for various input images. Future research using additional breast cancer slides from various sources for further training and validation of the developed models is necessary for more inclusive, rigorous, and robust clinical applications.

Wide-field optical coherence tomography for microstructural analysis of key tissue types: a proof-of-concept evaluation.

Introduction: The presence of positive margins following tumor resection is a frequent cause of re-excision surgery. Nondestructive, real-time intraoperative histopathological imaging methods may improve margin status assessment at the time of surgery; optical coherence tomography (OCT) has been identified as a potential solution but has not been tested with the most common tissue types in surgical oncology using a single, standardized platform. Methods: This was a proof-of-concept evaluation of a novel device that employs wide-field OCT (WF-OCT; OTIS 2.0 System) to image tissue specimens. Various cadaveric tissues were obtained from a single autopsy and were imaged with WF-OCT then processed for permanent histology. The quality and resolution of the WF-OCT images were evaluated and compared to histology and with images in previous literature. Results: A total of 30 specimens were collected and tissue-specific microarchitecture consistent with previous literature were identified on both WF-OCT images and histology slides for all specimens, and corresponding sections were correlated. Application of vacuum pressure during scanning did not affect specimen integrity. On average, specimens were scanned at a speed of 10.3s/cm2 with approximately three features observed per tissue type. Conclusion: The WF-OCT images captured in this study displayed the key features of the most common human tissue types encountered in surgical oncology with utility comparable to histology, confirming the utility of an FDA-cleared imaging platform. With further study, WF-OCT may have the potential to bridge the gap between the immediate information needs of the operating room and the longer timeline inherent to histology workflow.

OralNet: Fused Optimal Deep Features Framework for Oral Squamous Cell Carcinoma Detection.

Humankind is witnessing a gradual increase in cancer incidence, emphasizing the importance of early diagnosis and treatment, and follow-up clinical protocols. Oral or mouth cancer, categorized under head and neck cancers, requires effective screening for timely detection. This study proposes a framework, OralNet, for oral cancer detection using histopathology images. The research encompasses four stages: (i) Image collection and preprocessing, gathering and preparing histopathology images for analysis; (ii) feature extraction using deep and handcrafted scheme, extracting relevant features from images using deep learning techniques and traditional methods; (iii) feature reduction artificial hummingbird algorithm (AHA) and concatenation: Reducing feature dimensionality using AHA and concatenating them serially and (iv) binary classification and performance validation with three-fold cross-validation: Classifying images as healthy or oral squamous cell carcinoma and evaluating the framework's performance using three-fold cross-validation. The current study examined whole slide biopsy images at 100× and 400× magnifications. To establish OralNet's validity, 3000 cropped and resized images were reviewed, comprising 1500 healthy and 1500 oral squamous cell carcinoma images. Experimental results using OralNet achieved an oral cancer detection accuracy exceeding 99.5%. These findings confirm the clinical significance of the proposed technique in detecting oral cancer presence in histology slides.