Oral Dysplasia Research Articles

Exploring the Efficacy of Various CNN Architectures in Diagnosing Oral Cancer from Squamous Cell Carcinoma

Oral cancer can result from mutations in cells located in the lips or mouth. Diagnosing oral cavity squamous cell carcinoma (OCSCC) is particularly challenging, often occurring at advanced stages. To address this, computer-aided diagnosis methods are increasingly being used. In this work, a deep learning-based approach utilizing models such as VGG16, ResNet50, LeNet-5, MobileNetV2, and Inception V3 is presented. NEOR and OCSCC datasets were used for feature extraction, with virtual slide images divided into tiles and classified as normal or squamous cell cancer. Performance metrics like accuracy, F1-score, AUC, precision, and recall were analyzed to determine the prerequisites for optimal CNN performance. The proposed CNN approaches were effective for classifying OCSCC and oral dysplasia, with the highest accuracy of 95.41% achieved using MobileNetV2. Key findingsDeep learning models, particularly MobileNetV2, achieved high classification accuracy (95.41%) for OCSCC.CNN-based methods show promise for early-stage OCSCC and oral dysplasia diagnosis. Performance parameters like precision, recall, and F1-score help optimize CNN model selection for this task.

Artificial Intelligence Applications in Oral Cancer and Oral Dysplasia.

Oral squamous cell carcinoma (OSCC) is a highly unpredictable disease with devastating mortality rates that have not changed over the past decades, in the face of advancements in treatments and biomarkers, which have improved survival for other cancers. Delays in diagnosis are frequent, leading to more disfiguring treatments and poor outcomes for patients. The clinical challenge lies in identifying those patients at the highest risk of developing OSCC. Oral epithelial dysplasia (OED) is a precursor of OSCC with highly variable behavior across patients. There is no reliable clinical, pathological, histological, or molecular biomarker to determine individual risk in OED patients. Similarly, there are no robust biomarkers to predict treatment outcomes or mortality in OSCC patients. This review aims to highlight advancements in artificial intelligence (AI)-based methods to develop predictive biomarkers of OED transformation to OSCC or predictive biomarkers of OSCC mortality and treatment response. Biomarkers such as S100A7 demonstrate promising appraisal for the risk of malignant transformation of OED. Machine learning-enhanced multiplex immunohistochemistry workflows examine immune cell patterns and organization within the tumor immune microenvironment to generate outcome predictions in immunotherapy. Deep learning (DL) is an AI-based method using an extended neural network or related architecture with multiple "hidden" layers of simulated neurons to combine simple visual features into complex patterns. DL-based digital pathology is currently being developed to assess OED and OSCC outcomes. The integration of machine learning in epigenomics aims to examine the epigenetic modification of diseases and improve our ability to detect, classify, and predict outcomes associated with epigenetic marks. Collectively, these tools showcase promising advancements in discovery and technology, which may provide a potential solution to addressing the current limitations in predicting OED transformation and OSCC behavior, both of which are clinical challenges that must be addressed in order to improve OSCC survival.

EMT-cancer stem cell expression in oral dysplasia and cancer

EMT-cancer stem cell expression in oral dysplasia and cancer

Fusobacterium nucleatum induces invasive growth and angiogenic responses in malignant oral keratinocytes that are cell line- and bacterial strain-specific.

Fusobacterium nucleatum is an anaerobic commensal of the oral cavity recently reported to be associated with cancers of the gastrointestinal tract and oral squamous cell carcinoma (OSCC). In this study, we investigate the impact on oral keratinocytes of infection with a genetically diverse set of strains of F. nucleatum subsp. polymorphum recovered from patients with oral dysplasia (n=6). We employed H357 oral keratinocytes derived from a stage 1 OSCC and H376 cells derived from a stage 3 OSCC. Adhesion phenotypes were strain specific, with 3/6 clinical isolates examined exhibiting higher adherence to the stage 3 H376 cell line. Conversely, intracellular invasion was greatest in the H357 cells and was associated with specific transcriptional responses including autophagy and keratinization. Infection of both H357 and H376 cell lines induced transcriptional and cytokine responses linked to cancer cell migration and angiogenesis. F. nucleatum infection induced greater levels of MMP9 secretion in the H376 cell line which was associated with enhanced motility and invasion phenotypes. Additionally, the degree of F. nucleatum induced invasive growth by H376 cells varied between different clinical isolates of F. nucleatum subsp. polymorphum. Blockage of CCL5 signalling using the inhibitor metCCL5 resulted in reduced keratinocyte invasion. F. nucleatum infection also induced expression of the pro-angiogenic chemokine MCP-1 and the angiogenic growth factor VEGF-A resulting in increased capillary-like tube formation in HUVEC cells, most significantly in H376 cells. Treatment of HUVEC cells with resveratrol, a VEGF-A signalling inhibitor, significantly attenuated F. nucleatum induced tube formation. Our data indicate that the outcomes of F. nucleatum-oral cell interactions can vary greatly depending on the bacterial genotype and the malignant phenotype of the host cell.

Immunohistochemical Verification of Oral Dysplasia, Premalignant Lesions and Oral Cancer by Use of Varied Expression of Cytokeratins; A Review

Oral cancer, predominantly squamous cell carcinoma, carries high morbidity and disease burden all over the world. Evaluation of the cellular basis of oral carcinogenesis has implied dysregulation in cytokeratins as one of the contributory factors. These intermediate filament proteins of the oral epithelium, guardian of cellular architecture, are known to maintain cellular interactions and are involved in various cell cycle regulation pathways. Oral premalignancy and neoplastic progression are often depicted by disturbed and/or haphazard expression of these cytokeratins. We present a summarized analysis of cytokeratins and explore their diagnostic potential using immunohistochemistry. CK 8/18 is a significant and reliable biomarker of oral precancer and cancer that can be utilized in addition to histopathology for early malignancy screening and intervention. Moreover, dysplastic oral lesions are attributed to downregulation and/or gradual disappearance of CK 4/13, CK 5/14, and expression of CK 17. Increasing the diagnostic strategy using immunohistochemical techniques on specific cytokeratins and extensive sample studies on premalignant lesions, dysplasia, and oral squamous cell carcinoma can open up many possibilities. Utilizing the enhanced diagnostic spectrum of specific cytokeratins can help clinicians and dental specialists diagnose early, thus timely managing and alleviating suffering associated with oral cancer.

Imaging Biomarkers of Oral Dysplasia and Carcinoma Measured with In Vivo Endoscopic Optical Coherence Tomography.

Optical coherence tomography is a noninvasive imaging technique that provides three-dimensional visualization of subsurface tissue structures. OCT has been proposed and explored in the literature as a tool to assess oral cancer status, select biopsy sites, or identify surgical margins. Our endoscopic OCT device can generate widefield (centimeters long) imaging of lesions at any location in the oral cavity-but it is challenging for raters to quantitatively assess and score large volumes of data. Leveraging a previously developed epithelial segmentation network, this work develops quantifiable biomarkers that provide direct measurements of tissue properties in three dimensions. We hypothesize that features related to morphology, tissue attenuation, and contrast between tissue layers will be able to provide a quantitative assessment of disease status (dysplasia through carcinoma). This work retrospectively assesses seven biomarkers on a lesion-contralateral matched OCT dataset of the lateral and ventral tongue (40 patients, 70 sites). Epithelial depth and loss of epithelial-stromal boundary visualization provide the strongest discrimination between disease states. The stroma optical attenuation coefficient provides a distinction between benign lesions from dysplasia and carcinoma. The stratification biomarkers visualize subsurface changes, which provides potential for future utility in biopsy site selection or treatment margin delineation.

Characterizing oral dysplasia immune microenvironment (DIME) using imaging mass cytometry

Characterizing oral dysplasia immune microenvironment (DIME) using imaging mass cytometry

Unraveling the Keratin Expression in Oral Leukoplakia: A Scoping Review.

Intermediate filaments are one of three polymeric structures that form the cytoskeleton of epithelial cells. In the epithelium, these filaments are made up of a variety of keratin proteins. Intermediate filaments complete a wide range of functions in keratinocytes, including maintaining cell structure, cell growth, cell proliferation, cell migration, and more. Given that these functions are intimately associated with the carcinogenic process, and that hyperkeratinization is a quintessential feature of oral leukoplakias, the utility of keratins in oral leukoplakia is yet to be fully explored. This scoping review aims to outline the current knowledge founded on original studies on human tissues regarding the expression and utility of keratins as diagnostic, prognostic, and predictive biomarkers in oral leukoplakias. After using a search strategy developed for several scientific databases, namely, PubMed, Scopus, Web of Science, and OVID, 42 papers met the inclusion and exclusion criteria. One more article was added when it was identified through manually searching the list of references. The included papers were published between 1989 and 2024. Keratins 1-20 were investigated in the 43 included studies, and their expression was assessed in oral leukoplakia and dysplasia cases. Only five studies investigated the prognostic role of keratins in relation to malignant transformation. No studies evaluated keratins as a diagnostic adjunct or predictive tool. Evidence supports the idea that dysplasia disrupts the terminal differentiation pathway of primary keratins. Gain of keratin 17 expression and loss of keratin 13 were significantly observed in differentiated epithelial dysplasia. Also, the keratin 19 extension into suprabasal cells has been associated with the evolving features of dysplasia. The loss of keratin1/keratin 10 has been significantly associated with high-grade dysplasia. The prognostic value of cytokeratins has shown conflicting results, and further studies are required to ascertain their role in predicting the malignant transformation of oral leukoplakia.

Stabilized oral lichen planus does not compromise dental implants survival: A systematic review and meta-analysis.

This study aimed to evaluate the incidence of implant failure in patients with oral lichen planus (OLP) and investigate the potential association between OLP and peri-implant diseases. Embase, Web of Science, PubMed, and Scopus databases were searched for studies with no time restrictions. Meta-analysis was performed calculating pooled proportion of peri-implantitis (PI), peri-implant mucositis (PIM), and bleeding on probing (BOP) prevalence using fixed-effects model. Odds ratio and corresponding 95% CI were calculated to assess the potential risk of PI, PIM, and BOP in dental implant patients with OLP compared to healthy controls. Implant failure rate was 4.38% at the patient level and 4.37% at the implant level. Six patients (3.92%) from five studies were diagnosed with oral cancer after receiving implant. The prevalence of PI, PIM, and BOP at the implant level were 14.00%, 20.00%, and 40.00%, respectively. There was no significant difference in the occurrence of PI and PIM between OLP patients and healthy controls. Stabilized OLP is not considered a significant risk factor for peri-implant diseases. It is advised against placing implants or prostheses during the acute phase of the disease. Histopathological investigation to differentiate OLP from oral lichenoid dysplasia is crucial.

An Update on the Use of Artificial Intelligence in Digital Pathology for Oral Epithelial Dysplasia Research.

Oral epithelial dysplasia (OED) is a precancerous histopathological finding which is considered the most important prognostic indicator for determining the risk of malignant transformation into oral squamous cell carcinoma (OSCC). The gold standard for diagnosis and grading of OED is through histopathological examination, which is subject to inter- and intra-observer variability, impacting accurate diagnosis and prognosis. The aim of this review article is to examine the current advances in digital pathology for artificial intelligence (AI) applications used for OED diagnosis. We included studies that used AI for diagnosis, grading, or prognosis of OED on histopathology images or intraoral clinical images. Studies utilizing imaging modalities other than routine light microscopy (e.g., scanning electron microscopy), or immunohistochemistry-stained histology slides, or immunofluorescence were excluded from the study. Studies not focusing on oral dysplasia grading and diagnosis, e.g., to discriminate OSCC from normal epithelial tissue were also excluded. A total of 24studies were included in this review. Nineteen studies utilized deep learning (DL) convolutional neural networks for histopathological OED analysis, and 4 used machine learning (ML) models. Studies were summarized by AI method, main study outcomes, predictive value for malignant transformation, strengths, and limitations. ML/DL studies for OED grading and prediction of malignant transformation are emerging as promising adjunctive tools in the field of digital pathology. These adjunctive objective tools can ultimately aid the pathologist in more accurate diagnosis and prognosis prediction. However, further supportive studies that focus on generalization, explainable decisions, and prognosis prediction are needed.

Analysis of differentially expressed genes in dysplastic oral keratinocyte cell line and their role in the development of HNSCC

Dysplasia is the presence of abnormal types of cells in a tissue precipitated by over or diminished expression of certain genes. These cells act as a precursor to cancer. Dysplastic oral keratinocyte (DOK) cell lines have an aneuploid complex karyotype. They provide an opportunity to study the action of specific carcinogens on malignant transformations. This study aimed to identify the differentially expressed genes in dysplastic cells and their possible association with head and neck squamous cell carcinoma (HNSCC). These genes can be developed as diagnostic, prognostic, or therapeutic leads.The list of genes related to oral keratinocyte dysplasia and head and neck cancer was accessed from the GEO (Gene Expression Omnibus) database. Gene expression profiling was done between dysplastic oral keratinocytes and normal human oral keratinocytes. Gene expression and Kaplan Meier survival analysis were performed using the UALCAN database to assess the correlations between dysregulated genes identified in dysplastic keratinocytes and primary tumors of HNSCC. The GEO omnibus dataset identified numerous differentially expressed genes of which the top 10 up and downregulated genes in dysplastic oral keratinocytes were curated for further analysis. The expression profile of these genes was assessed using the HNSCC dataset (TCGA, Firehose Legacy). Among all the genes assessed, only one gene, the OLR1 gene encoding oxidized low-density lipoprotein, was found to be overexpressed in both the groups viz., dysplastic keratinocytes and HNSCC cases with a strong correlation with the survival status of patients. There was significant correlation between the gene expression pattern observed in dysplastic keratinocytes and the primary tumor of the HNSCC group, with an exotic gene that was seldom discussed in association with cancer, viz., OLR1. Exploration into other top-ranking differentially expressed genes in dysplastic cases would aid in identifying the candidate gene associated with both phenotypes.

Calcitriol Treatment Decreases Cell Migration, Viability and β-Catenin Signaling in Oral Dysplasia.

Nearly 90% of oral cancers are characterized as oral squamous cell carcinoma (OSCC), representing the sixth most common type of cancer. OSCC usually evolves from oral potentially malignant disorders that, in some cases, are histologically consistent with a oral dysplasia. The levels of 1α,25 dihydroxyvitamin D3 (1,25-(OH)2D3; calcitriol), the active form of vitamin D3, have been shown to be decreased in patients with oral dysplasia and OSCC. Moreover, treatment with 1,25-(OH)2D3 has been proven beneficial in OSCC by inhibiting the Wnt/β-catenin pathway, a signaling route that promotes cell migration, proliferation, and viability. However, whether this inhibition mechanism occurs in oral dysplasia is unknown. To approach this question, we used dysplastic oral keratinocyte cultures and oral explants (ex vivo model of oral dysplasia) treated with 1,25-(OH)2D3 for 48 h. Following treatment with 1,25-(OH)2D3, both in vitro and ex vivo models of oral dysplasia showed decreased levels of nuclear β-catenin by immunofluorescence (IF) and immunohistochemistry (IHC). Consistently, reduced protein and mRNA levels of the Wnt/β-catenin target gene survivin were observed after treatment with 1,25-(OH)2D3. Moreover, 1,25-(OH)2D3 promoted membranous localization of E-cadherin and nuclear localization of vitamin D receptor (VDR). Functionally, DOK cells treated with 1,25-(OH)2D3 displayed diminished cell migration and viability in vitro.

Knowledge, Attitude, and Practice of the Interns About Exfoliative Cytology in a Tertiary Care Dental Hospital in Bhubaneswar City: A Questionnaire-Based Survey.

The most significant accomplishment of dentists is the early detection of oral dysplasia or the rapid diagnosis of potentially malignant lesions. An additional test that assists in the diagnosis of numerous oral illnesses is exfoliative cytology. This study was conducted among dental professionals in the city of Bhubaneswar to assess their exfoliative cytology knowledge, attitudes, and practices. Fifty-eight dental interns participated in a cross-sectional, descriptive, Web-based questionnaire study. The findings reveal that while some interns were knowledgeable about some components of exfoliative cytology in oral cancer (OC) screening, they lacked a thorough understanding of other crucial elements. To enhance early diagnosis and lessen the prevalence of OC, it is critical that dental practitioners increase their understanding of and use of exfoliative cytology. To address the misconcepts found in the study, it draws attention to the necessity of patient education and public awareness campaigns. The study recommends more awareness and education among dental practitioners and emphasizes the need of tackling OC through preventive measures and appropriate diagnostic techniques such as exfoliative cytology.

Fusobacterium nucleatum subsp. polymorphum recovered from malignant and potentially malignant oral disease exhibit heterogeneity in adhesion phenotypes and adhesin gene copy number, shaped by inter-subspecies horizontal gene transfer and recombination-derived mosaicism.

Fusobacterium nucleatum is an anaerobic commensal of the oral cavity associated with periodontitis and extra-oral diseases, including colorectal cancer. Previous studies have shown an increased relative abundance of this bacterium associated with oral dysplasia or within oral tumours. Using direct culture, we found that 75 % of Fusobacterium species isolated from malignant or potentially malignant oral mucosa were F. nucleatum subsp. polymorphum. Whole genome sequencing and pangenome analysis with Panaroo was carried out on 76 F. nucleatum subsp. polymorphum genomes. F. nucleatum subsp. polymorphum was shown to possesses a relatively small core genome of 1604 genes in a pangenome of 7363 genes. Phylogenetic analysis based on the core genome shows the isolates can be separated into three main clades with no obvious genotypic associations with disease. Isolates recovered from healthy and diseased sites in the same patient are generally highly related. A large repertoire of adhesins belonging to the type V secretion system (TVSS) could be identified with major variation in repertoire and copy number between strains. Analysis of intergenic recombination using fastGEAR showed that adhesin complement is shaped by horizontal gene transfer and recombination. Recombination events at TVSS adhesin genes were not only common between lineages of subspecies polymorphum, but also between different subspecies of F. nucleatum. Strains of subspecies polymorphum with low copy numbers of TVSS adhesin encoding genes tended to have the weakest adhesion to oral keratinocytes. This study highlights the genetic heterogeneity of F. nucleatum subsp. polymorphum and provides a new framework for defining virulence in this organism.

Abstract 310: Cadherin-dependent modulation of chemosensitivity in oral tumor progression

Abstract The progression of healthy oral epithelia to an oral carcinoma often involves aberrant expression of a group of calcium-dependent transmembrane proteins called cadherins. Histological studies of human tumors have demonstrated increased expression of the epithelial-resident placental (P-cadherin) and ectopic expression of the mesenchymal neuronal (N-cadherin) cadherin in oral dysplasia. We hypothesized that these changes in cadherin expression during early oral tumor development could facilitate increases in survival signaling. We genetically modified dysplastic oral keratinocytes to overexpress P-cadherin or suppress endogenous expression of N-cadherin. P-cadherin overexpression and N-cadherin suppression independently elicited a 3-6-fold decrease in apoptosis levels in response to docetaxel treatment compared to similarly treated controls. We found that in untreated dysplastic cells, P- and N-cadherin expression had positive and negative correlations, respectively, with the activation of the survival-associated enzymes focal adhesion kinase (FAK), mitogen-activated protein kinase (MAPK), and AKT kinase. The total levels of nuclear factor kappa B (NFkb) were positively correlated with P-cadherin overexpression and reciprocally correlated with levels of N-cadherin. In docetaxel-treated dysplastic cells, we observed the reverse effect compared to untreated cells: the increased level of N-cadherin was positively correlated with FAK activity whereas FAK activity was suppressed by P-cadherin. P-cadherin expression in docetaxel-treated cells also increased activation of both MAPK and AKT, whereas alteration of N-cadherin levels had no effect. Because N-cadherin levels are known to increase in late-stage oral tumors, we extended these signaling studies to oral squamous carcinoma cells (OSCCs) expressing low or high levels of N-cadherin. In contrast to its suppression of survival pathways in docetaxel-treated dysplastic cells, in docetaxel-treated OSCCs, N-cadherin expression strongly enhanced phosphorylation of MAPK and AKT, and increased expression of NFkb. N-cadherin overexpression during this treatment also increased expression of the antiapoptotic proteins Mcl-1 and suppressed phosphorylation of the apoptotic regulator signal transducer and activator of transcription 3 (pSTAT3). Increase cadherin expression in all untreated cell lines increased expression of an inducer of epithelial to mesenchymal transition, the transmembrane protein mucin-1 (MUC1). N-cadherin overexpression also increased motility in both cell lines. Taken together, our data demonstrate antiapoptotic signaling directed by increased levels of P-cadherin in oral dysplasia and N-cadherin in OSCC, whereas increased levels of N-cadherin in dysplastic cells directed cellular signaling that were associated with a proapoptotic phenotype. Citation Format: Aaron Mody, Cassandra Decker, Kathryn R. Lawson. Cadherin-dependent modulation of chemosensitivity in oral tumor progression [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 310.

Bacteriome and mycobiome dysbiosis in oral mucosal dysplasia and oral cancer.

It has long been considered that the oral microbiome is tightly connected to oral health and that dysbiotic changes can be detrimental to the occurrence and progression of dysplastic oral mucosal lesions or oral cancer. Improved understanding of the concepts of microbial dysbiosis together with advances in high-throughput molecular sequencing of these pathologies have charted in greater microbiological detail the nature of their clinical state. This review discusses the bacteriome and mycobiome associated with oral mucosal lesions, oral candidiasis, and oral squamous cell carcinoma, aiming to delineate the information available to date in pursuit of advancing diagnostic and prognostic utilities for oral medicine.

Effect of valproic acid on histone deacetylase expression in oral cancer (Review).

Oral squamous cell carcinoma (OSCC) is a frequent human malignancy that demonstrates a range of genetic and epigenetic alterations. Histone deacetylases (HDACs) are key epigenetic regulators of cell-cycle progression, differentiation and apoptosis and their dysregulation is implicated in cancer development. HDACs are promising targets for anticancer therapy through the utilisation of HDAC inhibitors (HDACis). OSCC cells have been shown to have low levels of histone acetylation, suggesting that HDACis may produce beneficial effects in patients with OSCC. Valproic acid (VPA) is a class I and IIa HDACi and, therefore, may be useful in anticancer therapy. VPA has been reported as a chemo-preventive epigenetic agent in individuals with high-risk oral dysplasia (OD) and thus associated with a reduced risk of HNSCC. It is hypothesised that HDAC inhibition by VPA triggers a change in the expression levels of different HDAC family gene-members. The present review summarises the current literature on HDAC expression changes in response to VPA in oral cancer patients and in vitro studies in an effort to better understand the potential epigenetic impact of VPA treatment. The present review outlined the need for exploring supportive evidence of the chemo-preventive role played by VPA-based epigenetic modification in treating oral pre-cancerous lesions and, thus, providing a novel tolerable chemotherapeutic strategy for patients with oral cancer.

Identification and Evaluation of Cancer Stem Cells in Oral Squamous Cell Carcinoma and Oral Epithelial Dysplasia Using NANOG: An Immunohistochemical Study.

Squamous cell carcinoma of the oral cavity may show precursor lesions, termed as potentially malignant disorders, of which leukoplakia is the most frequent one. Oral leukoplakia is a clinical diagnosis for which the histological diagnosis may be either hyperplasia or oral epithelial dysplasia (OED) and sometimes even oral squamous cell carcinoma (OSCC). Cancer stem cells (CSCs), identified in various tumors, are a specific group of cells that exhibit the properties of self-renewal and differentiation. Among the various biomarkers that identify CSCs, the transcription factor NANOG is considered to be a significant one. In this study, we intend to identify and compare the immunohistochemical expression of NANOG in OSCC, OED, and normal oral mucosa. Tissue blocks of OSCC (n=28), OED (n=28), and normal oral mucosa (n=28) were used in this study. Specimens were immunohistochemically analyzed for NANOG expression. The results were statistically analyzed using one-way ANOVA, Games-Howell post hoc, and Student t-test. Statistical Product and Service Solutions (SPSS, version 21; IBM SPSS Statistics for Windows, Armonk, NY) software was used for performing the statistical analysis, and the level of significance was set as 0.05. NANOG expression was higher in OSCC when compared to oral dysplasias and normal oral mucosa, in decreasing order. A significantly higher histo-score and labeling index score were observed in OSCC and oral dysplasias compared to normal oral mucosa (p=<0.001). The expression levels of NANOG were positively correlated with disease progression in OSCC, implicating that NANOG can be used as a surrogate marker of oral oncogenesis and prognosis. Therefore, decoding the molecular mechanisms ofNANOG regulation in the progression of cancer helps in developing new therapeutic strategies for oral cancer.

OralEpitheliumDB: A Dataset for Oral Epithelial Dysplasia Image Segmentation and Classification.

Early diagnosis of potentially malignant disorders, such as oral epithelial dysplasia, is the most reliable way to prevent oral cancer. Computational algorithms have been used as an auxiliary tool to aid specialists in this process. Usually, experiments are performed on private data, making it difficult to reproduce the results. There are several public datasets of histological images, but studies focused on oral dysplasia images use inaccessible datasets. This prevents the improvement of algorithms aimed at this lesion. This study introduces an annotated public dataset of oral epithelial dysplasia tissue images. The dataset includes 456 images acquired from 30 mouse tongues. The images were categorized among the lesion grades, with nuclear structures manually marked by a trained specialist and validated by a pathologist. Also, experiments were carried out in order to illustrate the potential of the proposed dataset in classification and segmentation processes commonly explored in the literature. Convolutional neural network (CNN) models for semantic and instance segmentation were employed on the images, which were pre-processed with stain normalization methods. Then, the segmented and non-segmented images were classified with CNN architectures and machine learning algorithms. The data obtained through these processes is available in the dataset. The segmentation stage showed the F1-score value of 0.83, obtained with the U-Net model using the ResNet-50 as a backbone. At the classification stage, the most expressive result was achieved with the Random Forest method, with an accuracy value of 94.22%. The results show that the segmentation contributed to the classification results, but studies are needed for the improvement of these stages of automated diagnosis. The original, gold standard, normalized, and segmented images are publicly available and may be used for the improvement of clinical applications of CAD methods on oral epithelial dysplasia tissue images.

NRF2 Activation in Trp53;p16-deficient Mice Drives Oral Squamous Cell Carcinoma.

Human squamous cancers frequently show constitutive NRF2 activation, associated with poorer outcomes and resistance to multiple therapies. Here, we report the first activated NRF2-driven and human-relevant mouse model of squamous cell carcinoma that develops in the background of p16 and p53 loss. The availability of this model will lead to a clearer understanding of how NRF2 contributes to the initiation, progression, and therapeutic response of OSCC.