The use of artificial intelligence, machine learning and deep learning in oncologic histopathology

BACKGROUND

Recently, there has been a momentous drive to apply advanced artificial intelligence (AI) technologies to diagnostic medicine. The introduction of AI has provided vast new opportunities to improve health care and has introduced a new wave of heightened precision in oncologic pathology. The impact of AI on oncologic pathology has now become apparent, and its use with respect to oral oncology is still in the nascent stage.

DISCUSSION

A foundational overview of AI classification systems used in medicine and a review of common terminology used in machine learning and computational pathology will be presented. This paper provides a focused review on the recent advances in AI and deep learning in oncologic histopathology and oral oncology. In addition, specific emphasis on recent studies that have applied these technologies to oral cancer prognostication will also be discussed.

CONCLUSION

Machine and deep learning methods designed to enhance prognostication of oral cancer have been proposed with much of the work focused on prediction models on patient survival and locoregional recurrences in patients with oral squamous cell carcinomas (OSCC). Few studies have explored machine learning methods on OSCC digital histopathologic images. It is evident that further research at the whole slide image level is needed and future collaborations with computer scientists may progress the field of oral oncology.

Macronutrients, vitamins and minerals intake and risk of esophageal squamous cell carcinoma: a case-control study in Iran

Background

Although Iran is a high-risk region for esophageal squamous cell carcinoma (ESCC), dietary factors that may contribute to this high incidence have not been thoroughly studied. The aim of this study was to evaluate the effect of macronutrients, vitamins and minerals on the risk of ESCC.

Methods

In this hospital-based case-control study, 47 cases with incident ESCC and 96 controls were interviewed and usual dietary intakes were collected using a validated food frequency questionnaire. Data were modeled through unconditional multiple logistic regression to estimate odds ratios (OR) and 95% confidence intervals (CI), controlling for age, sex, gastrointestinal reflux, body mass index, smoking history (status, intensity and duration), physical activity, and education.

Results

ESCC cases consumed significantly more hot foods and beverages and fried and barbecued meals, compared to the controls (p < 0.05). After adjusting for potential confounders, the risk of ESCC increased significantly in the highest tertiles of saturated fat [OR:2.88,95%CI:1.15-3.08], cholesterol [OR:1.53, 95%CI: 1.41-4.13], discretionary calorie [OR:1.51, 95%CI: 1.06-3.84], sodium [OR:1.49,95%CI:1.12-2.89] and total fat intakes [OR:1.48, 95%CI:1.09-3.04]. In contrast, being in the highest tertile of carbohydrate, dietary fiber and (n-3) fatty acid intake reduced the ESCC risk by 78%, 71% and 68%, respectively. The most cancer-protective effect was observed for the combination of high folate and vitamin E intakes (OR: 0.02, 95%CI: 0.00-0.87; p < 0.001). Controls consumed 623.5 times higher selenium, 5.48 times as much β-carotene and 1.98 times as much α-tocopherol as the amount ESCC cases consumed.

Conclusion

This study suggests that high intake of nutrients primarily found in plant-based foods is associated with a reduced esophageal cancer risk. Some nutrients such as folate, vitamin E and selenium might play major roles in the etiology of ESCC and their status may eventually be used as an epidemiological marker for esophageal cancer in Iran, and perhaps other high-risk regions.

Next generation sequencing and its application in deciphering head and neck cancer

Head and neck squamous cell carcinoma (HNSCC) are a group of heterogeneous tumours mainly attributable to tobacco use, alcohol consumption and infection with human papillomavirus. Based on the stage of cancer at the time of diagnosis, patients are managed by surgery, radiotherapy, chemotherapy or a combination of these. Early diagnosis usually improves patient prognosis. Since their first commercial application in 2005, next generation sequencing (NGS) platforms are rapidly changing the face of basic science laboratories; however prior to progressing to clinical applications, clinicians should carefully examine currently available data and guidelines for technical and ethical matters concerning NGS. In this review, we compare various commercially available NGS platforms, with special consideration given to their clinical application in the management of HNSCC.

hMSH6: a potential diagnostic marker for oral carcinoma in situ

Oral medicine specialists rely upon accurate assessment of pathology to rationalise lesion management, especially for high-risk oral epithelial dysplasia, carcinoma in situ (CIS) and oral squamous cell carcinoma. Cross-discipline cancer research has highlighted the role of genetic instability in neoplasia. Improved diagnostic stringency from translation of immunostaining for DNA repair defects into current pathology practice has potential to benefit pathologists, clinicians and patients. The focus of this study was the obligatory and non-obligatory components of the MutLα and MutSα mismatch repair heterodimers, namely hMLH1, hMSH2, hPMS2 and hMSH6, which were studied in 274 formalin-fixed paraffin-embedded sections. A readily apparent inverse correlation between oral disease severity and both obligatory and non-obligatory components of MutLα and MutSα was observed (hMLH1, ρ=-0.715; hPMS2, ρ=-0.692; hMSH2, ρ=-0.728; and hMSH6, ρ=-0.702), with particularly conspicuous loss of hMSH6 expression from the stratum basale of CIS.

Nutrient patterns and risk of esophageal squamous cell carcinoma: a case-control study

Although Iran is a high-risk country for esophageal squamous cell carcinoma (ESCC), the contribution of overall nutrient intakes to this high incidence rate is not yet clear. The aim of this study was to examine the association between nutrient patterns and risk of ESCC in Iran. Forty-seven patients with ESCC and 96 frequency-matched hospital controls underwent private interviews, and dietary habits were collected using a validated food-frequency questionnaire. Factor analysis was conducted and two major nutrient patterns were retained; factor 1 (high in pantothenic acid, vitamin C, potassium, vitamin B(6), magnesium, folate, thiamin, copper, carbohydrate, vitamin K, niacin, α-tocopherol, zinc, total fiber, fluoride, and polyunsaturated fatty acids) and factor 2 (high in saturated fatty acid, biotin, selenium, monounsaturated fatty acids, riboflavin, sodium, fat, cholesterol, calcium, phosphorus, protein, iron, vitamin E, manganese, vitamin D, and vitamin B(12)). Factor 2 was inversely associated with ESCC (OR = 0.06, 95% CI: 0.01-0.28; P = 0.008), whereas no significant association was found for factor 1 (OR = 0.45, 95% CI: 0.11-1.82). The results of the present study suggested a possible role for a nutrient pattern similar to factor 2 in reducing the risk of ESCC.

Measuring professionalism in residency training programs in Iran

BACKGROUND AND AIM

To assess the Farsi (Persian) translated and modified version of the questionnaire of the American Board of Internal Medicine (ABIM) for measuring professionalism, and also, measuring the professional attitudes and behaviors associated with the medical residency training environment in Iran.

METHODS

After a pilot study, a 17-item questionnaire was distributed to 282 medical residents of two major universities of Iran, from December 2006 through February 2007.

RESULTS

Of the 282 distributed questionnaires, 259 (95.2%) were used in the analysis. Based on the analyses, two items were omitted and 15 items were retained for further analysis. The mean score was 106 (Standard Deviation (SD), 22.4) out of maximum 150, whereas the item mean was 6.12 (SD, 0.37) out of maximum 10. Corrected item-to-total correlations ranged from low to moderate. The internal reliability of the scale, based on Cronbach's alpha, meets Nunnally's minimal requirement. A factor analysis was performed, based on principal components and varimax rotation. The solution identified three factors (subscales) including excellence, honor/integrity and altruism/respect. Together these factors represented 58.8% of the common variance.

CONCLUSION

This study showed the content validity and internal reliability of the Farsi version of the ABIM questionnaire. Therefore, it can be considered as an encouraging step toward developing a short, reliable and valid instrument for measuring professionalism in medical environments.

Electronic nicotine delivery systems: Oral health implications and oral cancer risk

There is a paucity of evidence surrounding the potential detrimental effects of electronic nicotine delivery systems (ENDS) for both systemic and oral health. The effects of conventional cigarettes on the development of oral cancer are well known; however, the role of ENDS in oral carcinogenesis is yet to be elucidated. Furthermore, the exponential rise of the use of ENDS by the general public means that dental healthcare providers are more likely to encounter questions on their safety in the oral cavity, and on their effectiveness as a smoking cessation aid. Herein, we review the most up to date literature on the systemic and oral health complications of ENDS. Moreover, evidence-based recommendations on the use of ENDS as a smoking cessation tool within the dental setting are discussed.

Harnessing massively parallel sequencing in personalized head and neck oncology

Advances in the management of patients with head and neck squamous cell carcinoma (HNSCC) have not significantly changed the prognosis of this tumor over the past five decades. Molecular heterogeneity of HNSCC and its association with HPV, in addition to the increase in the number of cancers arising in traditionally low-risk patients, are among some of the obstacles to the successful management of this group of tumors. Massively parallel sequencing, otherwise known as next-generation sequencing (NGS), is rapidly changing conventional patient management by providing detailed information about each patient's genome and transcriptome. Despite major advances in technology and a significant reduction in the cost of sequencing, NGS remains mainly limited to research facilities. In addition, there are only a few published studies that have utilized this technology in HNSCC. This paper aims to report briefly on current commercially available NGS platforms and discuss their clinical applications, ethical considerations, and utilization in personalized patient care, particularly as this relates to head and neck cancer.

Central mechanisms of pain in orofacial pain patients: Implications for management

BACKGROUND

Central sensitization (CS) is a form of neuroplasticity characterized by changes in the neural sensitivity, responsiveness, and/or output that are not contingent on peripheral input nor activity-dependent. CS is characterized by activation of unmyelinated C-fibers resulting in a cascade of events at molecular and cellular levels which eventuate into generation of synaptic currents at rest. CS, therefore, contributes to heightened generalized pain sensitivity, further complicates the process of reaching a diagnosis, and increases the possibility of treatment failure. BODY: Trigeminal nerve is the main sensory supplier of the anterior part of the head, including the intraoral structures. Primary afferent nociceptors of the trigeminal nerve and low threshold mechanoreceptors synapse with wide dynamic range (WDR) neurons in the pons. This multifaceted network of nerve interactions which is further complicated by the modulatory circuits that can suppress or heighten the activity of WDR neurons is one of the main contributors to CS. The importance of CS in orofacial pain disorders is emphasized in the context of chronic pain development. As for all chronic pain conditions, it is crucial to consider the biopsychosocial aspects of chronic orofacial pain in managing this diverse group of conditions. This review highlights current understanding of the biopsychosocial model and central mechanisms contributing to the pathogenesis of chronic orofacial pain.

Deficient double-strand break repair in oral squamous cell carcinoma cell lines

BACKGROUND

Approximately 20% of oral squamous cell carcinoma (OSCC) cases arise without any identifiable environmental cause, suggesting involvement of genetic influences in their aetiology. DNA double-strand breaks (DSBs) sever both strands of DNA and pose a potential threat to genomic integrity. A hastened accumulation of somatic mutations consequent to DSB repair is deemed to be a likely event in tumorigenesis of OSCC.

METHODS

Two discrete chemical approaches, namely hydrogen peroxide and camptothecin, were used to induce DSB in oral cell lines derived from normal through dysplastic to OSCC tissues. After optimization, gamma histone 2Ax (γH2Ax) foci were counted as an indirect measure of kinetics of DSB and confirmed with Western blot of γH2Ax, Nbs1 and ATM.

RESULTS

Maximal number of γH2Ax foci was detected 1 and 2 hours post-exposure to camptothecin and hydrogen peroxide, respectively; when adjusted for the baseline number of γH2Ax, neoplastic cell lines showed the lowest number of maximal DSB and slowest rate of repair compared to other cell lines. γH₂ Ax Western blot closely mirrored the trend observed in immunofluorescent staining for γH₂ Ax foci. Changes in the expression level of ATM and Nbs1 were minimal; however, ATM expression showed a slight gradual increase in normal cells which reached its peak at 2 hours after exposure to camptothecin.

CONCLUSIONS

There is a difference in efficiency of DSB repair pathways in cell lines derived from different stages of oral tumorigenesis with neoplastic cell lines having the most defective DSB repair system.

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

INTRODUCTION

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.

MATERIALS AND METHODS

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.

RESULTS

A total of 24 studies 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.

CONCLUSION

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.

Oral Chronic Graft-Versus-Host Disease: Pathogenesis, Diagnosis, Current Treatment, and Emerging Therapies

Chronic graft-versus-host disease (cGvHD) is a multisystem disorder that occurs in recipients of allogeneic hematopoietic (alloHCT) stem cell transplants and is characterized by both inflammatory and fibrotic manifestations. It begins with the recognition of host tissues by the non-self (allogeneic) graft and progresses to tissue inflammation, organ dysfunction and fibrosis throughout the body. Oral cavity manifestations of cGVHD include mucosal features, salivary gland dysfunction and fibrosis. This review synthesizes current knowledge on the pathogenesis, diagnosis and management of oral cGVHD, with a focus on emerging trends and novel therapeutics. Data from various clinical studies and expert consensus are integrated to provide a comprehensive overview.

World Workshop on Oral Medicine VIII: barriers to research in oral medicine: lessons learned from a bibliometric analysis of the oral potentially malignant disorders literature

OBJECTIVE

This study aimed to assess the impact of oral medicine (OM) practitioners on the literature regarding oral potentially malignant disorders (OPMDs), focusing on oral leukoplakia.

STUDY DESIGN

Using a bibliometric approach on the Scopus database until September 1, 2022, the top 100 cited articles were analyzed for article type, subtopic, specialty contributions, author metrics, and keywords. The Bibliometrix package for R and VOSviewer were used to evaluate interactions and generate science maps.

RESULTS

OM practitioners, comprising 39% of contributors, played a significant role in studies related to nomenclature and screening of OPMDs. Notably, 4 OM specialists ranked among the most prolific authors, demonstrating denser collaboration with OM co-authors compared to other cancer specialists. However, there was a scarcity of OPMD management studies authored by OM practitioners.

CONCLUSIONS

Despite the paucity of OM practitioners, the findings underscored the substantial contribution of OM practitioners in developing OPMD nomenclature and classification, emphasizing the need for increased collaboration with cancer specialists to conduct comprehensive clinical trials for OPMD management. The study highlights the importance of standardized criteria in OPMDs research for better data comparison and encourages further efforts from the OM scientific community.

Nicotine e-vaping and cardiovascular consequences: a case series and literature review

Background

Cardiovascular toxicity as a consequence of nicotine from conventional tobacco cigarette smoking is well documented. However, little is known about the cardiovascular consequences of nicotine e-vaping. We review the literature and report a case series of three cases of major adverse cardiovascular clinical effects post nicotine e-vaping.

Case summary

Three patients with known peripheral arterial disease who switched from heavy cigarette smoking consumption to a high-intensity dose of nicotine e-vaping all developed further arterial complications within 6-30 months.

Discussion

With the recent epidemic of e-vaping in young individuals and the national outbreak of e-vaping use-associated lung injury (EVALI), the dangers of e-vaping are now coming to light. The pulmonary effects are now well described, and this paper highlights three new cases of cardiovascular toxicity associated with e-vaping. The potential role of nicotine e-vaping and the risk of coronavirus disease-2019 (COVID-19) will also be discussed.

Deep Learning-Based Image Classification and Segmentation on Digital Histopathology for Oral Squamous Cell Carcinoma: A Systematic Review and Meta-Analysis

BACKGROUND

Artificial intelligence (AI)-based tools have shown promise in histopathology image analysis in improving the accuracy of oral squamous cell carcinoma (OSCC) detection with intent to reduce human error.

OBJECTIVES

This systematic review and meta-analysis evaluated deep learning (DL) models for OSCC detection on histopathology images by assessing common diagnostic performance evaluation metrics for AI-based medical image analysis studies.

METHODS

Diagnostic accuracy studies that used DL models for the analysis of histopathological images of OSCC compared to the reference standard were analyzed. Six databases (PubMed, Google Scholar, Scopus, Embase, ArXiv, and IEEE) were screened for publications without any time limitation. The QUADAS-2 tool was utilized to assess quality. The meta-analyses included only studies that reported true positives (TP), true negatives (TN), false positives (FP), and false negatives (FN) in their test sets.

RESULTS

Of 1267 screened studies, 17 studies met the final inclusion criteria. DL methods such as image classification (n = 11) and segmentation (n = 3) were used, and some studies used combined methods (n = 3). On QUADAS-2 assessment, only three studies had a low risk of bias across all applicability domains. For segmentation studies, 0.97 was reported for accuracy, 0.97 for sensitivity, 0.98 for specificity, and 0.92 for Dice. For classification studies, accuracy was reported as 0.99, sensitivity 0.99, specificity 1.0, Dice 0.95, F1 score 0.98, and AUC 0.99. Meta-analysis showed pooled estimates of 0.98 sensitivity and 0.93 specificity.

CONCLUSION

Application of AI-based classification and segmentation methods on image analysis represents a fundamental shift in digital pathology. DL approaches demonstrated significantly high accuracy for OSCC detection on histopathology, comparable to that of human experts in some studies. Although AI-based models cannot replace a well-trained pathologist, they can assist through improving the objectivity and repeatability of the diagnosis while reducing variability and human error as a consequence of pathologist burnout.

Generative Adversarial Networks in Digital Histopathology: Current Applications, Limitations, Ethical Considerations, and Future Directions

Generative adversarial networks (GANs) have gained significant attention in the field of image synthesis, particularly in computer vision. GANs consist of a generative model and a discriminative model trained in an adversarial setting to generate realistic and novel data. In the context of image synthesis, the generator produces synthetic images, whereas the discriminator determines their authenticity by comparing them with real examples. Through iterative training, the generator allows the creation of images that are indistinguishable from real ones, leading to high-quality image generation. Considering their success in computer vision, GANs hold great potential for medical diagnostic applications. In the medical field, GANs can generate images of rare diseases, aid in learning, and be used as visualization tools. GANs can leverage unlabeled medical images, which are large in size, numerous in quantity, and challenging to annotate manually. GANs have demonstrated remarkable capabilities in image synthesis and have the potential to significantly impact digital histopathology. This review article focuses on the emerging use of GANs in digital histopathology, examining their applications and potential challenges. Histopathology plays a crucial role in disease diagnosis, and GANs can contribute by generating realistic microscopic images. However, ethical considerations arise because of the reliance on synthetic or pseudogenerated images. Therefore, the manuscript also explores the current limitations and highlights the ethical considerations associated with the use of this technology. In conclusion, digital histopathology has seen an emerging use of GANs for image enhancement, such as color (stain) normalization, virtual staining, and ink/marker removal. GANs offer significant potential in transforming digital pathology when applied to specific and narrow tasks (preprocessing enhancements). Evaluating data quality, addressing biases, protecting privacy, ensuring accountability and transparency, and developing regulation are imperative to ensure the ethical application of GANs.

Oral plasmacytoma in multiple myeloma patients: report of 18 cases

OBJECTIVES

Plasmacytoma is a malignant neoplasm of plasma cells with or without an underlying multiple myeloma (MM). Plasmacytoma rarely involves the oral mucosa and jaw bones. We aimed to describe oral plasmacytoma in MM patients.

STUDY DESIGN

Patients with oral plasmacytoma seen at the Oral Medicine clinic from 2007 to 2024 were included in this study.

RESULTS

There were 18 patients (26 oral plasmacytomas). The median age was 60.5 years (72.3% females). The most common symptoms were pain and bleeding. Ten patients had gnathic involvement, and 8 patients had soft tissue disease without bony involvement. The gingiva was the most common site (50.0%) and buccolingual involvement was seen in 38.5% of cases. Treatment directed at the oral disease was rendered in 11 patients. Over 90% of patients died with a median of 5 months.

CONCLUSIONS

MM patients may present with oral involvement most commonly on the gingiva with frequent buccal and lingual involvement.

Unusual presentation of granular cell tumour of buccal mucosa

Granular cell tumour (GCT) is a rare soft tissue lesion which many consider to have malignant potential of yet unknown aetiopathogenesis. Oral GCT lesions may occur in an area of leucoplakia and are predominantly present on the tongue. This case study highlights an uncommon presentation of this condition located on the buccal mucosa and illustrates the need for meticulous evaluation of suspicious lesions. Due to the malignant risk associated with GCT lesions, it is important to make the correct diagnosis and ensure complete surgical excision for these cases. Ongoing long-term follow-up is also indicated to monitor for recurrence or malignancy.

Efficacy and empathy of AI chatbots in answering frequently asked questions on oral oncology

OBJECTIVES

Artificial intelligence chatbots have demonstrated feasibility and efficacy in improving health outcomes. In this study, responses from 5 different publicly available AI chatbots-Bing, GPT-3.5, GPT-4, Google Bard, and Claude-to frequently asked questions related to oral cancer were evaluated.

STUDY DESIGN

Relevant patient-related frequently asked questions about oral cancer were obtained from two main sources: public health websites and social media platforms. From these sources, 20 oral cancer-related questions were selected. Four board-certified specialists in oral medicine/oral and maxillofacial pathology assessed the answers using modified version of the global quality score on a 5-point Likert scale. Additionally, readability was measured using the Flesch-Kincaid Grade Level and Flesch Reading Ease scores. Responses were also assessed for empathy using a validated 5-point scale.

RESULTS

Specialists ranked GPT-4 with highest total score of 17.3 ± 1.5, while Bing received the lowest at 14.9 ± 2.2. Bard had the highest Flesch Reading Ease score of 62 ± 7; and ChatGPT-3.5 and Claude received the lowest scores (more challenging readability). GPT-4 and Bard emerged as the most superior chatbots in terms of empathy and accurate citations on patient-related frequently asked questions pertaining to oral cancer. GPT-4 had highest overall quality, whereas Bing showed the lowest level of quality, empathy, and accuracy for citations.

CONCLUSION

GPT-4 demonstrated the highest quality responses to frequently asked questions pertaining to oral cancer. Although impressive in their ability to guide patients on common oral cancer topics, most chatbots did not perform well when assessed for empathy or citation accuracy.