Objective The aim of this study was to employ artificial intelligence (AI) via convolutional neural network (CNN) for the separation of oral lichen planus (OLP) and non-OLP in biopsy-proven clinical cases of OLP and non-OLP. Materials and Methods Data comprised of clinical photographs of 609 OLP and 480 non-OLP which diagnosis has been confirmed histopathologically. Fifty-five photographs from the OLP and non-OLP groups were randomly selected for use as the test dataset, while the remaining were used as training and validation datasets. Data augmentation was performed on the training dataset to increase the number and variation of photographs. Performance metrics for the CNN model performance included accuracy, positive predictive value, negative predictive value, sensitivity, specificity, and F1-score. Gradient-weighted class activation mapping was also used to visualize the important regions associated with discriminative clinical features on which the model relies. Results All the selected CNN models were able to diagnose OLP and non-OLP lesions using photographs. The performance of the Xception model was significantly higher than that of the other models in terms of overall accuracy and F1-score. Conclusions Our demonstration shows that CNN models can achieve an accuracy of 82 to 88%. Xception model performed the best in terms of both accuracy and F1-score.

Benign fibro-osseous lesion (BFOL) is a distinct group of jaw entities composed of fibrocellular tissue and mineralized materials. In this study, we examined the epidemiological, clinical, and pathological features of patients with BFOL. Records and microslides of 207 BFOLs submitted to pathology service were retrospectively reviewed. Overall, fibrous dysplasia (FD) was the most prevalent (36.7%), followed by ossifying fibroma (OF; 32.4%), osseous dysplasia (OD; 24.6%), and juvenile ossifying fibroma (JOF; 6.3%). Female predilection was noted. FD and JOF were common in maxilla, whereas most OF and OD affected the mandible. Most patients with FD and OF presented with painless swelling, while patients with OD were symptomless. The majority of FD specimens showed woven bone, while a mixture of woven bone and cementum-like materials was often noted in OF and OD. Our data show variations in the clinicopathologic features of BFOLs. A thorough examination of all aspects of BFOL patients is imperative for accurate diagnosis.

Claudins constitute a group of principal proteins forming the tight junctional complex. The altered expression of selected claudins has been reported in several human cancers. The purpose of this study was to investigate the expression of claudin-1 and claudin-4 in oral squamous cell carcinoma (OSCC) and examine its relationship with patient clinical-pathologic features. Forty-five OSCC cases were enrolled. Patient clinical, pathologic and follow-up data were reviewed and the claudin-1 and claudin-4 expression was analyzed immunohistochemically. Positive claudin-1 and claudin-4 immunoreactivities were noted in 86.7 and 80 % of cases, respectively. The majority of cases showed the staining in less than 25 % of cancer cells. The increased claudin-1 expression was significantly associated with the high pathologic grade, the presence of microscopic perineural invasion, vascular invasion, nodal metastasis, and advanced clinical stage. No relationship between various clinico-pathologic parameters and differential claudin-4 expression was observed. Claudin-1 may play a role in OSCC progression and could serve as a prognostic marker of advanced disease.

Click on the article title to read more.

The overlapping characteristics of oral lichen planus (OLP), a chronic oral mucosal inflammatory condition, with those of other oral lesions, present diagnostic challenges. Large language models (LLMs) with integrated computer-vision capabilities and convolutional neural networks (CNNs) constitute an alternative diagnostic modality. We evaluated the ability of seven LLMs, including both proprietary and open-source models, to detect OLP from intraoral images and generate differential diagnoses. Using a dataset with 1,142 clinical photographs of histopathologically confirmed OLP, non-OLP lesions, and normal mucosa. The LLMs were tested using three experimental designs: zero-shot recognition, example-guided recognition, and differential diagnosis. Performance was measured using accuracy, precision, recall, F1-score, and discounted cumulative gain (DCG). Furthermore, the performance of LLMs was compared with three previously published CNN-based models for OLP detection on a subset of 110 photographs, which were previously used to test the CNN models. Gemini 1.5 Pro and Flash demonstrated the highest accuracy (69.69%) in zero-shot recognition, whereas GPT-4o ranked first in the F1 score (76.10%). With example-guided prompts, which improved consistency and reduced refusal rates, Gemini 1.5 Flash achieved the highest accuracy (80.53%) and F1-score (84.54%); however, Claude 3.5 Sonnet achieved the highest DCG score of 0.63. Although the proprietary models generally excelled, the open-source Llama model demonstrated notable strengths in ranking relevant diagnoses despite moderate performance in detection tasks. All LLMs were outperformed by the CNN models. The seven evaluated LLMs lack sufficient performance for clinical use. CNNs trained to detect OLP outperformed the LLMs tested in this study.

Background: Aberrant expression of stem cell markers has been observed in several types of neoplasms. This trait attributes to the acquired stem-like property of tumor cells and can impact patient prognosis. The objective of this study was to comparatively analyze the expression and significance of SOX2 and OCT4 in various types of odontogenic cysts and tumors. Methods: Fifty-five cases of odontogenic cysts and tumors, including 15 ameloblastomas (AM), 5 adenomatoid odontogenic tumors (AOT), 5 ameloblastic fibromas (AF), 5 calcifying odontogenic cysts (COC), 10 dentigerous cysts (DC) and 15 odontogenic keratocysts (OKC) were investigated for the expression of SOX2 and OCT4 immunohistochemically. Results: Most OKCs (86.7 %) and all AFs expressed SOX2 in more than 50 % of epithelial cells. Its immunoreactivity was moderate-to-strong in all epithelial cell types in both lesions. In contrast, SOX2 expression was undetectable in AOTs and limited to the ameloblast-like cells in a minority of AM and COC cases. Most DCs showed positive staining in less than 25 % of cystic epithelium. Significantly greater SOX2 expression was noted in OKC compared with DC or AM, and in AF compared with COC or AOT. OCT4 rarely expressed in odontogenic lesions with the immunoreactivity being mild and present exclusively in OKCs. Conclusions: SOX2 is differentially expressed in odontogenic cysts and tumors. This could be related to their diverse cells of origin or stages of histogenesis. The overexpression of SOX2 and OCT4 in OKC indicates the acquired stem-like property. Future studies should investigate whether the overexpression of OCT4 and SOX2 contributes to the aggressive behaviors of the tumors.