Adjunctive Diagnostic Techniques for Oral and Oropharyngeal Cancer Discovery

Artificial Intelligence-Based Diagnosis of Oral Lichen Planus Using Deep Convolutional Neural Networks

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.

Smart Diagnostics: Combining Artificial Intelligence and In Vitro Diagnostics

We are beginning a new era of Smart Diagnostics-integrated biosensors powered by recent innovations in embedded electronics, cloud computing, and artificial intelligence (AI). Universal and AI-based in vitro diagnostics (IVDs) have the potential to exponentially improve healthcare decision making in the coming years. This perspective covers current trends and challenges in translating Smart Diagnostics. We identify essential elements of Smart Diagnostics platforms through the lens of a clinically validated platform for digitizing biology and its ability to learn disease signatures. This platform for biochemical analyses uses a compact instrument to perform multiclass and multiplex measurements using fully integrated microfluidic cartridges compatible with the point of care. Image analysis digitizes biology by transforming fluorescence signals into inputs for learning disease/health signatures. The result is an intuitive Score reported to the patients and/or providers. This AI-linked universal diagnostic system has been validated through a series of large clinical studies and used to identify signatures for early disease detection and disease severity in several applications, including cardiovascular diseases, COVID-19, and oral cancer. The utility of this Smart Diagnostics platform may extend to multiple cell-based oncology tests via cross-reactive biomarkers spanning oral, colorectal, lung, bladder, esophageal, and cervical cancers, and is well-positioned to improve patient care, management, and outcomes through deployment of this resilient and scalable technology. Lastly, we provide a future perspective on the direction and trajectory of Smart Diagnostics and the transformative effects they will have on health care.

A novel lightweight deep convolutional neural network for early detection of oral cancer

OBJECTIVES

To develop a lightweight deep convolutional neural network (CNN) for binary classification of oral lesions into benign and malignant or potentially malignant using standard real-time clinical images.

METHODS

A small deep CNN, that uses a pretrained EfficientNet-B0 as a lightweight transfer learning model, was proposed. A data set of 716 clinical images was used to train and test the proposed model. Accuracy, specificity, sensitivity, receiver operating characteristics (ROC) and area under curve (AUC) were used to evaluate performance. Bootstrapping with 120 repetitions was used to calculate arithmetic means and 95% confidence intervals (CIs).

RESULTS

The proposed CNN model achieved an accuracy of 85.0% (95% CI: 81.0%-90.0%), a specificity of 84.5% (95% CI: 78.9%-91.5%), a sensitivity of 86.7% (95% CI: 80.4%-93.3%) and an AUC of 0.928 (95% CI: 0.88-0.96).

CONCLUSIONS

Deep CNNs can be an effective method to build low-budget embedded vision devices with limited computation power and memory capacity for diagnosis of oral cancer. Artificial intelligence (AI) can improve the quality and reach of oral cancer screening and early detection.

Nuclear F-actin Cytology in Oral Epithelial Dysplasia and Oral Squamous Cell Carcinoma

Oral cavity cancer has a low 5-y survival rate, but outcomes improve when the disease is detected early. Cytology is a less invasive method to assess oral potentially malignant disorders relative to the gold-standard scalpel biopsy and histopathology. In this report, we aimed to determine the utility of cytological signatures, including nuclear F-actin cell phenotypes, for classifying the entire spectrum of oral epithelial dysplasia and oral squamous cell carcinoma. We enrolled subjects with oral potentially malignant disorders, subjects with previously diagnosed malignant lesions, and healthy volunteers without lesions and obtained brush cytology specimens and matched scalpel biopsies from 486 subjects. Histopathological assessment of the scalpel biopsy specimens classified lesions into 6 categories. Brush cytology specimens were analyzed by machine learning classifiers trained to identify relevant cytological features. Multimodal diagnostic models were developed using cytology results, lesion characteristics, and risk factors. Squamous cells with nuclear F-actin staining were associated with early disease (i.e., lower proportions in benign lesions than in more severe lesions), whereas small round parabasal-like cells and leukocytes were associated with late disease (i.e., higher proportions in severe dysplasia and carcinoma than in less severe lesions). Lesions with the impression of oral lichen planus were unlikely to be either dysplastic or malignant. Cytological features substantially improved upon lesion appearance and risk factors in predicting squamous cell carcinoma. Diagnostic models accurately discriminated early and late disease with AUCs (95% CI) of 0.82 (0.77 to 0.87) and 0.93 (0.88 to 0.97), respectively. The cytological features identified here have the potential to improve screening and surveillance of the entire spectrum of oral potentially malignant disorders in multiple care settings.

Oral Potentially Malignant Disorders and Oral Cavity Cancer

The Surveillance, Epidemiology, and End Results program from the National Cancer Institute reports that the aggregate number of oral cavity and pharyngeal cancer cases has been increasing over the past decade and, despite an overall decline in oral cavity cancers, this increase is largely related to a dramatic increase in cancers involving oropharyngeal subsites. Early detection of oral cavity cancers is commensurate with improved survival, and opportunistic screening by trained clinicians to detect oral cavity cancer and oral potentially malignant disorders is recommended by the American Dental Association and the American Academy of Oral Medicine.

Point-of-care oral cytology tool for the screening and assessment of potentially malignant oral lesions

BACKGROUND

The effective detection and monitoring of potentially malignant oral lesions (PMOL) are critical to identifying early-stage cancer and improving outcomes. In the current study, the authors described cytopathology tools, including machine learning algorithms, clinical algorithms, and test reports developed to assist pathologists and clinicians with PMOL evaluation.

METHODS

Data were acquired from a multisite clinical validation study of 999 subjects with PMOLs and oral squamous cell carcinoma (OSCC) using a cytology-on-a-chip approach. A machine learning model was trained to recognize and quantify the distributions of 4 cell phenotypes. A least absolute shrinkage and selection operator (lasso) logistic regression model was trained to distinguish PMOLs and cancer across a spectrum of histopathologic diagnoses ranging from benign, to increasing grades of oral epithelial dysplasia (OED), to OSCC using demographics, lesion characteristics, and cell phenotypes. Cytopathology software was developed to assist pathologists in reviewing brush cytology test results, including high-content cell analyses, data visualization tools, and results reporting.

RESULTS

Cell phenotypes were determined accurately through an automated cytological assay and machine learning approach (99.3% accuracy). Significant differences in cell phenotype distributions across diagnostic categories were found in 3 phenotypes (type 1 ["mature squamous"], type 2 ["small round"], and type 3 ["leukocytes"]). The clinical algorithms resulted in acceptable performance characteristics (area under the curve of 0.81 for benign vs mild dysplasia and 0.95 for benign vs malignancy).

CONCLUSIONS

These new cytopathology tools represent a practical solution for rapid PMOL assessment, with the potential to facilitate screening and longitudinal monitoring in primary, secondary, and tertiary clinical care settings.