Journal of pharmacy & bioallied sciences · 2025
Evaluation of Artificial Intelligence-Based Diagnostic Tools for Accurate Detection and Classification of Oral Lesions: A Clinical Study.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Shah Khushali, Dhimole Ankit, Pandey Neetu, Naorem Robium, Srivastava Rangoli, Gupta Sameer
Аннотация
BACKGROUND: Early detection and accurate classification of oral lesions are essential for effective clinical management and prevention of malignant transformation. Recent advancements in artificial intelligence (AI) have led to the development of diagnostic tools aimed at enhancing clinical decision-making.
METHODS: A prospective clinical study was conducted involving 180 patients presenting with suspicious oral lesions at a tertiary dental care center. Digital images of lesions were captured and analyzed using an AI-powered diagnostic software trained on a dataset of over 10,000 labeled images. Each case was also evaluated by two oral pathologists, followed by histopathological confirmation. Sensitivity, specificity, accuracy, and Cohen's kappa coefficient were calculated to assess diagnostic concordance.
RESULTS: The AI system achieved an overall diagnostic accuracy of 91.1% (±4.3), sensitivity of 93.6%, and specificity of 89.2%. The concordance between AI and histopathological diagnosis was substantial (κ =0.83; P < 0.001). The highest performance was observed in detecting leukoplakia (95.2% accuracy) and oral lichen planus (92.4%). Misclassification occurred primarily in cases of mild dysplasia and nonspecific ulcerations.
CONCLUSION: The AI-based diagnostic tool demonstrated high diagnostic accuracy and substantial agreement with histopathological findings. It shows promise as a chairside adjunct to enhance clinical decision-making and early detection of potentially malignant oral lesions.
METHODS: A prospective clinical study was conducted involving 180 patients presenting with suspicious oral lesions at a tertiary dental care center. Digital images of lesions were captured and analyzed using an AI-powered diagnostic software trained on a dataset of over 10,000 labeled images. Each case was also evaluated by two oral pathologists, followed by histopathological confirmation. Sensitivity, specificity, accuracy, and Cohen's kappa coefficient were calculated to assess diagnostic concordance.
RESULTS: The AI system achieved an overall diagnostic accuracy of 91.1% (±4.3), sensitivity of 93.6%, and specificity of 89.2%. The concordance between AI and histopathological diagnosis was substantial (κ =0.83; P < 0.001). The highest performance was observed in detecting leukoplakia (95.2% accuracy) and oral lichen planus (92.4%). Misclassification occurred primarily in cases of mild dysplasia and nonspecific ulcerations.
CONCLUSION: The AI-based diagnostic tool demonstrated high diagnostic accuracy and substantial agreement with histopathological findings. It shows promise as a chairside adjunct to enhance clinical decision-making and early detection of potentially malignant oral lesions.