NUMERO: #20240078

Título: Artificial neural networks to predict malignant transformation of oral leukoplakia

Nome do Apresentador: Caíque Mariano PEDROSO

Categoria do Trabalho: Painel de pesquisa científica (PPC)

Área Temática: Estomatologia

Resumo: Objective: This study aimed to evaluate the performance of a neural network (NN) in predicting the malignant transformation of oral leukoplakia (OL). Study Design: A retrospective analysis was conducted using clinical and pathological data from patients diagnosed with OL. The NN was trained on relevant features to predict the malignant transformation. Performance metrics including accuracy, sensitivity, specificity, precision, and area under the curve (AUC) were calculated to assess the models predictive ability. Results: Data consisted of 97 patients with 10 input layer. The sample was composed predominantly by men, with average age of 59 years, smokers, with tongue lesions higher than 2 cm, being 64 with epithelial dysplasia (mainly moderate); Twenty-one recurrences and eleven malignant transformations occurred. The NN exhibited promising performance in predicting malignant transformation. The model showed higher performance with an accuracy of 0.85, sensitivity of 0.92, specificity of 0.77, precision of 0.80, F1-score of 0.86, and AUC of 0.85. Conclusion: The NN showed particularly strong performance in predicting malignant transformation, with high sensitivity and specificity, demonstrating as a potential diagnostic tool. These findings suggest that the NN could aid clinicians in improve predict outcomes for OL patients.
Financial support: The São Paulo Research Foundation/FAPESP

Autor 1:  Caíque Mariano PEDROSO

E-mail 1:  [email protected]

Autor 2:   Rafael Vasconcelos Costa MACHADO

E-mail 2:  [email protected]

Autor 3 :  Cristina SALDIVIA-SIRACUSA

E-mail 3:  [email protected]

Autor 4:  Anna Luiza Damaceno ARAUJO

E-mail 4:  [email protected]

Autor 5:  Márcio Ajudarte LOPES

E-mail 5:  [email protected]

Autor 6:  Matheus Cardoso MORAES

E-mail 6:  [email protected]

Autor 7:  Alan Roger SANTOS-SILVA

E-mail 7:  [email protected]