Machine Learning for Predicting Malignant Transformation in Actinic Cheilitis: A Prognostic Support System Based on Demographic and Clinical Descriptors

article
Autores

Correia‐Neto, Ivan José

Da Costa, Alex Franco

Araújo, Anna Luíza Damaceno

Saldivia‐Siracusa, Cristina

De Sá, Raísa Sales

Pereira, Thiago Martini

Vargas, Pablo Agustin

Santos‐Silva, Alan Roger

Kowalski, Luiz Paulo

Moraes, Matheus Cardoso

Lopes, Marcio Ajudarte

Data de Publicação

19 de janeiro de 2026

Resumo

ABSTRACT Objective This study aimed to develop and evaluate Machine Learning models to predict the malignant transformation (MT) in patients with actinic cheilitis (AC). Methods Three hundred forty patients diagnosed with AC (322 in the no MT group, and 18 in the MT group) were carefully documented. The study used the Adaptive Synthetic Sampling to adaptively balance the dataset (322 in the no MT group and 319 in the MT group). Four supervised Machine Learning classifiers (Random Forest, Xtreme Gradient Boosting, Multilayer Perceptron, and Support Vector Machine) were trained and tested using 5‐fold cross‐validation to correlate inputs (clinical descriptors and demographic data) to outputs (MT). SHAP values were used to identify the most influential predictors of MT. Results The Xtreme Gradient Boosting model stood out, achieving 96.72% accuracy, 96.87% sensitivity, 96.57% specificity, 96.61% precision, 96.73% of F1‐Score, and 0.9498 AUC. Multilayer Perceptron showed the best sensitivity (98.44%), and Random Forest presented comparable results. In contrast, Support Vector Machine underperformed, with higher values of false negatives and false positives. Across models, ulceration, multifocality, and long‐standing lesions were the strongest predictors of MT, while small, asymptomatic, or solitary lesions were associated with lower risk. Conclusion The results revealed promising performance metrics for Xtreme Gradient Boosting and Multilayer Perceptron suggesting their potential value as tools in a support system for monitoring AC. Additionally, synthetic data proved constructive in training, enhancing the models’ robustness and predictive capabilities.

Citação

BibTeX
@online{ivan_josé2026,
  author = {Ivan José , Correia‐Neto and Costa, Alex Franco, Da and Anna
    Luíza Damaceno , Araújo and Cristina , Saldivia‐Siracusa and Sá,
    Raísa Sales, De and Thiago Martini , Pereira and Pablo Agustin ,
    Vargas and Alan Roger , Santos‐Silva and Luiz Paulo , Kowalski and
    Matheus Cardoso , Moraes and Marcio Ajudarte , Lopes},
  title = {Machine Learning for Predicting Malignant Transformation in
    Actinic Cheilitis: A Prognostic Support System Based on Demographic
    and Clinical Descriptors},
  date = {2026-01-19},
  doi = {10.1111/jop.70113},
  langid = {pt-BR},
  abstract = {ABSTRACT Objective This study aimed to develop and
    evaluate Machine Learning models to predict the malignant
    transformation (MT) in patients with actinic cheilitis (AC). Methods
    Three hundred forty patients diagnosed with AC (322 in the no MT
    group, and 18 in the MT group) were carefully~documented. The study
    used the Adaptive Synthetic Sampling to adaptively balance the
    dataset (322 in the no MT group and 319 in the MT group). Four
    supervised Machine Learning classifiers (Random Forest, Xtreme
    Gradient Boosting, Multilayer Perceptron, and Support Vector
    Machine) were trained and tested using 5‐fold cross‐validation to
    correlate inputs (clinical descriptors and demographic data) to
    outputs (MT). SHAP values were used to identify the most influential
    predictors of MT. Results The Xtreme Gradient Boosting model stood
    out, achieving 96.72\% accuracy, 96.87\% sensitivity, 96.57\%
    specificity, 96.61\% precision, 96.73\% of F1‐Score, and 0.9498 AUC.
    Multilayer Perceptron showed the best sensitivity (98.44\%), and
    Random Forest presented comparable results. In contrast, Support
    Vector Machine underperformed, with higher values of false negatives
    and false positives. Across models, ulceration, multifocality, and
    long‐standing lesions were the strongest predictors of MT, while
    small, asymptomatic, or solitary lesions were associated with lower
    risk. Conclusion The results revealed promising performance metrics
    for Xtreme Gradient Boosting and Multilayer Perceptron suggesting
    their potential value as tools in a support system for monitoring
    AC. Additionally, synthetic data proved constructive in training,
    enhancing the models’ robustness and predictive capabilities.}
}
Por favor, cite este trabalho como:
Ivan José, Correia‐Neto, Da Costa, Alex Franco, Araújo Anna Luíza Damaceno, Saldivia‐Siracusa Cristina, De Sá, Raísa Sales, Pereira Thiago Martini, Vargas Pablo Agustin, et al. 2026. “Machine Learning for Predicting Malignant Transformation in Actinic Cheilitis: A Prognostic Support System Based on Demographic and Clinical Descriptors.” Journal of Oral Pathology & Medicine. January 19, 2026. https://doi.org/10.1111/jop.70113.