Scholarlink

Objective <p>This study aimed to evaluate the coherence between data heterogeneity and model complexity by comparing seven convolutional neural network (CNN) architectures—trained with and without ImageNet pretraining—in a multiclass framework for the histopathological classification of three odontogenic tumors: adenomatoid odontogenic tumor, ameloblastoma, and ameloblastic carcinoma. The goal was to investigate how transfer learning influences performance and diagnostic reliability in a clinically relevant context characterized by overlapping histological patterns.</p> Methods <p>An international, multicenter cross-sectional dataset of 64 hematoxylin- and eosin-stained whole slide images was analyzed, including adenomatoid odontogenic tumor (<i>n</i> = 16), ameloblastoma (<i>n</i> = 27), and ameloblastic carcinoma (<i>n</i> = 21). Seven CNN models (DenseNet121, EfficientNetV2B0, InceptionV3, MobileNet, ResNet50, VGG16, and Xception) were trained and tested on 455,107 patches (224 × 224 pixels). Performance was assessed using accuracy, balanced accuracy, sensitivity, specificity, F1-score, and AUC.</p> Results <p>Without ImageNet pretraining, DenseNet121 achieved the highest performance (accuracy = 0.73, balanced accuracy = 0.74, AUC = 0.78, specificity = 0.84, sensitivity = 0.65), followed by EfficientNetV2B0 (accuracy = 0.67, balanced accuracy = 0.68, sensitivity = 0.54). When ImageNet pretraining was applied, performance improved across all architectures. EfficientNetV2B0 reached the best overall results (accuracy = 0.79, balanced accuracy = 0.81, AUC = 0.91, specificity = 0.88, sensitivity = 0.74), while DenseNet121 maintained consistent performance (accuracy = 0.72, balanced accuracy = 0.74, AUC = 0.85, specificity = 0.84, sensitivity = 0.64).</p> Conclusion <p>Transfer learning with ImageNet weights enhanced the performance of most CNNs, with EfficientNetV2B0 showing the greatest responsiveness to pretraining and DenseNet121 demonstrating intrinsic robustness to initialization. These results highlight the potential of CNN-based frameworks to support the differential diagnosis of odontogenic tumors—an inherently challenging task due to morphological overlap—while establishing reproducible methodological baselines that contribute to the global development of explainable, ensemble-based, and clinically reliable AI systems in oral pathology.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Impact of Transfer Learning on Convolutional Neural Networks for Odontogenic Tumor Diagnosis

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月19日

查看全文
  • Daniela Giraldo-Roldán,
  • Thaís Cerqueira Reis Nakamura,
  • Anderson Faria Claret,
  • Giovanna Calabrese dos Santos,
  • Katya Pulido-Díaz,
  • Roberto Gerber-Mora,
  • Leonor Victoria Gonzalez-Perez,
  • Jeconias Câmara,
  • Hélder Antônio Rebelo Pontes,
  • Manoela Domingues Martin,
  • Márcio Campos Oliveira,
  • Vanesa Pereira Prado,
  • Felipe Martins Silveira,
  • Ronell Bologna-Molina,
  • Anna Luíza Damaceno Araújo,
  • Matheus Cardoso Moraes,
  • Pablo Agustin Vargas

摘要

Objective

This study aimed to evaluate the coherence between data heterogeneity and model complexity by comparing seven convolutional neural network (CNN) architectures—trained with and without ImageNet pretraining—in a multiclass framework for the histopathological classification of three odontogenic tumors: adenomatoid odontogenic tumor, ameloblastoma, and ameloblastic carcinoma. The goal was to investigate how transfer learning influences performance and diagnostic reliability in a clinically relevant context characterized by overlapping histological patterns.

Methods

An international, multicenter cross-sectional dataset of 64 hematoxylin- and eosin-stained whole slide images was analyzed, including adenomatoid odontogenic tumor (n = 16), ameloblastoma (n = 27), and ameloblastic carcinoma (n = 21). Seven CNN models (DenseNet121, EfficientNetV2B0, InceptionV3, MobileNet, ResNet50, VGG16, and Xception) were trained and tested on 455,107 patches (224 × 224 pixels). Performance was assessed using accuracy, balanced accuracy, sensitivity, specificity, F1-score, and AUC.

Results

Without ImageNet pretraining, DenseNet121 achieved the highest performance (accuracy = 0.73, balanced accuracy = 0.74, AUC = 0.78, specificity = 0.84, sensitivity = 0.65), followed by EfficientNetV2B0 (accuracy = 0.67, balanced accuracy = 0.68, sensitivity = 0.54). When ImageNet pretraining was applied, performance improved across all architectures. EfficientNetV2B0 reached the best overall results (accuracy = 0.79, balanced accuracy = 0.81, AUC = 0.91, specificity = 0.88, sensitivity = 0.74), while DenseNet121 maintained consistent performance (accuracy = 0.72, balanced accuracy = 0.74, AUC = 0.85, specificity = 0.84, sensitivity = 0.64).

Conclusion

Transfer learning with ImageNet weights enhanced the performance of most CNNs, with EfficientNetV2B0 showing the greatest responsiveness to pretraining and DenseNet121 demonstrating intrinsic robustness to initialization. These results highlight the potential of CNN-based frameworks to support the differential diagnosis of odontogenic tumors—an inherently challenging task due to morphological overlap—while establishing reproducible methodological baselines that contribute to the global development of explainable, ensemble-based, and clinically reliable AI systems in oral pathology.