The Use of Machine Learning to Support the Diagnosis of Oral Alterations
Objective: To verify the accuracy of deep learning models in detecting cellular alterations in histological images of oral mucosa. Material and Methods: The study compares three convolutional neural network (CNN) architectures for classifying histological images: EfficientNet-B3, MobileNet-V2, and...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Association of Support to Oral Health Research (APESB)
2025-01-01
|
| Series: | Pesquisa Brasileira em Odontopediatria e Clínica Integrada |
| Subjects: | |
| Online Access: | https://revista.uepb.edu.br/PBOCI/article/view/4227 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832586127965421568 |
|---|---|
| author | Rosana Leal do Prado Juliane Avansini Marsicano Amanda Keren Frois Jacques Duílio Brancher |
| author_facet | Rosana Leal do Prado Juliane Avansini Marsicano Amanda Keren Frois Jacques Duílio Brancher |
| author_sort | Rosana Leal do Prado |
| collection | DOAJ |
| description |
Objective: To verify the accuracy of deep learning models in detecting cellular alterations in histological images of oral mucosa. Material and Methods: The study compares three convolutional neural network (CNN) architectures for classifying histological images: EfficientNet-B3, MobileNet-V2, and VGG16. Efficient and focused on computer vision, each has specific advantages. A Kaggle database with 5192 images was used, divided into training (70%), validation (15%), and test (15%) sets. The CNNs were implemented using the Keras library, trained with pre-trained ImageNet weights, and evaluated using accuracy and AUC metrics. Results: The findings indicate that EfficientNet-B3 achieved the lowest training and validation losses at epoch 30, with the highest accuracy and stability during training. Evaluation metrics showed EfficientNet-B3 with 98% accuracy and 99% sensitivity for oral squamous cell carcinoma (OSCC) images, outperforming MobileNet-V2 and VGG16. MobileNet-V2 achieved 97% accuracy and 96% sensitivity, while VGG16 reached 94% accuracy and 93% sensitivity for OSCC images. All models exhibited high sensitivity and specificity in differentiating between normal and OSCC images, as demonstrated by ROC curves. EfficientNet-B3 had the highest AUC (0.982), followed by MobileNet-V2 (AUC=0.967) and VGG16 (AUC=0.937). These findings underscore the effectiveness of EfficientNet-B3 for accurately detecting cellular alterations in histological images of oral mucosa. Conclusion: Our study reveals the superior performance of CNNs, particularly EfficientNet-B3, in classifying histological images of OSCC.
|
| format | Article |
| id | doaj-art-5daff4c5831248fd87995448aa2a8dbc |
| institution | Kabale University |
| issn | 1519-0501 1983-4632 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Association of Support to Oral Health Research (APESB) |
| record_format | Article |
| series | Pesquisa Brasileira em Odontopediatria e Clínica Integrada |
| spelling | doaj-art-5daff4c5831248fd87995448aa2a8dbc2025-01-26T10:23:50ZengAssociation of Support to Oral Health Research (APESB)Pesquisa Brasileira em Odontopediatria e Clínica Integrada1519-05011983-46322025-01-0125The Use of Machine Learning to Support the Diagnosis of Oral AlterationsRosana Leal do PradoJuliane Avansini MarsicanoAmanda Keren FroisJacques Duílio Brancher Objective: To verify the accuracy of deep learning models in detecting cellular alterations in histological images of oral mucosa. Material and Methods: The study compares three convolutional neural network (CNN) architectures for classifying histological images: EfficientNet-B3, MobileNet-V2, and VGG16. Efficient and focused on computer vision, each has specific advantages. A Kaggle database with 5192 images was used, divided into training (70%), validation (15%), and test (15%) sets. The CNNs were implemented using the Keras library, trained with pre-trained ImageNet weights, and evaluated using accuracy and AUC metrics. Results: The findings indicate that EfficientNet-B3 achieved the lowest training and validation losses at epoch 30, with the highest accuracy and stability during training. Evaluation metrics showed EfficientNet-B3 with 98% accuracy and 99% sensitivity for oral squamous cell carcinoma (OSCC) images, outperforming MobileNet-V2 and VGG16. MobileNet-V2 achieved 97% accuracy and 96% sensitivity, while VGG16 reached 94% accuracy and 93% sensitivity for OSCC images. All models exhibited high sensitivity and specificity in differentiating between normal and OSCC images, as demonstrated by ROC curves. EfficientNet-B3 had the highest AUC (0.982), followed by MobileNet-V2 (AUC=0.967) and VGG16 (AUC=0.937). These findings underscore the effectiveness of EfficientNet-B3 for accurately detecting cellular alterations in histological images of oral mucosa. Conclusion: Our study reveals the superior performance of CNNs, particularly EfficientNet-B3, in classifying histological images of OSCC. https://revista.uepb.edu.br/PBOCI/article/view/4227Deep LearningMouth NeoplasmsNeural Networks, ComputerMachine Learning |
| spellingShingle | Rosana Leal do Prado Juliane Avansini Marsicano Amanda Keren Frois Jacques Duílio Brancher The Use of Machine Learning to Support the Diagnosis of Oral Alterations Pesquisa Brasileira em Odontopediatria e Clínica Integrada Deep Learning Mouth Neoplasms Neural Networks, Computer Machine Learning |
| title | The Use of Machine Learning to Support the Diagnosis of Oral Alterations |
| title_full | The Use of Machine Learning to Support the Diagnosis of Oral Alterations |
| title_fullStr | The Use of Machine Learning to Support the Diagnosis of Oral Alterations |
| title_full_unstemmed | The Use of Machine Learning to Support the Diagnosis of Oral Alterations |
| title_short | The Use of Machine Learning to Support the Diagnosis of Oral Alterations |
| title_sort | use of machine learning to support the diagnosis of oral alterations |
| topic | Deep Learning Mouth Neoplasms Neural Networks, Computer Machine Learning |
| url | https://revista.uepb.edu.br/PBOCI/article/view/4227 |
| work_keys_str_mv | AT rosanalealdoprado theuseofmachinelearningtosupportthediagnosisoforalalterations AT julianeavansinimarsicano theuseofmachinelearningtosupportthediagnosisoforalalterations AT amandakerenfrois theuseofmachinelearningtosupportthediagnosisoforalalterations AT jacquesduiliobrancher theuseofmachinelearningtosupportthediagnosisoforalalterations AT rosanalealdoprado useofmachinelearningtosupportthediagnosisoforalalterations AT julianeavansinimarsicano useofmachinelearningtosupportthediagnosisoforalalterations AT amandakerenfrois useofmachinelearningtosupportthediagnosisoforalalterations AT jacquesduiliobrancher useofmachinelearningtosupportthediagnosisoforalalterations |