Effective data selection via deep learning processes and corresponding learning strategies in ultrasound image classification
Abstract In this study, we propose a novel approach to enhancing transfer learning by optimizing data selection through deep learning techniques and corresponding innovative learning strategies. This method is particularly beneficial when the available dataset has reached its limit and cannot be fur...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-05-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-00416-5 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849312023236575232 |
|---|---|
| author | Hyunju Lee Jin Young Kwak Eunjung Lee |
| author_facet | Hyunju Lee Jin Young Kwak Eunjung Lee |
| author_sort | Hyunju Lee |
| collection | DOAJ |
| description | Abstract In this study, we propose a novel approach to enhancing transfer learning by optimizing data selection through deep learning techniques and corresponding innovative learning strategies. This method is particularly beneficial when the available dataset has reached its limit and cannot be further expanded. Our approach focuses on maximizing the use of existing data to improve learning outcomes which offers an effective solution for data-limited applications in medical imaging classification. The proposed method consists of two stages. In the first stage, an original network performs the initial classification. When the original network exhibits low confidence in its predictions, ambiguous classifications are passed to a secondary decision-making step involving a newly trained network, referred to as the True network. The True network shares the same architecture as the original network but is trained on a subset of the original dataset that is selected based on consensus among multiple independent networks. It is then used to verify the classification results of the original network, identifying and correcting any misclassified images. To evaluate the effectiveness of our approach, we conducted experiments using thyroid nodule ultrasound images with the ResNet101 and Vision Transformer architectures along with eleven other pre-trained neural networks. The proposed method led to performance improvements across all five key metrics, accuracy, sensitivity, specificity, F1-score, and AUC, compared to using only the original or True networks in ResNet101. Additionally, the True network showed strong performance when applied to the Vision Transformer and similar enhancements were observed across multiple convolutional neural network architectures. Furthermore, to assess the robustness and adaptability of our method across different medical imaging modalities, we applied it to dermoscopic images and observed similar performance enhancements. These results provide evidence of the effectiveness of our approach in improving transfer learning-based medical image classification without requiring additional training data. |
| format | Article |
| id | doaj-art-a2d64d84c35b4b5ea08b197c90c5d8bf |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-a2d64d84c35b4b5ea08b197c90c5d8bf2025-08-20T03:53:12ZengNature PortfolioScientific Reports2045-23222025-05-0115111210.1038/s41598-025-00416-5Effective data selection via deep learning processes and corresponding learning strategies in ultrasound image classificationHyunju Lee0Jin Young Kwak1Eunjung Lee2Department of Radiology, Severance Hospital, Research Institute of Radiological Science, College of Medicine, Yonsei UniversityDepartment of Radiology, Severance Hospital, Research Institute of Radiological Science, College of Medicine, Yonsei UniversitySchool of Mathematics and Computing, Yonsei UniversityAbstract In this study, we propose a novel approach to enhancing transfer learning by optimizing data selection through deep learning techniques and corresponding innovative learning strategies. This method is particularly beneficial when the available dataset has reached its limit and cannot be further expanded. Our approach focuses on maximizing the use of existing data to improve learning outcomes which offers an effective solution for data-limited applications in medical imaging classification. The proposed method consists of two stages. In the first stage, an original network performs the initial classification. When the original network exhibits low confidence in its predictions, ambiguous classifications are passed to a secondary decision-making step involving a newly trained network, referred to as the True network. The True network shares the same architecture as the original network but is trained on a subset of the original dataset that is selected based on consensus among multiple independent networks. It is then used to verify the classification results of the original network, identifying and correcting any misclassified images. To evaluate the effectiveness of our approach, we conducted experiments using thyroid nodule ultrasound images with the ResNet101 and Vision Transformer architectures along with eleven other pre-trained neural networks. The proposed method led to performance improvements across all five key metrics, accuracy, sensitivity, specificity, F1-score, and AUC, compared to using only the original or True networks in ResNet101. Additionally, the True network showed strong performance when applied to the Vision Transformer and similar enhancements were observed across multiple convolutional neural network architectures. Furthermore, to assess the robustness and adaptability of our method across different medical imaging modalities, we applied it to dermoscopic images and observed similar performance enhancements. These results provide evidence of the effectiveness of our approach in improving transfer learning-based medical image classification without requiring additional training data.https://doi.org/10.1038/s41598-025-00416-5Data selectionTwo-step decision making processTrue network |
| spellingShingle | Hyunju Lee Jin Young Kwak Eunjung Lee Effective data selection via deep learning processes and corresponding learning strategies in ultrasound image classification Scientific Reports Data selection Two-step decision making process True network |
| title | Effective data selection via deep learning processes and corresponding learning strategies in ultrasound image classification |
| title_full | Effective data selection via deep learning processes and corresponding learning strategies in ultrasound image classification |
| title_fullStr | Effective data selection via deep learning processes and corresponding learning strategies in ultrasound image classification |
| title_full_unstemmed | Effective data selection via deep learning processes and corresponding learning strategies in ultrasound image classification |
| title_short | Effective data selection via deep learning processes and corresponding learning strategies in ultrasound image classification |
| title_sort | effective data selection via deep learning processes and corresponding learning strategies in ultrasound image classification |
| topic | Data selection Two-step decision making process True network |
| url | https://doi.org/10.1038/s41598-025-00416-5 |
| work_keys_str_mv | AT hyunjulee effectivedataselectionviadeeplearningprocessesandcorrespondinglearningstrategiesinultrasoundimageclassification AT jinyoungkwak effectivedataselectionviadeeplearningprocessesandcorrespondinglearningstrategiesinultrasoundimageclassification AT eunjunglee effectivedataselectionviadeeplearningprocessesandcorrespondinglearningstrategiesinultrasoundimageclassification |