Towards a Better Understanding of Transfer Learning for Medical Imaging: A Case Study
Open accessThe DOI matches its Crossref record; 62 of the 62 references it cites were found in a registry or an index.
Abstract
One of the main challenges of employing deep learning models in the field of medicine is a lack of training data due to difficulty in collecting and labeling data, which needs to be performed by experts. To overcome this drawback, transfer learning (TL) has been utilized to solve several medical imaging tasks using pre-trained state-of-the-art models from the ImageNet dataset. However, there are primary divergences in data features, sizes, and task characteristics between the natural image classification and the targeted medical imaging tasks. Therefore, TL can slightly improve performance if the source domain is completely different from the target domain. In this paper, we explore the benefit of TL from the same and different domains of the target tasks. To do so, we designed a deep convolutional neural network (DCNN) model that integrates three ideas including traditional and parallel convolutional layers and residual connections along with global average pooling. We trained the proposed model against several scenarios. We utilized the same and different domain TL with the diabetic foot ulcer (DFU) classification task and with the animal classification task. We have empirically shown that the source of TL from the same domain can significantly improve the performance considering a reduced number of images in the same domain of the target dataset. The proposed model with the DFU dataset achieved F1-score value of 86.6% when trained from scratch, 89.4% with TL from a different domain of the targeted dataset, and 97.6% with TL from the same domain of the targeted dataset.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Increase the training time.”
From Review of deep learning: concepts, CNN architectures, challenges, applications, future directions (2021) · cites this paper as [87]
In the citing paper
“Add regularization.”
From Review of deep learning: concepts, CNN architectures, challenges, applications, future directions (2021) · cites this paper as [87]
In the citing paper
“Increase the depth (or width) of the model.”
From Review of deep learning: concepts, CNN architectures, challenges, applications, future directions (2021) · cites this paper as [87]
In the citing paper
“Expand the dataset with data augmentation or use transfer learning (explained in latter sections).”
From Review of deep learning: concepts, CNN architectures, challenges, applications, future directions (2021) · cites this paper as [87]
In the citing paper
“Increase hyperparameters tuning.”
From Review of deep learning: concepts, CNN architectures, challenges, applications, future directions (2021) · cites this paper as [87]
2 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 10×
“Hence, studying and designing a better network is necessary [35,39].”
InfluentialImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 12×
“Krizhevsky et al. [30] developed the CNN further when they introduced the AlexNet network.”
InfluentialLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
ImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333 · cited 2×
“For example, DCNN models trained on the ImageNet dataset [60], which comprises of natural images, are utilized to enhance the performance of the medical image classification task.”
Going deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562
(2014) · Cited by 33,419
Rethinking the Inception Architecture for Computer Vision
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., & Wojna, Z. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.308 · Cited by 32,138
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Ioffe, S., & Szegedy, C. (2015) · arXiv · doi:10.48550/arxiv.1502.03167 · Cited by 27,818 · cited 1×
“The DCNN attained fast and excellent advancements in different attributes involving activation function [48], regularization strategies [49], and optimization techniques [50].”
Visualizing and understanding convolutional networks
Zeiler, M. D., & Fergus, R. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10590-1_53 · Cited by 15,839 · cited 1×
“These CNNs extract features via the convolutional layers and executes the classification tasks by the fully connected layers [37,40,41].”
A survey on Image Data Augmentation for Deep Learning
Shorten, C., & Khoshgoftaar, T. M. (2019) · Journal of Big Data · doi:10.1186/s40537-019-0197-0 · Cited by 13,203 · cited 1×
“One of the most common techniques is data augmentation, where data is created virtually [63].”
Research works printed but not linked: 7
- Simonyan, K., and Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv.
- Targ, S., Almeida, D., and Lyman, K. (2016). ResNet in ResNet: Generalizing residual architectures. arXiv.
- Larsson, G., Maire, M., and Shakhnarovich, G. (2016). FractalNet: Ultra-Deep Neural Networks Without Residuals. arXiv.
- Zhao, L., Wang, J., Li, X., Tu, Z., and Zeng, W. (2016). On the connection of deep fusion to ensembling. arXiv.
- Wang, J., Wei, Z., Zhang, T., and Zeng, W. (2016). Deeply-fused nets. arXiv.
- Combalia, M., Codella, N.C., Rotemberg, V., Helba, B., Vilaplana, V., Reiter, O., Carrera, C., Barreiro, A., Halpern, A.C., and Puig, S. (2019). BCN20000: Dermoscopic lesions in the wild. arXiv.
- Lin, M., Chen, Q., and Yan, S. (2013). Network in network. arXiv.
Cited by
Review of deep learning: concepts, CNN architectures, challenges, applications, future directions
Alzubaidi, L., Zhang, J., Humaidi, A. J., Al-Dujaili, A., Duan, Y., Al-Shamma, O., Santamaría, J., Fadhel, M. A., Al-Amidie, M., & Farhan, L. (2021) · Journal of Big Data · doi:10.1186/s40537-021-00444-8 · Cited by 7,906
Self-citationTransfer learning for medical image classification: a literature review
Kim, H. E., Cosa-Linan, A., Santhanam, N., Jannesari, M., Maros, M. E., & Ganslandt, T. (2022) · BMC Medical Imaging · doi:10.1186/s12880-022-00793-7 · Cited by 965
“Also, we did not evaluate aspects of potential model improvements leveraged by the differences of the source-and the target domain of the training data used for TL [60].”
MethodAlzubaidi, L., Bai, J., Al-Sabaawi, A., Santamaría, J., Albahri, A. S., Al-dabbagh, B. S. N., Fadhel, M. A., Manoufali, M., Zhang, J., Al-Timemy, A. H., et al. (2023) · Journal of Big Data · doi:10.1186/s40537-023-00727-2 · Cited by 907
“• Transfer learning (TL): as TL defined previously, the learned model can then be reused in other DL tasks to improve their performance [23, 124].”
BackgroundMethodSelf-citationBig Self-Supervised Models Advance Medical Image Classification
Azizi, S., Mustafa, B., Ryan, F., Beaver, Z., Freyberg, J., Deaton, J., Loh, A., Karthikesalingam, A., Kornblith, S., Chen, T., et al. (2021) · 2021 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00346 · Cited by 685
“Despite the differences in image statistics, scale, and taskrelevant features, transfer learning from natural images is commonly used in medical image analysis [29, 31, 32, 46], and multiple empirical studies show that this improves performance [1, 15, 20].”
BackgroundQiu, S., Zhao, H., Jiang, N., Wang, Z., Liu, L., An, Y., Zhao, H., Miao, X., Liu, R., & Fortino, G. (2022) · Information Fusion · doi:10.1016/j.inffus.2021.11.006 · Cited by 671
Novel Transfer Learning Approach for Medical Imaging with Limited Labeled Data
Alzubaidi, L., Al-Amidie, M., Al-Asadi, A., Humaidi, A. J., Al-Shamma, O., Fadhel, M. A., Zhang, J., Santamaría, J., & Duan, Y. (2021) · Cancers · doi:10.3390/cancers13071590 · Cited by 292
“The authors empirically showed that in-domain TL can improve the performance against the pretrained ImageNet models [12–14].”
BackgroundSelf-citationPham, T. N., Tran, L. V., & Dao, S. V. T. (2020) · IEEE Access · doi:10.1109/access.2020.3031914 · Cited by 262
“Transfer learning (TL) is a solution for the lack of training data in deep learning [33].”
BackgroundRobust and data-efficient generalization of self-supervised machine learning for diagnostic imaging
Azizi, S., Culp, L., Freyberg, J., Mustafa, B., Baur, S., Kornblith, S., Chen, T., Tomasev, N., Mitrović, J., Strachan, P., et al. (2023) · Nature Biomedical Engineering · doi:10.1038/s41551-023-01049-7 · Cited by 250
“Multiple empirical studies suggest that such transfer-learning strategy improves performance 70,159,160 for medical imaging tasks.”
BackgroundTahir, M., Naeem, A., Malik, H., Tanveer, J., Naqvi, R. A., & Lee, S.-W. (2023) · Cancers · doi:10.3390/cancers15072179 · Cited by 240
“There are 33,126 der-moscopic images, 579 images of malignant skin lesions, and 32,542 images of benign skin lesions.”
BackgroundReview of the State of the Art of Deep Learning for Plant Diseases: A Broad Analysis and Discussion
Hasan, R. I., Yusuf, S. M., & Alzubaidi, L. (2020) · Plants · doi:10.3390/plants9101302 · Cited by 213
“In [84,85], the researchers suggested a first round of training using transfer learning from the original and augmented samples of a large-scale dataset which is in the same domain of the targeted dataset, followed by re-training the model using the original and augmented samples of the target dataset.”
BackgroundSelf-citationAhmad, S., Shakeel, I., Mehfuz, S., & Ahmad, J. (2023) · Computer Science Review · doi:10.1016/j.cosrev.2023.100568 · Cited by 197
A Comprehensive Survey of Deep Learning Approaches in Image Processing
Τρίγκα, Μ., & Dritsas, E. (2025) · Sensors · doi:10.3390/s25020531 · Cited by 207
“By fine-tuning pre-trained models on smaller, task-specific datasets, practitioners can achieve significant performance improvements, often surpassing models trained from scratch [69,70].”
A CNN-RNN Combined Structure for Real-World Violence Detection in Surveillance Cameras
Vosta, S., & Yow, K. (2022) · Applied Sciences · doi:10.3390/app12031021 · Cited by 158
“Due to difficulties in collecting and labeling anomalous events, we used Transfer Learning in our model [59].”
ADD-Net: An Effective Deep Learning Model for Early Detection of Alzheimer Disease in MRI Scans
Fareed, M. M. S., Zikria, S., Ahmed, G., Mui-zzud-din, Mahmood, S., Aslam, M., Jilani, S. F., Moustafa, A., & Asad, M. (2022) · IEEE Access · doi:10.1109/access.2022.3204395 · Cited by 144
“The proposed 493 ADD-Net is compared using the AD dataset’s ROC curve 494 with DenseNet169, InceptionResNet V2, and VGG19.”
BackgroundMethodInfluentialAtasever, S., Azgınoğlu, N., Terzi, D. S., & Terzi, R. (2022) · Clinical Imaging · doi:10.1016/j.clinimag.2022.11.003 · Cited by 148
Diabetic foot ulcer detection using deep learning approaches
Thotad, P. N., Bharamagoudar, G. R., & Anami, B. S. (2022) · Sensors International · doi:10.1016/j.sintl.2022.100210 · Cited by 142
Towards Risk‐Free Trustworthy Artificial Intelligence: Significance and Requirements
Alzubaidi, L., Al-Sabaawi, A., Bai, J., Dukhan, A., Alkenani, A. H., Al‐Asadi, A., Alwzwazy, H. A., Manoufali, M., Fadhel, M. A., Albahri, A. S., et al. (2023) · International Journal of Intelligent Systems · doi:10.1155/2023/4459198 · Cited by 147
Self-citationA Comprehensive Survey on Deep-Learning-Based Breast Cancer Diagnosis
Mridha, M. F., Hamid, M. A., Monowar, M. M., Keya, A. J., Ohi, A. Q., Islam, M. R., & Kim, J.-M. (2021) · Cancers · doi:10.3390/cancers13236116 · Cited by 114
Luca, A. R., Ursuleanu, T. F., Gheorghe, L., Grigorovici, R., Iancu, S., Hlusneac, M., & Grigorovici, A. (2022) · Informatics in Medicine Unlocked · doi:10.1016/j.imu.2022.100911 · Cited by 121
Mitigating Bias in Radiology Machine Learning: 2. Model Development
Zhang, K., Khosravi, B., Vahdati, S., Faghani, S., Nugen, F., Rassoulinejad-Mousavi, S. M., Moassefi, M., Jagtap, J. M. M., Singh, Y., Rouzrokh, P., et al. (2022) · Radiology: Artificial Intelligence · doi:10.1148/ryai.220010 · Cited by 82
“On the contrary, transfer learning from in-domain datasets such as ChestX-ray8 (43) and CheXpert (44) has been shown to improve performance compared with ImageNet pretrained models (45).”
ResultApplications of Artificial Intelligent and Machine Learning Techniques in Image Processing
Boopathi, S., & Kanike, U. K. (2023) · Advances in computational intelligence and robotics book series · doi:10.4018/978-1-6684-8618-4.ch010 · Cited by 82
Alzubaidi, L., Al-Dulaimi, K., Salhi, A., Alammar, Z., Fadhel, M. A., Albahri, A. S., Alamoodi, A. H., Albahri, O. S., Hasan, A. F., Bai, J., et al. (2024) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2024.102935 · Cited by 84
Self-citationKumar, A., Yadav, S. P., & Kumar, A. (2025) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2025.109822 · Cited by 78
Supervised Transfer Learning at Scale for Medical Imaging
Mustafa, B., Loh, A., Freyberg, J., MacWilliams, P., Wilson, M., McKinney, S. M., Sieniek, M., Winkens, J., Liu, Y., Bui, P., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2101.05913 · Cited by 21
“In some cases, detailed experiments have been performed to demonstrate the benefits of adopting this transfer [2, 13, 18], but it is not clear if this is always beneficial.”
BackgroundReview on Deep Learning Approaches for Anomaly Event Detection in Video Surveillance
Jebur, S. A., Hussein, K. A., Hoomod, H. K., Alzubaidi, L., & Santamaría, J. (2022) · Electronics · doi:10.3390/electronics12010029 · Cited by 69
Self-citationJiménez, D. R., Casanova-Lozano, L., Grau, S., & Reig-Bolaño, R. (2025) · Journal of Medical Systems · doi:10.1007/s10916-025-02153-8 · Cited by 80
“Moreover, Alzubaidi et al. explored transfer learning in DFU classification, achieving a remarkable F1-score of 97.6% with a dataset of 1,200 images, emphasizing the effectiveness of domain-specific approaches [79].”
InfluentialRobust application of new deep learning tools: an experimental study in medical imaging
Alzubaidi, L., Fadhel, M. A., Al-Shamma, O., Zhang, J., Santamaría, J., & Duan, Y. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-021-10942-9 · Cited by 68
“Having a few images for TL in the same domain as the target dataset is better than having a large number of images in a different domain [4, 47].”
BackgroundMethodSelf-citationSurpassing the Human Accuracy: Detecting Gallbladder Cancer from USG Images with Curriculum Learning
Basu, S., Gupta, M., Rana, P., Gupta, P., & Arora, C. (2022) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52688.2022.02022 · Cited by 87
“Transfer Learning and Data Augmentation: Studies show that pre-training on natural image data improves network performance on medical image data [2, 15].”
BackgroundAlzubaidi, L., Duan, Y., Al-Dujaili, A. Q., Ibraheem, I. K., Alkenani, A. H., Santamaría, J., Fadhel, M. A., Al-Shamma, O., & Zhang, J. (2021) · PeerJ Computer Science · doi:10.7717/peerj-cs.715 · Cited by 73
“More speci fi cally, in the natural image setting ( He, Girshick & Dollár, 2019 ; Kornblith, Shlens & Le, 2019 ; Geirhos et al., 2018 ; Wen et al., 2021 ; Alzubaidi et al., 2020b , 2021a ), the current work-based TL has faced several common dilemmas.”
BackgroundMethodResultSelf-citationZhou, P., Cao, Y., Li, M., Ma, Y., Chen, C., Gan, X., Wu, J., Lv, X., & Chen, C. (2022) · Scientific Reports · doi:10.1038/s41598-022-18879-1 · Cited by 68
Multistage transfer learning for medical images
Ayana, G., Dese, K., Abagaro, A. M., Jeong, K. C., Yoon, S.-D., & Choe, S.-W. (2024) · Artificial Intelligence Review · doi:10.1007/s10462-024-10855-7 · Cited by 57
“Fine-tuning the approach to suit specific medical imaging tasks and domains is essential for maximizing the benefits while mitigating the risks associated with transfer learning in the medical image domain (Alzubaidi et al. 2020; Kora et al. 2022).”
Jones, M. A., Islam, W., Faiz, R., Chen, X., & Zheng, B. (2022) · Frontiers in Oncology · doi:10.3389/fonc.2022.980793 · Cited by 63
“The deeper a model is, the more complex representations can be learned, so the question of how deep a CNN must be to suf fi ciently capture features for a large classi fi cation task remains (128).”
BackgroundPunn, N. S., & Agarwal, S. (2022) · Machine Learning · doi:10.1007/s10994-022-06219-3 · Cited by 72
“Though this strategy works very well with natural images, but is ineffectual in biomedical image analysis (Alzubaidi et al, 2020; Raghu et al, 2019) due to large variation in the associated complex patterns of biomedical imaging as compared to natural images.”
BackgroundFace Recognition Based on Deep Learning and FPGA for Ethnicity Identification
AlBdairi, A. J. A., Xiao, Z., Alkhayyat, A., Humaidi, A. J., Fadhel, M. A., Taher, B. H., Alzubaidi, L., Santamaría, J., & Al-Shamma, O. (2022) · Applied Sciences · doi:10.3390/app12052605 · Cited by 54
Self-citationAutomatic Early Diagnosis of Alzheimer’s Disease Using 3D Deep Ensemble Approach
Gamal, A., Elattar, M., & Selim, S. (2022) · IEEE Access · doi:10.1109/access.2022.3218621 · Cited by 64
“In medical imaging, such as cancer classification, heart function quan-tification, and lung disease classification, transfer learning has yielded promising results [67], [68].”
BackgroundSilva, F., Pereira, T., Morgado, J., Frade, J., Mendes, J., Freitas, C., Negrão, E., de Lima, B. F., da Silva, M. C., Madureira, A. J., et al. (2021) · IEEE Access · doi:10.1109/access.2021.3070701 · Cited by 60
“In this work, given the scarce dataset size available to perform the target task, we considered an alternative approach, consisting of developing and training the feature extractor in the same domain of the final task [25].”
MethodChowdhury, D., Das, A., Dey, A., Sarkar, S., Dwivedi, A. D., Mukkamala, R. R., & Murmu, L. (2022) · Sensors · doi:10.3390/s22030832 · Cited by 52
“[21], they presented a thorough study of how pretraining Transfer Learning models with different datasets can affect the performance of the model.”
BackgroundAlexnet architecture variations with transfer learning for classification of wound images
Eldem, H., Ülker, E., & Işıklı, O. Y. (2023) · Engineering Science and Technology an International Journal · doi:10.1016/j.jestch.2023.101490 · Cited by 69
Colon Tissues Classification and Localization in Whole Slide Images Using Deep Learning
Gupta, P., Huang, Y., Sahoo, P. K., You, J., Chiang, S.-F., Onthoni, D. D., Chern, Y., Chao, K., Chiang, J., Yeh, C.-Y., et al. (2021) · Diagnostics · doi:10.3390/diagnostics11081398 · Cited by 55
“On the other hand, the work in [13] experimentally proved that TL considering a model with the same domain of the targeted image achieved better accuracy than TL considering a model with the different domain of the target dataset.”
BackgroundBadgujar, C., Armstrong, P. R., Gerken, A. R., Pordesimo, L. O., & Campbell, J. F. (2023) · Journal of Stored Products Research · doi:10.1016/j.jspr.2023.102196 · Cited by 50
A Novel Method for the Classification of Butterfly Species Using Pre-Trained CNN Models
P., F. R. P., Orban, R., Kogilavani, S. V., Subramanian, M., Muthusamy, S., AbdElminaam, D. S., Nabil, A. M., Abulaigh, L., Ahmadi, M., & Ali, M. A. S. (2022) · Electronics · doi:10.3390/electronics11132016 · Cited by 47
COVID-19 detection from scarce chest x-ray image data using few-shot deep learning approach
Jadon, S. (2021) · SPIE Medical Imaging: Imaging Informatics for Healthcare, Research, and Applications · doi:10.1117/12.2581496 · Cited by 49
Biswas, S. K., Mostafiz, R., Paul, B. K., Uddin, K. M. M., Hadi, M. A., & Khanom, F. (2024) · Biomedical Materials & Devices · doi:10.1007/s44174-024-00165-5 · Cited by 46
“The dataset comprises four folders listed in Kaggle online data repositories [19], and the source of this dataset is the works [10, 20].”
Amended Convolutional Neural Network with Global Average Pooling for Image Classification
Al-Sabaawi, A., Ibrahim, H. M., Arkah, Z. M., Al‐Amidie, M., & Alzubaidi, L. (2021) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-71187-0_16 · Cited by 50
Self-citationDas, S. K., Roy, P., & Mishra, A. K. (2021) · International Journal of Imaging Systems and Technology · doi:10.1002/ima.22598 · Cited by 45
“In another work, 22 the authors have proposed a hybrid deep CNN architecture to explore the benefits of domain-specific transfer learning.”
BackgroundTrustworthy deep learning framework for the detection of abnormalities in X-ray shoulder images
Alzubaidi, L., Salhi, A., Fadhel, M. A., Bai, J., Hollman, F., Italia, K. R., Pareyón, R., Albahri, A. S., Ouyang, C., Santamaría, J., et al. (2024) · PLoS ONE · doi:10.1371/journal.pone.0299545 · Cited by 41
Self-citationIrkham, I., Ibrahim, A. U., Nwekwo, C. W., Al-Turjman, F. M., & Hartati, Y. W. (2022) · Sensors · doi:10.3390/s23010426 · Cited by 51
“The use of TL models, also known as pretrained models, has shown to outperform models developed from scratch [120,121].”
MethodDeep Representation Learning: Fundamentals, Technologies, Applications, and Open Challenges
Payandeh, A., Baghaei, K. T., Fayyazsanavi, P., Ramezani, S. B., Chen, Z., & Rahimi, S. (2023) · IEEE Access · doi:10.1109/access.2023.3335196 · Cited by 64
“In [449], [450], [451], and [452] the authors try to transfer the knowledge learned from the pre-trained models, Resnet [453] or ALexNet [454] trained on ImageNet, and transfer it for different tasks like Brain Tumor Segmentation, 3d medical image analysis and Alzheimer.”
BackgroundReyes-Luévano, J., Guerrero-Viramontes, J. A., Romo-Andrade, J. R., & Funes-Gallanzi, M. (2023) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2023.105341 · Cited by 39
Enhanced Visual Attention-Guided Deep Neural Networks for Image Classification
Yeh, C., Lin, M.-H., Chang, P.-C., & Kang, L.-W. (2020) · IEEE Access · doi:10.1109/access.2020.3021729 · Cited by 39
“On the other hand, the selections of both the public Animals-10 dataset [33], also used in recent works, e.g., [40], [41], and the public Stanford Cars dataset [34], also used in recent works, e.g., [42], [43], are mainly based on the former two principles.”
MethodAzam, M. A., Sampieri, C., Ioppi, A., Benzi, P., Giordano, G., Vecchi, M. D., Campagnari, V., Li, S., Guastini, L., Paderno, A., et al. (2022) · Frontiers in Oncology · doi:10.3389/fonc.2022.900451 · Cited by 45
“While most methods for medical image analysis employ TL from general natural images (e.g., ImageNet) (32) this strategy has been proven to be less effective compared to in-domain TL due to the mismatch in learned features between natural and medical images (50).”
MethodGhafariasl, P., Zeinalnezhad, M., & Chang, S. I. (2024) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2024.109666 · Cited by 42
Accurate Deep Learning-based Gastrointestinal Disease Classification via Transfer Learning Strategy
Escobar, J., Sánchez, K., Hinojosa, C. A., Argüello, H., & Castillo, S. (2021) · Symposium on Image, Signal Processing and Artificial Vision (STSIVA) · doi:10.1109/stsiva53688.2021.9591995 · Cited by 38
“One of the main challenges of employing deep learning models in the medical area is the lack of training data [22].”
BackgroundKhanna, N. N., Maindarkar, M. A., Viswanathan, V., Puvvula, A., Paul, S., Bhagawati, M., Ahluwalia, P., Ruzsa, Z., Sharma, A., Kolluri, R., et al. (2022) · Journal of Clinical Medicine · doi:10.3390/jcm11226844 · Cited by 41
Deep Transfer Learning with Enhanced Feature Fusion for Detection of Abnormalities in X-ray Images
Alammar, Z., Alzubaidi, L., Zhang, J., Li, Y., Lafta, W. M., & Gu, Y. (2023) · Cancers · doi:10.3390/cancers15154007 · Cited by 45
“This distinction between colour and grayscale images highlights the need for careful consideration and customisation when applying TL techniques to medical imaging tasks, which often involve specific imaging modalities and grayscale representations [16].”
BackgroundMethodSelf-citationAutomated brain tumor classification using various deep learning models: a comparative study
Abbood, A. A., Shallal, Q. M., & Fadhel, M. A. (2021) · Indonesian Journal of Electrical Engineering and Computer Science · doi:10.11591/ijeecs.v22.i1.pp252-259 · Cited by 49
“CNNs are heavy task and required high computational tools to handle them such as GPUs [31-34] and FPGA [23, 35-37].”
BackgroundMethodSelf-citationLung cancer histopathological image classification using wavelets and AlexNet
Sethy, P. K., Devi, A. G., Padhan, B., Behera, S. K., Sreedhar, S., & Das, K. (2022) · Journal of X-Ray Science and Technology · doi:10.3233/xst-221301 · Cited by 42
“Transfer learning is another name for this method [30].”
Deep Learning-Based Classification of Chest Diseases Using X-rays, CT Scans, and Cough Sound Images
Malik, H., Anees, T., Al‐Shamayleh, A. S., Alharthi, S. Z., Khalil, W., & Akhunzada, A. (2023) · Diagnostics · doi:10.3390/diagnostics13172772 · Cited by 43
Deep learning-based CNN for multiclassification of ocular diseases using transfer learning
Deepak, G. D., & Bhat, S. K. (2024) · Computer Methods in Biomechanics and Biomedical Engineering Imaging & Visualization · doi:10.1080/21681163.2024.2335959 · Cited by 35
“Transfer learning implements CNN models that have already been trained on challenging datasets such as ImageNet (Soekhoe et al. 2016; Alzubaidi et al. 2020; Morid et al. 2021; Gupta et al. 2022).”
Karthik, R., Ajay, A., Jhalani, A., Ballari, K., & Suganthi, K. (2025) · Scientific Reports · doi:10.1038/s41598-025-87519-1 · Cited by 42
Jordanian banknote data recognition: A CNN-based approach with attention mechanism
Nasayreh, A. N., Jaradat, A. S., Gharaibeh, H., Dawaghreh, W., Mamlook, R. E. A., Alqudah, Y. E., Al-Na’amneh, Q., Daoud, M. S., Migdady, H., & Abualigah, L. (2024) · Journal of King Saud University - Computer and Information Sciences · doi:10.1016/j.jksuci.2024.102038 · Cited by 41
“It has such a great reputation and attention in research since it has the ability to transfer already-gained knowledge, allowing researchers not to worry about the need for a massive dataset and expensive learning time (Alzubaidi, Jan. 2020).”
Ensemble of Deep Learning Models for Multi-plant Disease Classification in Smart Farming
Vo, H.-T., Quach, L.-D., & Ngoc, H. T. (2023) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2023.01405108 · Cited by 34
“Agriculture imaging tasks [32], [5], [26], [33], [31]; Medical imaging tasks [34], [2], [47], etc.”
BackgroundMartis, J. E., Sannidhan, M. S., Balasubramani, R., Mutawa, A. M., & Murugappan, M. (2024) · Bioengineering · doi:10.3390/bioengineering11080799 · Cited by 35
Mushroom Image Classification with CNNs: A Case-Study of Different Learning Strategies
Kiss, N., & Czúni, L. (2021) · International symposium on image and signal processing and analysis/ISPA ... · doi:10.1109/ispa52656.2021.9552053 · Cited by 32
MedNet: Pre-trained Convolutional Neural Network Model for the Medical Imaging TasksPreprint
Alzubaidi, L., Santamaría, J., Manoufali, M., Mohammed, B. J., Fadhel, M. A., Zhang, J., Al-Timemy, A. H., Al-Shamma, O., & Duan, Y. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2110.06512 · Cited by 31
“Recently, many researchers have demonstrated that the use of TL in medical image classification tasks is effective and efficient [4,5,6].”
BackgroundMethodInfluentialSelf-citationTowards Home-Based Diabetic Foot Ulcer Monitoring: A Systematic Review
Kairys, A., Pauliukienė, R., Raudonis, V., & Čeponis, J. (2023) · Sensors · doi:10.3390/s23073618 · Cited by 30
“Networks within the last group were designed by using concepts of 3D [26] and parallel convolutions [30,47], repeatable network modules [44,81], and other concepts [66].”
MethodInfluentialArtificial Intelligence in Oncology: Applications, Challenges and Future Frontiers
Alsharif, F. H. (2024) · International Journal of Pharmaceutical Investigation · doi:10.5530/ijpi.14.3.76 · Cited by 35
“31,32 Transfer learning has enabled the advancement of breast cancer diagnosis through ultrasound imaging, addressing the common issue of acquiring an extensive training dataset.”
Jones, M. A., Faiz, R., Qiu, Y., & Zheng, B. (2022) · Physics in Medicine and Biology · doi:10.1088/1361-6560/ac5297 · Cited by 40
Supervised fine-tuned approach for automated detection of diabetic retinopathy
Ohri, K., & Kumar, M. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-16049-7 · Cited by 33
Compatible-domain Transfer Learning for Breast Cancer Classification with Limited Annotated Data
Shamshiri, M., Krzyżak, A., Kowal, M., & Korbicz, J. (2023) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2023.106575 · Cited by 28
Applications of machine learning in time-domain fluorescence lifetime imaging: a review
Gouzou, D., Taimori, A., Haloubi, T., Finlayson, N., Wang, Q., Hopgood, J. R., & Vallejo, M. (2023) · Methods and Applications in Fluorescence · doi:10.1088/2050-6120/ad12f7 · Cited by 31
A Transfer Learning Approach for Lumbar Spine Disc State Classification
Al-Kubaisi, A., & Khamiss, N. N. (2021) · Electronics · doi:10.3390/electronics11010085 · Cited by 26
Alam, M. N., Yamashita, R., Ramesh, V., Prabhune, T., Lim, J. I., Chan, R. V. P., Hallak, J., Leng, T., & Rubin, D. L. (2023) · Scientific Reports · doi:10.1038/s41598-023-33365-y · Cited by 30
“While transfer learning through Imagenet can be useful, recent work has shown that the role of Imagenet weights for medical imaging tasks can be limited [32, 33] since it was developed and validated”
BackgroundMahmood, S., Fareed, M. M. S., Ahmed, G., Dawood, F., Zikria, S., Mostafa, A., Jilani, S. F., Asad, M., & Aslam, M. (2022) · Biomedicines · doi:10.3390/biomedicines10092195 · Cited by 24
Hernández-Guedes, A., Santana-Pérez, I., Arteaga-Marrero, N., Fabelo, H. A., Callicó, G. M., & Ruiz‐Alzola, J. (2022) · IEEE Access · doi:10.1109/access.2022.3225107 · Cited by 24
“Transfer learning should be applied in the same domain of the target dataset [50].”
BackgroundVerma, G. (2024) · Polish Journal of Radiology · doi:10.5114/pjr/189412 · Cited by 24
“Medical imaging [8-11] is a cornerstone for diagnosing various patient conditions in today’s digital healthcare landscape.”
BackgroundAlzubaidi, L., Fadhel, M. A., Hollman, F., Salhi, A., Santamaría, J., Duan, Y., Gupta, A., Cutbush, K., Abbosh, A. M., & Gu, Y. (2024) · Artificial Intelligence Review · doi:10.1007/s10462-024-10878-0 · Cited by 23
“Alzubaidi et al. (2020, 2021) proposed two different types of TL which showed excellent results in several medical applications.”
Self-citationMirjebreili, S. M., Shalbaf, R., & Shalbaf, A. (2024) · Physical and Engineering Sciences in Medicine · doi:10.1007/s13246-024-01392-2 · Cited by 28
“In order to overcome this problem, transfer learning utilizes knowledge acquired for one task for another task [26–28].”
FocusMAE: Gallbladder Cancer Detection from Ultrasound Videos with Focused Masked Autoencoders
Basu, S., Gupta, M., Madan, C., Gupta, P. K., & Arora, C. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.01113 · Cited by 28
“Although there is a domain gap in natural and medical image data, studies show that pretraining on natural image data improves network performance on medical imaging tasks [2, 12].”
BackgroundBasu, S., Singla, S., Gupta, M., Rana, P., Gupta, P., & Arora, C. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-16440-8_41 · Cited by 24
Zhang, A.-Q., Zhao, H.-P., Li, F., Liang, P., Gao, J., & Cheng, M. Y. (2022) · Frontiers in Oncology · doi:10.3389/fonc.2022.969707 · Cited by 22
Transfer-Learning-Based Approach for the Diagnosis of Lung Diseases from Chest X-ray Images
Fan, R., & Bu, S. (2022) · Entropy · doi:10.3390/e24030313 · Cited by 22
“A deep convolutional neural network (DCNN) was proposed to study the advantages of transfer learning in medicine [15].”
BackgroundSelf-supervised endoscopic image key-points matching
Farhat, M., Chaabouni-Chouayakh, H., & Hamadou, A. B. (2022) · Expert Systems with Applications · doi:10.1016/j.eswa.2022.118696 · Cited by 23
Wang, X., Gao, M., Xie, J., Deng, Y., Tu, W., Yang, H., Liang, S., Xu, P., Zhang, M., Lu, Y., et al. (2022) · Frontiers in Oncology · doi:10.3389/fonc.2022.892890 · Cited by 21
“As proved by many studies (38 – 40), the transfer learning strategy can effectively train a well performing classi fi cation model with relatively small sample size.”
BackgroundA systematic approach to deep learning-based nodule detection in chest radiographs
Behrendt, F., Bengs, M., Bhattacharya, D., Krüger, J., Opfer, R., & Schlaefer, A. (2023) · Scientific Reports · doi:10.1038/s41598-023-37270-2 · Cited by 26
Latif, G., Morsy, H., Hassan, A., & Alghazo, J. M. (2022) · Viruses · doi:10.3390/v14081667 · Cited by 22
“A well-known CNN is GoogleNet, consisting of 22 deep layers with computational resources utilizing inception modules [42].”
MethodUrsuleanu, T. F., Luca, A. R., Gheorghe, L., Grigorovici, R., Iancu, S., Hlusneac, M., Preda, C., & Grigorovici, A. (2021) · Diagnostics · doi:10.3390/diagnostics11081373 · Cited by 26
Mazhar, F., Aslam, N., Naeem, A., Ahmad, H., Fuzail, M., & Imran, M. (2025) · Journal of Computing Theories and Applications · doi:10.62411/jcta.11954 · Cited by 25
An explainable AI for breast cancer classification using vision Transformer (ViT)
Naas, M., Mzoughi, H., Njeh, I., & BenSlima, M. (2025) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2025.108011 · Cited by 23
Jang, S. J., Flevas, D. A., Kunze, K. N., Anderson, C. G., Fontana, M. A., Boettner, F., Sculco, T. P., Baldini, A., & Sculco, P. K. (2023) · The Journal of Arthroplasty · doi:10.1016/j.arth.2023.02.015 · Cited by 17
Enhancing Transfer Learning for Medical Image Classification with SMOTE: A Comparative Study
Alam, M. Z., Roy, T., Kawsar, H. M. N., & Rimi, I. F. (2024) · International Conference on Computer and Information Technology (ICCIT) · doi:10.1109/iccit64611.2024.11022326 · Cited by 17
“Transfer Learning (TL) offers a solution by leveraging pre-trained models on extensive, varied datasets, improving their adaptability to specialized tasks like medical image classification [8].”
Semantic Segmentation of Periocular Near-Infra-Red Eye Images Under Alcohol Effects
Tapia, J. E., Droguett, E. L., Valenzuela, A., Benalcazar, D., Causa, L., & Busch, C. (2021) · IEEE Access · doi:10.1109/access.2021.3101256 · Cited by 21
“Recently, automatic pupil segmentation is attracting many researchers to find the precise measurement of the pupil radii to apply to biometric analytics as part of medical, entertainment applications, virtual reality lenses, among others [9], [10].”
BackgroundAn Overview of Deep Learning Based Small Sample Medical Imaging Classification
Wang, K. (2021) · doi:10.1109/conf-spml54095.2021.00060 · Cited by 22
Developing an image-based deep learning framework for automatic scoring of the pentagon drawing test
Li, Y., Guo, J., & Yang, P. (2022) · Journal of Alzheimer's Disease · doi:10.3233/jad-210714 · Cited by 16
“However, recentstudieshaveshownthatsuchanapproachmay notalwaysbeeffective,especiallyonimagesthatare distinct from the previous training data (e.g., transfer learning from natural images to medical images) [23–25].”
Fazilov, S. K., Abdieva, K., Yusupov, O. R., Yusupov, R., & Qodirov, A. (2024) · AIP conference proceedings · doi:10.1063/5.0241418 · Cited by 15
Nieniewski, M. J., Chmielewski, L. J., Patrzyk, S., & Woźniacka, A. J. (2023) · EURASIP Journal on Image and Video Processing · doi:10.1186/s13640-023-00607-y · Cited by 17
“Alzubaidi et al. [10] proposed an intelligent medical system for the diagnosis of diabetic foot ulcer (DFU).”
BackgroundToofanee, M. S. A., Dowlut, S., Hamroun, M., Tamine, K., Duong, A. K., Petit, V., & Sauveron, D. (2023) · Applied Sciences · doi:10.3390/app131810310 · Cited by 15
A Deep Convolutional Neural Network Model for Multi-class Fruits Classification
Alzubaidi, L., Al-Shamma, O., Fadhel, M. A., Arkah, Z. M., & Awad, F. H. (2020) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-49342-4_9 · Cited by 14
Self-citationFarhad, M., Masud, M. M., Beg, A., Ahmad, A., Ahmed, L. A., & Memon, S. (2023) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2023.107129 · Cited by 14
Sarwar, N., Al-Otaibi, S. T., & Irshad, A. (2025) · International Journal of Imaging Systems and Technology · doi:10.1002/ima.70033 · Cited by 17
“TL has been used extensively in medical image analysis since it allows using models pretrained on vast datasets like ImageNet and fine-tuning them for particular tasks in question with scarce data [9, 10].”
Metadata
- Authors
- Alzubaidi, L., Fadhel, M. A., Al-Shamma, O., Zhang, J., Santamaría, J., Duan, Y., & R. Oleiwi, S.
- Affiliations
- Queensland University of Technology (AU)
- University of Information Technology and Communications (IQ)
- Universidad de Jaén (ES)
- University of Missouri (US)
- Nursing College, Muthanna University, Muthanna 76218, Iraq
- Corresponding author
- Alzubaidi, Laith
- Venue
- Applied Sciences
- Publisher
- MDPI AG
- Published
- 2020-06-29
- Volume
- 10
- Issue
- 13
- Pages
- 4523
- ISSN
- 2076-3417 (ISSN-L)
- OpenAlex
- W3038835370
- Cited by
- 258
- References
- 62
- Citations by year
- 2026: 27 · 2025: 30 · 2024: 59 · 2023: 54 · 2022: 49 · 2021: 34 · 2020: 5
- FWCI
- 13.95
- Type
- journal article
- Language
- en
- Open access
- gold
- dx.doi.org
- Licence
- Locations
- Applied Sciences (publishedVersion, cc-by)
- QUT ePrints (Queensland University of Technology) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Keywords
- transfer learning, diagnostic imaging, diabetic foot ulcer classification, domain shift, ImageNet pre-trained models, convolutional neural networks, skip connections, global average pooling, limited training data, animal image classification
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.