Novel Transfer Learning Approach for Medical Imaging with Limited Labeled Data

@article{Alzubaidi2021NovelTL,
  title={Novel Transfer Learning Approach for Medical Imaging with Limited Labeled Data},
  author={Laith Alzubaidi and Muthana Al-Amidie and Ahmed Al-Asadi and Amjad J. Humaidi and Omran Al-Shamma and Mohammed Abdulraheem Fadhel and Jinglan Zhang and Jesus Santamar{\'i}a and Ye Duan},
  journal={Cancers},
  year={2021},
  volume={13}
}
Deep learning requires a large amount of data to perform well. However, the field of medical image analysis suffers from a lack of sufficient data for training deep learning models. Moreover, medical images require manual labeling, usually provided by human annotators coming from various backgrounds. More importantly, the annotation process is time-consuming, expensive, and prone to errors. Transfer learning was introduced to reduce the need for the annotation process by transferring the deep… 
MedNet: Pre-trained Convolutional Neural Network Model for the Medical Imaging Tasks
TLDR
A novel DL model to be used for addressing classification tasks of medical imaging, called MedNet, which performs as the pre-trained model to tackle any real-world application from medical imaging and achieve the level of generalization needed for dealing with medical imaging tasks, e.g. classification.
Transfer learning for medical image classification: a literature review
TLDR
The investigated studies demonstrated the efficacy of transfer learning despite the data scarcity and encourage data scientists and practitioners to use deep models as feature extractors, which can save computational costs and time without degrading the predictive power.
Deepening into the suitability of using pre-trained models of ImageNet against a lightweight convolutional neural network in medical imaging: an experimental study
TLDR
Surprisingly, it has been found that the lightweight model trained from scratch achieved a more competitive performance when compared to the pre-trained model and more importantly, the lightweight CNN model can be successfully trained and tested using basic computational tools and provide high-quality results, specifically when using medical imaging datasets.
Big transfer learning for automated skin cancer classification
TLDR
A novel approach of transfer learning is introduced by training the pre-trained models of the ImageNet (VGG, GoogleNet, and ResNet50) on a large number of unlabelled skin cancer images, first, and then train them on a small number of labeled skin images.
Deep Learning Application for Analyzing of Constituents and Their Correlations in the Interpretations of Medical Images
TLDR
The novelty in this paper consists primarily in the unitary approach, of the constituent elements of DL models, namely, data, tools used by DL architectures or specifically constructed DL architecture combinations and highlighting their “key” features, for completion of tasks in current applications in the interpretation of medical images.
ABCanDroid: A Cloud Integrated Android App for Noninvasive Early Breast Cancer Detection Using Transfer Learning
TLDR
The proposed framework aims to be an efficient tool for all doctors and society as a whole and help the user in early detection of breast cancer and provide an accuracy of 99.58%.
Natural Images Allow Universal Adversarial Attacks on Medical Image Classification Using Deep Neural Networks with Transfer Learning
TLDR
This work investigated whether medical DNN models with transfer learning are vulnerable to universal adversarial perturbations (UAPs), generated using natural images, and found that UAPs from natural images are useful for both non-targeted and targeted attacks.
Long-distance contextual attention network for skin disease segmentation
TLDR
A long-distance contextual attention network (LCA-Net) is proposed, which connects the non-local module and the channel attention (CAM) in parallel to form a non- local operation to enhance the network’s ability to extract features of skin diseases.
A Transfer Learning Approach for Lumbar Spine Disc State Classification
TLDR
It has been demonstrated that using transfer learning from the same domain as the target dataset may increase performance dramatically, and applying the ROI method improved the disc state classification results in VGG19 2%, ResNet50 16%, MobileNetV2 5%, and VGG16 2%.
Automating cell counting in fluorescent microscopy through deep learning with c-ResUnet
TLDR
A Deep Learning approach that exploits a fully-convolutional network in a binary segmentation fashion to localize the objects of interest in fluorescent microscopy, suggesting human-level performance inasmuch even erroneous predictions seem to fall within the limits of operator interpretation.
...
1
2
3
4
...

References

SHOWING 1-10 OF 76 REFERENCES
Towards a Better Understanding of Transfer Learning for Medical Imaging: A Case Study
TLDR
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 is designed that can significantly improve the performance considering a reduced number of images in the same domain of the target dataset.
Transfusion: Understanding Transfer Learning for Medical Imaging
TLDR
Investigating the learned representations and features finds that some of the differences from transfer learning are due to the over-parametrization of standard models rather than sophisticated feature reuse, and isolate where useful feature reuse occurs, and outline the implications for more efficient model exploration.
Med3D: Transfer Learning for 3D Medical Image Analysis
TLDR
A heterogeneous 3D network called Med3D is designed to co-train multi-domain 3DSeg-8 so as to make a series of pre-trained models which can accelerate the training convergence speed of target 3D medical tasks and improve accuracy ranging from 3% to 20%.
Joint Liver Lesion Segmentation and Classification via Transfer Learning
TLDR
It is shown that a simple joint framework outperforms the commonly used multi-task architecture (Y-Net), achieving an improvement of 10% in classification accuracy, compared to a 3% improvement with Y-Net.
Optimizing the Performance of Breast Cancer Classification by Employing the Same Domain Transfer Learning from Hybrid Deep Convolutional Neural Network Model
TLDR
A hybrid model of parallel convolutional layers and residual links is utilized to classify hematoxylin–eosin-stained breast biopsy images into four classes: invasive carcinoma, in-situ carcinomas, benign tumor and normal tissue.
CheXpert: A Large Chest Radiograph Dataset with Uncertainty Labels and Expert Comparison
TLDR
A labeler is designed to automatically detect the presence of 14 observations in radiology reports, capturing uncertainties inherent in radiograph interpretation, in CheXpert, a large dataset that contains 224,316 chest radiographs of 65,240 patients.
Classification of Breast Cancer Histology Images Through Transfer Learning Using a Pre-trained Inception Resnet V2
TLDR
A deep neural network approach using transfer learning to classify breast cancer histology images using an Inception Resnet V2 and the accuracy is 0.76 in the blind test set.
Automatic Skin Lesion Segmentation Using Deep Fully Convolutional Networks With Jaccard Distance
TLDR
A fully automatic method for skin lesion segmentation by leveraging 19-layer deep convolutional neural networks that is trained end-to-end and does not rely on prior knowledge of the data to ensure effective and efficient learning with limited training data is presented.
Dual Network Architecture for Few-view CT - Trained on ImageNet Data and Transferred for Medical Imaging
TLDR
A dual network architecture (DNA) for reconstructing images directly from sinograms using natural images can be used to pre-train DNA to avoid overfitting when the amount of real patient images is limited.
...
1
2
3
4
5
...