Convolutional neural networks (CNNs) have been successfully applied to medical image classification, segmentation, and related tasks. Among the many CNNs architectures, U-Net and its improved versions based are widely used and achieve state-of-the-art performance these years. These improved architectures focus on structural improvements and the size of the convolution kernel is generally fixed. In this paper, we propose a module that combines the benefits of multiple kernel sizes and we apply the proposed module to U-Net and its variants. We test our module on three segmentation benchmark datasets and experimental results show significant improvement.
Deep learning for medical image classification faces three major challenges: 1) the number of annotated medical images for training are usually small; 2) regions of interest (ROIs) are relatively small with unclear boundaries in the whole medical images, and may appear in arbitrary positions across the x,y (and also z in 3D images) dimensions. However often only labels of the whole images are annotated, and localized ROIs are unavailable; and 3) ROIs in medical images often appear in varying sizes (scales). We approach these three challenges with a Multi-Instance Multi-Scale (MIMS) CNN: 1) We propose a multi-scale convolutional layer, which extracts patterns of different receptive fields with a shared set of convolutional kernels, so that scale-invariant patterns are captured by this compact set of kernels. As this layer contains only a small number of parameters, training on small datasets becomes feasible; 2) We propose a "top-k pooling" to aggregate the feature maps in varying scales from multiple spatial dimensions, allowing the model to be trained using weak annotations within the multiple instance learning (MIL) framework. Our method is shown to perform well on three classification tasks involving two 3D and two 2D medical image datasets.
Medical image analysis using supervised deep learning methods remains problematic because of the reliance of deep learning methods on large amounts of labelled training data. Although medical imaging data repositories continue to expand there has not been a commensurate increase in the amount of annotated data. Hence, we propose a new unsupervised feature learning method that learns feature representations to then differentiate dissimilar medical images using an ensemble of different convolutional neural networks (CNNs) and K-means clustering. It jointly learns feature representations and clustering assignments in an end-to-end fashion. We tested our approach on a public medical dataset and show its accuracy was better than state-of-the-art unsupervised feature learning methods and comparable to state-of-the-art supervised CNNs. Our findings suggest that our method could be used to tackle the issue of the large volume of unlabelled data in medical imaging repositories.
Convolutional Neural Networks (ConvNets) are commonly developed at a fixed resource budget, and then scaled up for better accuracy if more resources are available. In this paper, we systematically study model scaling and identify that carefully balancing network depth, width, and resolution can lead to better performance. Based on this observation, we propose a new scaling method that uniformly scales all dimensions of depth/width/resolution using a simple yet highly effective compound coefficient. We demonstrate the effectiveness of this method on scaling up MobileNets and ResNet. To go even further, we use neural architecture search to design a new baseline network and scale it up to obtain a family of models, called EfficientNets, which achieve much better accuracy and efficiency than previous ConvNets. In particular, our EfficientNet-B7 achieves state-of-the-art 84.3% top-1 accuracy on ImageNet, while being 8.4x smaller and 6.1x faster on inference than the best existing ConvNet. Our EfficientNets also transfer well and achieve state-of-the-art accuracy on CIFAR-100 (91.7%), Flowers (98.8%), and 3 other transfer learning datasets, with an order of magnitude fewer parameters. Source code is at https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet.
Annotated training data insufficiency remains to be one of the challenges of applying deep learning in medical data classification problems. Transfer learning from an already trained deep convolutional network can be used to reduce the cost of training from scratch and to train with small training data for deep learning. This raises the question of whether we can use transfer learning to overcome the training data insufficiency problem in deep learning based medical data classifications. Deep convolutional networks have been achieving high performance results on the ImageNet Large Scale Visual Recognition Competition (ILSVRC) image classification challenge. One example is the Inception-V3 model that was the first runner up on the ILSVRC 2015 challenge. Inception modules that help to extract different sized features of input images in one level of convolution are the unique features of the Inception-V3. In this work, we have used a pretrained Inception-V3 model to take advantage of its Inception modules for Diabetic Retinopathy detection. In order to tackle the labelled data insufficiency problem, we sub-sampled a smaller version of the Kaggle Diabetic Retinopathy classification challenge dataset for model training, and tested the model's accuracy on a previously unseen data subset. Our technique could be used in other deep learning based medical image classification problems facing the challenge of labeled training data insufficiency.
In this paper, we proposed Transferable Ranking Convolutional Neural Network (TRk-CNN) that can be effectively applied when the classes of images to be classified show a high correlation with each other. The multi-class classification method based on the softmax function, which is generally used, is not effective in this case because the inter-class relationship is ignored. Although there is a Ranking-CNN that takes into account the ordinal classes, it cannot reflect the inter-class relationship to the final prediction. TRk-CNN, on the other hand, combines the weights of the primitive classification model to reflect the inter-class information to the final classification phase. We evaluated TRk-CNN in glaucoma image dataset that was labeled into three classes: normal, glaucoma suspect, and glaucoma eyes. Based on the literature we surveyed, this study is the first to classify three status of glaucoma fundus image dataset into three different classes. We compared the evaluation results of TRk-CNN with Ranking-CNN (Rk-CNN) and multi-class CNN (MC-CNN) using the DenseNet as the backbone CNN model. As a result, TRk-CNN achieved an average accuracy of 92.96%, specificity of 93.33%, sensitivity for glaucoma suspect of 95.12% and sensitivity for glaucoma of 93.98%. Based on average accuracy, TRk-CNN is 8.04% and 9.54% higher than Rk-CNN and MC-CNN and surprisingly 26.83% higher for sensitivity for suspicious than multi-class CNN. Our TRk-CNN is expected to be effectively applied to the medical image classification problem where the disease state is continuous and increases in the positive class direction.
Representing features at multiple scales is of great importance for numerous vision tasks. Recent advances in backbone convolutional neural networks (CNNs) continually demonstrate stronger multi-scale representation ability, leading to consistent performance gains on a wide range of applications. However, most existing methods represent the multi-scale features in a layer-wise manner. In this paper, we propose a novel building block for CNNs, namely Res2Net, by constructing hierarchical residual-like connections within one single residual block. The Res2Net represents multi-scale features at a granular level and increases the range of receptive fields for each network layer. The proposed Res2Net block can be plugged into the state-of-the-art backbone CNN models, e.g., ResNet, ResNeXt, and DLA. We evaluate the Res2Net block on all these models and demonstrate consistent performance gains over baseline models on widely-used datasets, e.g., CIFAR-100 and ImageNet. Further ablation studies and experimental results on representative computer vision tasks, i.e., object detection, class activation mapping, and salient object detection, further verify the superiority of the Res2Net over the state-of-the-art baseline methods. The source code and trained models are available on https://mmcheng.net/res2net/.
Convolutional Neural Networks (CNNs) have revolutionized performances in several machine learning tasks such as image classification, object tracking, and keyword spotting. However, given that they contain a large number of parameters, their direct applicability into low resource tasks is not straightforward. In this work, we experiment with an application of CNN models to gastrointestinal landmark classification with only a few thousands of training samples through transfer learning. As in a standard transfer learning approach, we train CNNs on a large external corpus, followed by representation extraction for the medical images. Finally, a classifier is trained on these CNN representations. However, given that several variants of CNNs exist, the choice of CNN is not obvious. To address this, we develop a novel metric that can be used to predict test performances, given CNN representations on the training set. Not only we demonstrate the superiority of the CNN based transfer learning approach against an assembly of knowledge driven features, but the proposed metric also carries an 87% correlation with the test set performances as obtained using various CNN representations.
The accuracy and robustness of image classification with supervised deep learning are dependent on the availability of large-scale, annotated training data. However, there is a paucity of annotated data available due to the complexity of manual annotation. To overcome this problem, a popular approach is to use transferable knowledge across different domains by: 1) using a generic feature extractor that has been pre-trained on large-scale general images (i.e., transfer-learned) but which not suited to capture characteristics from medical images; or 2) fine-tuning generic knowledge with a relatively smaller number of annotated images. Our aim is to reduce the reliance on annotated training data by using a new hierarchical unsupervised feature extractor with a convolutional auto-encoder placed atop of a pre-trained convolutional neural network. Our approach constrains the rich and generic image features from the pre-trained domain to a sophisticated representation of the local image characteristics from the unannotated medical image domain. Our approach has a higher classification accuracy than transfer-learned approaches and is competitive with state-of-the-art supervised fine-tuned methods.
Deep neural networks (DNNs) have produced state-of-the-art results in many benchmarks and problem domains. However, the success of DNNs depends on the proper configuration of its architecture and hyperparameters. Such a configuration is difficult and as a result, DNNs are often not used to their full potential. In addition, DNNs in commercial applications often need to satisfy real-world design constraints such as size or number of parameters. To make configuration easier, automatic machine learning (AutoML) systems for deep learning have been developed, focusing mostly on optimization of hyperparameters. This paper takes AutoML a step further. It introduces an evolutionary AutoML framework called LEAF that not only optimizes hyperparameters but also network architectures and the size of the network. LEAF makes use of both state-of-the-art evolutionary algorithms (EAs) and distributed computing frameworks. Experimental results on medical image classification and natural language analysis show that the framework can be used to achieve state-of-the-art performance. In particular, LEAF demonstrates that architecture optimization provides a significant boost over hyperparameter optimization, and that networks can be minimized at the same time with little drop in performance. LEAF therefore forms a foundation for democratizing and improving AI, as well as making AI practical in future applications.
Deep neural networks (DNNs) have achieved great success in a wide variety of medical image analysis tasks. However, these achievements indispensably rely on the accurately-annotated datasets. If with the noisy-labeled images, the training procedure will immediately encounter difficulties, leading to a suboptimal classifier. This problem is even more crucial in the medical field, given that the annotation quality requires great expertise. In this paper, we propose an effective iterative learning framework for noisy-labeled medical image classification, to combat the lacking of high quality annotated medical data. Specifically, an online uncertainty sample mining method is proposed to eliminate the disturbance from noisy-labeled images. Next, we design a sample re-weighting strategy to preserve the usefulness of correctly-labeled hard samples. Our proposed method is validated on skin lesion classification task, and achieved very promising results.
We propose a new deep learning approach for medical imaging that copes with the problem of a small training set, the main bottleneck of deep learning, and apply it for classification of healthy and cancer cells acquired by quantitative phase imaging. The proposed method, called transferring of pre-trained generative adversarial network (TOP-GAN), is a hybridization between transfer learning and generative adversarial networks (GANs). Healthy cells and cancer cells of different metastatic potential have been imaged by low-coherence off-axis holography. After the acquisition, the optical path delay maps of the cells have been extracted and directly used as an input to the deep networks. In order to cope with the small number of classified images, we have used GANs to train a large number of unclassified images from another cell type (sperm cells). After this preliminary training, and after transforming the last layer of the network with new ones, we have designed an automatic classifier for the correct cell type (healthy/primary cancer/metastatic cancer) with 90-99% accuracy, although small training sets of down to several images have been used. These results are better in comparison to other classic methods that aim at coping with the same problem of a small training set. We believe that our approach makes the combination of holographic microscopy and deep learning networks more accessible to the medical field by enabling a rapid, automatic and accurate classification in stain-free imaging flow cytometry. Furthermore, our approach is expected to be applicable to many other medical image classification tasks, suffering from a small training set.
Supervised deep learning algorithms have enabled significant performance gains in medical image classification tasks. But these methods rely on large labeled datasets that require resource-intensive expert annotation. Semi-supervised generative adversarial network (GAN) approaches offer a means to learn from limited labeled data alongside larger unlabeled datasets, but have not been applied to discern fine-scale, sparse or localized features that define medical abnormalities. To overcome these limitations, we propose a patch-based semi-supervised learning approach and evaluate performance on classification of diabetic retinopathy from funduscopic images. Our semi-supervised approach achieves high AUC with just 10-20 labeled training images, and outperforms the supervised baselines by upto 15% when less than 30% of the training dataset is labeled. Further, our method implicitly enables interpretation of the SSL predictions. As this approach enables good accuracy, resolution and interpretability with lower annotation burden, it sets the pathway for scalable applications of deep learning in clinical imaging.
Convolutional Neural Networks (CNNs) usually use the same activation function, such as RELU, for all convolutional layers. There are performance limitations of just using RELU. In order to achieve better classification performance, reduce training and testing times, and reduce power consumption and memory usage, a new "Compressed Multi-function CNN" is developed. Google's Inception-V4, for example, is a very deep CNN that consists of 4 Inception-A blocks, 7 Inception-B blocks, and 3 Inception-C blocks. RELU is used for all convolutional layers. A new "Compressed Multi-function Inception-V4" (CMI) that can use different activation functions is created with k Inception-A blocks, m Inception-B blocks, and n Inception-C blocks where k in {1, 2, 3, 4}, m in {1, 2, 3, 4, 5, 6, 7}, n in {1, 2, 3}, and (k+m+n)<14. For performance analysis, a dataset for classifying brain MRI images into one of the four stages of Alzheimer's disease is used to compare three CMI architectures with Inception-V4 in terms of F1-score, training and testing times (related to power consumption), and memory usage (model size). Overall, simulations show that the new CMI models can outperform both the commonly used Inception-V4 and Inception-V4 using different activation functions. In the future, other "Compressed Multi-function CNNs", such as "Compressed Multi-function ResNets and DenseNets" that have a reduced number of convolutional blocks using different activation functions, will be developed to further increase classification accuracy, reduce training and testing times, reduce computational power, and reduce memory usage (model size) for building more effective healthcare systems, such as implementing accurate and convenient disease diagnosis systems on mobile devices that have limited battery power and memory.
We propose a novel attention gate (AG) model for medical image analysis that automatically learns to focus on target structures of varying shapes and sizes. Models trained with AGs implicitly learn to suppress irrelevant regions in an input image while highlighting salient features useful for a specific task. This enables us to eliminate the necessity of using explicit external tissue/organ localisation modules when using convolutional neural networks (CNNs). AGs can be easily integrated into standard CNN models such as VGG or U-Net architectures with minimal computational overhead while increasing the model sensitivity and prediction accuracy. The proposed AG models are evaluated on a variety of tasks, including medical image classification and segmentation. For classification, we demonstrate the use case of AGs in scan plane detection for fetal ultrasound screening. We show that the proposed attention mechanism can provide efficient object localisation while improving the overall prediction performance by reducing false positives. For segmentation, the proposed architecture is evaluated on two large 3D CT abdominal datasets with manual annotations for multiple organs. Experimental results show that AG models consistently improve the prediction performance of the base architectures across different datasets and training sizes while preserving computational efficiency. Moreover, AGs guide the model activations to be focused around salient regions, which provides better insights into how model predictions are made. The source code for the proposed AG models is publicly available.
Disease progression modeling (DPM) using longitudinal data is a challenging task in machine learning for healthcare that can provide clinicians with better tools for diagnosis and monitoring of disease. Existing DPM algorithms neglect temporal dependencies among measurements and make parametric assumptions about biomarker trajectories. In addition, they do not model multiple biomarkers jointly and need to align subjects' trajectories. In this paper, recurrent neural networks (RNNs) are utilized to address these issues. However, in many cases, longitudinal cohorts contain incomplete data, which hinders the application of standard RNNs and requires a pre-processing step such as imputation of the missing values. We, therefore, propose a generalized training rule for the most widely used RNN architecture, long short-term memory (LSTM) networks, that can handle missing values in both target and predictor variables. This algorithm is applied for modeling the progression of Alzheimer's disease (AD) using magnetic resonance imaging (MRI) biomarkers. The results show that the proposed LSTM algorithm achieves a lower mean absolute error for prediction of measurements across all considered MRI biomarkers compared to using standard LSTM networks with data imputation or using a regression-based DPM method. Moreover, applying linear discriminant analysis to the biomarkers' values predicted by the proposed algorithm results in a larger area under the receiver operating characteristic curve (AUC) for clinical diagnosis of AD compared to the same alternatives, and the AUC is comparable to state-of-the-art AUCs from a recent cross-sectional medical image classification challenge. This paper shows that built-in handling of missing values in LSTM network training paves the way for application of RNNs in disease progression modeling.
A key component to the success of deep learning is the availability of massive amounts of training data. Building and annotating large datasets for solving medical image classification problems is today a bottleneck for many applications. Recently, capsule networks were proposed to deal with shortcomings of Convolutional Neural Networks (ConvNets). In this work, we compare the behavior of capsule networks against ConvNets under typical datasets constraints of medical image analysis, namely, small amounts of annotated data and class-imbalance. We evaluate our experiments on MNIST, Fashion-MNIST and medical (histological and retina images) publicly available datasets. Our results suggest that capsule networks can be trained with less amount of data for the same or better performance and are more robust to an imbalanced class distribution, which makes our approach very promising for the medical imaging community.
The widely used ChestX-ray14 dataset addresses an important medical image classification problem and has the following caveats: 1) many lung pathologies are visually similar, 2) a variant of diseases including lung cancer, tuberculosis, and pneumonia are present in a single scan, i.e. multiple labels and 3) The incidence of healthy images is much larger than diseased samples, creating imbalanced data. These properties are common in medical domain. Existing literature uses stateof- the-art DensetNet/Resnet models being transfer learned where output neurons of the networks are trained for individual diseases to cater for multiple diseases labels in each image. However, most of them don't consider relationship between multiple classes. In this work we have proposed a novel error function, Multi-label Softmax Loss (MSML), to specifically address the properties of multiple labels and imbalanced data. Moreover, we have designed deep network architecture based on fine-grained classification concept that incorporates MSML. We have evaluated our proposed method on various network backbones and showed consistent performance improvements of AUC-ROC scores on the ChestX-ray14 dataset. The proposed error function provides a new method to gain improved performance across wider medical datasets.
Good results on image classification and retrieval using support vector machines (SVM) with local binary patterns (LBPs) as features have been extensively reported in the literature where an entire image is retrieved or classified. In contrast, in medical imaging, not all parts of the image may be equally significant or relevant to the image retrieval application at hand. For instance, in lung x-ray image, the lung region may contain a tumour, hence being highly significant whereas the surrounding area does not contain significant information from medical diagnosis perspective. In this paper, we propose to detect salient regions of images during training and fold the data to reduce the effect of irrelevant regions. As a result, smaller image areas will be used for LBP features calculation and consequently classification by SVM. We use IRMA 2009 dataset with 14,410 x-ray images to verify the performance of the proposed approach. The results demonstrate the benefits of saliency-based folding approach that delivers comparable classification accuracies with state-of-the-art but exhibits lower computational cost and storage requirements, factors highly important for big data analytics.
Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers---8x deeper than VGG nets but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.