21 January 2020 in IEEE Transactions on Medical Imaging
Breast cancer is one of the most frequently diagnosed solid cancers. Mammography is the most commonly used screening technology for detecting breast cancer. Traditional machine learning methods of mammographic image classification or segmentation using manual features require a great quantity of manual segmentation annotation data to train the model and test the results. But manual labeling is expensive, time-consuming, and laborious, and greatly increases the cost of system construction. To reduce this cost and the workload of radiologists, an end-to-end full-image mammogram classification method based on deep neural networks was proposed for classifier building, which can be constructed without bounding boxes or mask ground truth label of training data. The only label required in this method is the classification of mammographic images, which can be relatively easy to collect from diagnostic reports. Because breast lesions usually take up a fraction of the total area visualized in the mammographic image, we propose different pooling structures for convolutional neural networks(CNNs) instead of the common pooling methods, which divide the image into regions and select the few with high probability of malignancy as the representation of the whole mammographic image. The proposed pooling structures can be applied on most CNN-based models, which may greatly improve the models' performance on mammographic image data with the same input. Experimental results on the publicly available INbreast dataset and CBIS dataset indicate that the proposed pooling structures perform satisfactorily on mammographic image data compared with previous state-of-the-art mammographic image classifiers and detection algorithm using segmentation annotations.
7 January 2023 in Medical Image Analysis
1 July 2018 in Applied Soft Computing
14 October 2022 in IEEE Transactions on Cybernetics
Deep neural network has shown a powerful performance in the medical image analysis of a variety of diseases. However, a number of studies over the past few years have demonstrated that these deep learning systems can be vulnerable to well-designed adversarial attacks, with minor disruptions added to the input. Since both the public and academia have focused on deep learning in the health information economy, these adversarial attacks would prove more important and raise security concerns. In this article, adversarial attacks on deep learning systems in medicine are analyzed from two different points of view: 1) white box and 2) black box. A fast adversarial sample generation method, Feature Space-Restricted Attention Attack is proposed to explore more confusing adversarial samples. It is based on a generative adversarial network with bound classification space to generate perturbations to achieve attacks. Meanwhile, it can employ an attention mechanism to focus this perturbation on the lesion region. This enables the perturbation closely associated with the classification information making the attack more efficient and invisible. The performance and specificity of the proposed attack method are demonstrated by conducting extensive experiments on three different types of medical images. Finally, it is expected that this work can assist practitioners become being of current weaknesses in the deployment of deep learning systems in clinical settings. And, it further investigates domain-specific features of medical deep learning systems to enhance model generalization and resistance to attacks.
16 December 2021 in IEEE Transactions on Artificial Intelligence
Mammography is one of the most commonly used methods for breast cancer screening. Most mammogram classification or segmentation models are trained with additional manual segmentation annotation data, which is difficult to obtain. Meanwhile, these models rarely take into account the fact that lesion features may vary at different scales. Moreover, most models have difficulty in quantifying the effect of single lesion regions on prediction results. To solve the above problems, we propose deep multiscale multi-instance networks with regional scoring. This model classifies mammograms automatically by image labels without using any other annotation information during the training procedure. The proposed multiscale structures include large scale, medium scale, and small scale, and they can extract features of mammograms from global to local. Because there may be more than one breast lesion region in a mammogram, the proposed multi-instance structure divides the image into multiple regions and selects few regions of interest to represent the whole image together. The regional scoring mechanism quantifies the contribution of each region to the prediction results, and thus, makes the results interpretable. In addition, the proposed method can roughly localize the lesion area by the score of each region. The experimental results on a publicly available dataset and a private dataset have shown that the proposed method achieved the highest AUC on both datasets compared with the state-of-the-art models.