Abstract
The development of accurate methods for OCT image analysis is highly dependent on the availability of large annotated datasets. As such datasets are usually expensive and hard to obtain, novel approaches based on deep generative models have been proposed for data augmentation. In this work, a flow-based network (SRFlow) and a generative adversarial network (ESRGAN) are used for synthesizing high-resolution OCT B-scans from low-resolution versions of real OCT images. The quality of the images generated by the two models is assessed using two standard fidelity-oriented metrics and a learned perceptual quality metric. The performance of two classification models trained on real and synthetic images is also evaluated. The obtained results show that the images generated by SRFlow preserve higher fidelity to the ground truth, while the outputs of ESRGAN present, on average, better perceptual quality. Independently of the architecture of the network chosen to classify the OCT B-scans, the model's performance always improves when images generated by SRFlow are included in the training set.

Abstract
Deep Learning (DL) algorithms allow fast results with high accuracy in medical imaging analysis solutions. However, to achieve a desirable performance, they require large amounts of high quality data. Active Learning (AL) is a subfield of DL that aims for more efficient models requiring ideally fewer data, by selecting the most relevant information for training. CheXpert is a Chest X-Ray (CXR) dataset, containing labels for different pathologic findings, alongside a Support Devices (SD) label. The latter contains several misannotations, which may impact the performance of a pathology detection model. The aim of this work is the detection of SDs in CheXpert CXR images and the comparison of the resulting predictions with the original CheXpert SD annotations, using AL approaches. A subset of 10,220 images was selected, manually annotated for SDs and used in the experimentations. In the first experiment, an initial model was trained on the seed dataset (6,200 images from this subset). The second and third approaches consisted in AL random sampling and least confidence techniques. In both of these, the seed dataset was used initially, and more images were iteratively employed. Finally, in the fourth experiment, a model was trained on the full annotated set. The AL least confidence experiment outperformed the remaining approaches, presenting an AUC of 71.10% and showing that training a model with representative information is favorable over training with all labeled data. This model was used to obtain predictions, which can be useful to limit the use of SD mislabelled images in future models.

Abstract

Abstract
Chest radiography is one of the most common imaging exams, but its interpretation is often challenging and timeconsuming, which has motivated the development of automated tools for pathology/abnormality detection. Deep learning models trained on large-scale chest X-ray datasets have shown promising results but are highly dependent on the quality of the data. However, these datasets often contain incorrect metadata and non-compliant or corrupted images. These inconsistencies are ultimately incorporated in the training process, impairing the validity of the results. In this study, a novel approach to detect non-compliant images based on deep features extracted from a patient position classification model and a pre-trained VGG16 model are proposed. This method is applied to CheXpert, a widely used public dataset. From a pool of 100 images, it is shown that the deep feature-based methods based on a patient position classification model are able to retrieve a larger number of non-compliant images (up to 81% of non-compliant images), when compared to the same methods but based on a pretrained VGG16 (up to 73%) and the state of the art uncertainty-based method (50%).

Abstract
The importance of X-Ray imaging analysis is paramount for healthcare institutions since it is the main imaging modality for patient diagnosis, and deep learning can be used to aid clinicians in image diagnosis or structure segmentation. In recent years, several articles demonstrate the capability that deep learning models have in classifying and segmenting chest x-ray images if trained in an annotated dataset. Unfortunately, for segmentation tasks, only a few relatively small datasets have annotations, which poses a problem for the training of robust deep learning strategies. In this work, a semi-supervised approach is developed which consists of using available information regarding other anatomical structures to guide the segmentation when the groundtruth segmentation for a given structure is not available. This semi-supervised is compared with a fully-supervised approach for the tasks of lung segmentation and for multi-structure segmentation (lungs, heart and clavicles) in chest x-ray images. The semi-supervised lung predictions are evaluated visually and show relevant improvements, therefore this approach could be used to improve performance in external datasets with missing groundtruth. The multi-structure predictions show an improvement in mean absolute and Hausdorff distances when compared to a fully supervised approach and visual analysis of the segmentations shows that false positive predictions are removed. In conclusion, the developed method results in a new strategy that can help solve the problem of missing annotations and increase the quality of predictions in new datasets.

Abstract
Explanations of a model's biases or predictions are essential to medical image analysis. Yet, explainable machine learning approaches for medical image analysis are challenged by needs to preserve privacy of patient data, and by current trends in deep learning to use unsustainably large models and large datasets. We propose DeepFixCX for explainable and privacy-preserving medical image compression that is nimble and performant. We contribute a review of the field and a conceptual framework for simultaneous privacy and explainability via tools of compression. DeepFixCX compresses images without learning by removing or obscuring spatial and edge information. DeepFixCX is ante-hoc explainable and gives privatized post hoc explanations of spatial and edge bias without accessing the original image. DeepFixCX privatizes images to prevent image reconstruction and mitigate patient re-identification. DeepFixCX is nimble. Compression can occur on a laptop CPU or GPU to compress and privatize 1700 images per second of size 320 x 320. DeepFixCX enables use of low memory MLP classifiers for vision data; permitting small performance loss gives end-to-end MLP performance over 70x faster and batch size over 100x larger. DeepFixCX consistently improves predictive classification performance of a Deep Neural Network (DNN) by 0.02 AUC ROC on Glaucoma and Cervix Type detection datasets, and can improve multi-label chest x-ray classification performance in seven of 10 tested settings. In all three datasets, compression to less than 5% of original number of pixels gives matching or improved performance. Our main novelty is to define an explainability versus privacy problem and address it with lossy compression.This article is categorized under:Fundamental Concepts of Data and Knowledge > Explainable AICommercial, Legal, and Ethical Issues > Security and PrivacyFundamental Concepts of Data and Knowledge > Big Data Mining