COMFormer: Classification of Maternal-Fetal and Brain Anatomy Using a Residual Cross-Covariance Attention Guided Transformer in Ultrasound.

Monitoring the healthy development of a fetus requires accurate and timely identification of different maternal-fetal structures as they grow. To facilitate this objective in an automated fashion, we propose a deep-learning-based image classification architecture called the COMFormer to classify maternal-fetal and brain anatomical structures present in 2-D fetal ultrasound (US) images. The proposed architecture classifies the two subcategories separately: maternal-fetal (abdomen, brain, femur, thorax, mother's cervix (MC), and others) and brain anatomical structures [trans-thalamic (TT), trans-cerebellum (TC), trans-ventricular (TV), and non-brain (NB)]. Our proposed architecture relies on a transformer-based approach that leverages spatial and global features using a newly designed residual cross-variance attention block. This block introduces an advanced cross-covariance attention (XCA) mechanism to capture a long-range representation from the input using spatial (e.g., shape, texture, intensity) and global features. To build COMFormer, we used a large publicly available dataset (BCNatal) consisting of 12 400 images from 1792 subjects. Experimental results prove that COMFormer outperforms the recent CNN and transformer-based models by achieving 95.64% and 96.33% classification accuracy on maternal-fetal and brain anatomy, respectively.

Neural Style Transfer as Data Augmentation for Improving COVID-19 Diagnosis Classification.

Coronavirus disease 2019 (COVID-19) has accounted for millions of causalities. While it affects not only individuals but also our collective healthcare and economic systems, testing is insufficient and costly hampering efforts to deal with the pandemic. Chest X-rays are routine radiographic imaging tests that are used for the diagnosis of respiratory conditions such as pneumonia and COVID-19. Convolutional neural networks have shown promise to be effective at classifying X-rays for assisting diagnosis of conditions; however, achieving robust performance demanded in most modern medical applications typically requires a large number of samples. While there exist datasets containing thousands of X-ray images of patients with healthy and pneumonia diagnoses, because COVID-19 is such a recent phenomenon, there are relatively few confirmed COVID-19 positive chest X-rays openly available to the research community. In this paper, we demonstrate the effectiveness of cycle-generative adversarial network, commonly used for neural style transfer, as a way to augment COVID-19 negative X-ray images to look like COVID-19 positive images for increasing the number of COVID-19 positive training samples. The statistical results show an increase in the mean macro f1-score over 21% on a one-tailed t score = 2.68 and p value = 0.01 to accept our alternative hypothesis for an α = 0.05 . We conclude that this approach, when used in conjunction with standard transfer learning techniques, is effective at improving the performance of COVID-19 classifiers for a variety of common convolutional neural networks.

Semi-Automated Data Labeling for Activity Recognition in Pervasive Healthcare.

Activity recognition, a key component in pervasive healthcare monitoring, relies on classification algorithms that require labeled data of individuals performing the activity of interest to train accurate models. Labeling data can be performed in a lab setting where an individual enacts the activity under controlled conditions. The ubiquity of mobile and wearable sensors allows the collection of large datasets from individuals performing activities in naturalistic conditions. Gathering accurate data labels for activity recognition is typically an expensive and time-consuming process. In this paper we present two novel approaches for semi-automated online data labeling performed by the individual executing the activity of interest. The approaches have been designed to address two of the limitations of self-annotation: (i) The burden on the user performing and annotating the activity, and (ii) the lack of accuracy due to the user labeling the data minutes or hours after the completion of an activity. The first approach is based on the recognition of subtle finger gestures performed in response to a data-labeling query. The second approach focuses on labeling activities that have an auditory manifestation and uses a classifier to have an initial estimation of the activity, and a conversational agent to ask the participant for clarification or for additional data. Both approaches are described, evaluated in controlled experiments to assess their feasibility and their advantages and limitations are discussed. Results show that while both studies have limitations, they achieve 80% to 90% precision.