Optimising Chest X-Rays for Image Analysis by Identifying and Removing Confounding Factors (preprint)

During the COVID-19 pandemic, the sheer volume of imaging performed in an emergency setting for COVID-19 diagnosis has resulted in a wide variability of clinical CXR acquisitions. This variation is seen in the CXR projections used, image annotations added and in the inspiratory effort and degree of rotation of clinical images. The image analysis community has attempted to ease the burden on overstretched radiology departments during the pandemic by developing automated COVID-19 diagnostic algorithms, the input for which has been CXR imaging. Large publicly available CXR datasets have been leveraged to improve deep learning algorithms for COVID-19 diagnosis. Yet the variable quality of clinically-acquired CXRs within publicly available datasets could have a profound effect on algorithm performance. COVID-19 diagnosis may be inferred by an algorithm from non-anatomical features on an image such as image labels. These imaging shortcuts may be dataset-specific and limit the generalisability of AI systems. Understanding and correcting key potential biases in CXR images is therefore an essential first step prior to CXR image analysis. In this study, we propose a simple and effective step-wise approach to pre-processing a COVID-19 chest X-ray dataset to remove undesired biases. We perform ablation studies to show the impact of each individual step. The results suggest that using our proposed pipeline could increase accuracy of the baseline COVID-19 detection algorithm by up to 13%.

Rapid Lung Ultrasound COVID-19 Severity Scoring with Resource-Efficient Deep Feature Extraction (preprint)

Artificial intelligence-based analysis of lung ultrasound imaging has been demonstrated as an effective technique for rapid diagnostic decision support throughout the COVID-19 pandemic. However, such techniques can require days- or weeks-long training processes and hyper-parameter tuning to develop intelligent deep learning image analysis models. This work focuses on leveraging 'off-the-shelf' pre-trained models as deep feature extractors for scoring disease severity with minimal training time. We propose using pre-trained initializations of existing methods ahead of simple and compact neural networks to reduce reliance on computational capacity. This reduction of computational capacity is of critical importance in time-limited or resource-constrained circumstances, such as the early stages of a pandemic. On a dataset of 49 patients, comprising over 20,000 images, we demonstrate that the use of existing methods as feature extractors results in the effective classification of COVID-19-related pneumonia severity while requiring only minutes of training time. Our methods can achieve an accuracy of over 0.93 on a 4-level severity score scale and provides comparable per-patient region and global scores compared to expert annotated ground truths. These results demonstrate the capability for rapid deployment and use of such minimally-adapted methods for progress monitoring, patient stratification and management in clinical practice for COVID-19 patients, and potentially in other respiratory diseases.

Lung Ultrasound Segmentation and Adaptation between COVID-19 and Community-Acquired Pneumonia (preprint)

Lung ultrasound imaging has been shown effective in detecting typical patterns for interstitial pneumonia, as a point-of-care tool for both patients with COVID-19 and other community-acquired pneumonia (CAP). In this work, we focus on the hyperechoic B-line segmentation task. Using deep neural networks, we automatically outline the regions that are indicative of pathology-sensitive artifacts and their associated sonographic patterns. With a real-world data-scarce scenario, we investigate approaches to utilize both COVID-19 and CAP lung ultrasound data to train the networks; comparing fine-tuning and unsupervised domain adaptation. Segmenting either type of lung condition at inference may support a range of clinical applications during evolving epidemic stages, but also demonstrates value in resource-constrained clinical scenarios. Adapting real clinical data acquired from COVID-19 patients to those from CAP patients significantly improved Dice scores from 0.60 to 0.87 (p < 0.001) and from 0.43 to 0.71 (p < 0.001), on independent COVID-19 and CAP test cases, respectively. It is of practical value that the improvement was demonstrated with only a small amount of data in both training and adaptation data sets, a common constraint for deploying machine learning models in clinical practice. Interestingly, we also report that the inverse adaptation, from labelled CAP data to unlabeled COVID-19 data, did not demonstrate an improvement when tested on either condition. Furthermore, we offer a possible explanation that correlates the segmentation performance to label consistency and data domain diversity in this point-of-care lung ultrasound application.

Image quality assessment for closed-loop computer-assisted lung ultrasound (preprint)

We describe a novel, two-stage computer assistance system for lung anomaly detection using ultrasound imaging in the intensive care setting to improve operator performance and patient stratification during coronavirus pandemics. The proposed system consists of two deep-learning-based models: a quality assessment module that automates predictions of image quality, and a diagnosis assistance module that determines the likelihood-oh-anomaly in ultrasound images of sufficient quality. Our two-stage strategy uses a novelty detection algorithm to address the lack of control cases available for training the quality assessment classifier. The diagnosis assistance module can then be trained with data that are deemed of sufficient quality, guaranteed by the closed-loop feedback mechanism from the quality assessment module. Using more than 25000 ultrasound images from 37 COVID-19-positive patients scanned at two hospitals, plus 12 control cases, this study demonstrates the feasibility of using the proposed machine learning approach. We report an accuracy of 86% when classifying between sufficient and insufficient quality images by the quality assessment module. For data of sufficient quality - as determined by the quality assessment module - the mean classification accuracy, sensitivity, and specificity in detecting COVID-19-positive cases were 0.95, 0.91, and 0.97, respectively, across five holdout test data sets unseen during the training of any networks within the proposed system. Overall, the integration of the two modules yields accurate, fast, and practical acquisition guidance and diagnostic assistance for patients with suspected respiratory conditions at point-of-care.

The challenges of deploying artificial intelligence models in a rapidly evolving pandemic (preprint)

The COVID-19 pandemic, caused by the severe acute respiratory syndrome coronavirus 2, emerged into a world being rapidly transformed by artificial intelligence (AI) based on big data, computational power and neural networks. The gaze of these networks has in recent years turned increasingly towards applications in healthcare. It was perhaps inevitable that COVID-19, a global disease propagating health and economic devastation, should capture the attention and resources of the world's computer scientists in academia and industry. The potential for AI to support the response to the pandemic has been proposed across a wide range of clinical and societal challenges, including disease forecasting, surveillance and antiviral drug discovery. This is likely to continue as the impact of the pandemic unfolds on the world's people, industries and economy but a surprising observation on the current pandemic has been the limited impact AI has had to date in the management of COVID-19. This correspondence focuses on exploring potential reasons behind the lack of successful adoption of AI models developed for COVID-19 diagnosis and prognosis, in front-line healthcare services. We highlight the moving clinical needs that models have had to address at different stages of the epidemic, and explain the importance of translating models to reflect local healthcare environments. We argue that both basic and applied research are essential to accelerate the potential of AI models, and this is particularly so during a rapidly evolving pandemic. This perspective on the response to COVID-19, may provide a glimpse into how the global scientific community should react to combat future disease outbreaks more effectively.