Acquisition-invariant brain MRI segmentation with informative uncertainties.

Combining multi-site data can strengthen and uncover trends, but is a task that is marred by the influence of site-specific covariates that can bias the data and, therefore, any downstream analyses. Post-hoc multi-site correction methods exist but have strong assumptions that often do not hold in real-world scenarios. Algorithms should be designed in a way that can account for site-specific effects, such as those that arise from sequence parameter choices, and in instances where generalisation fails, should be able to identify such a failure by means of explicit uncertainty modelling. This body of work showcases such an algorithm that can become robust to the physics of acquisition in the context of segmentation tasks while simultaneously modelling uncertainty. We demonstrate that our method not only generalises to complete holdout datasets, preserving segmentation quality but does so while also accounting for site-specific sequence choices, which also allows it to perform as a harmonisation tool.

Knowledge barriers in a national symptomatic-COVID-19 testing programme.

Symptomatic testing programmes are crucial to the COVID-19 pandemic response. We sought to examine United Kingdom (UK) testing rates amongst individuals with test-qualifying symptoms, and factors associated with not testing. We analysed a cohort of untested symptomatic app users (N = 1,237), nested in the Zoe COVID Symptom Study (Zoe, N = 4,394,948); and symptomatic respondents who wanted, but did not have a test (N = 1,956), drawn from a University of Maryland survey administered to Facebook users (The Global COVID-19 Trends and Impact Survey [CTIS], N = 775,746). The proportion tested among individuals with incident test-qualifying symptoms rose from ~20% to ~75% from April to December 2020 in Zoe. Testing was lower with one vs more symptoms (72.9% vs 84.6% p<0.001), or short vs long symptom duration (69.9% vs 85.4% p<0.001). 40.4% of survey respondents did not identify all three test-qualifying symptoms. Symptom identification decreased for every decade older (OR = 0.908 [95% CI 0.883-0.933]). Amongst symptomatic UMD-CTIS respondents who wanted but did not have a test, not knowing where to go was the most cited factor (32.4%); this increased for each decade older (OR = 1.207 [1.129-1.292]) and for every 4-years fewer in education (OR = 0.685 [0.599-0.783]). Despite current UK messaging on COVID-19 testing, there is a knowledge gap about when and where to test, and this may be contributing to the ~25% testing gap. Risk factors, including older age and less education, highlight potential opportunities to tailor public health messages. The testing gap may be ever larger in countries that do not have extensive, free testing, as the UK does.

Imitation learning for improved 3D PET/MR attenuation correction.

The assessment of the quality of synthesised/pseudo Computed Tomography (pCT) images is commonly measured by an intensity-wise similarity between the ground truth CT and the pCT. However, when using the pCT as an attenuation map (µ-map) for PET reconstruction in Positron Emission Tomography Magnetic Resonance Imaging (PET/MRI) minimising the error between pCT and CT neglects the main objective of predicting a pCT that when used as µ-map reconstructs a pseudo PET (pPET) which is as similar as possible to the gold standard CT-derived PET reconstruction. This observation motivated us to propose a novel multi-hypothesis deep learning framework explicitly aimed at PET reconstruction application. A convolutional neural network (CNN) synthesises pCTs by minimising a combination of the pixel-wise error between pCT and CT and a novel metric-loss that itself is defined by a CNN and aims to minimise consequent PET residuals. Training is performed on a database of twenty 3D MR/CT/PET brain image pairs. Quantitative results on a fully independent dataset of twenty-three 3D MR/CT/PET image pairs show that the network is able to synthesise more accurate pCTs. The Mean Absolute Error on the pCT (110.98 HU ± 19.22 HU) compared to a baseline CNN (172.12 HU ± 19.61 HU) and a multi-atlas propagation approach (153.40 HU ± 18.68 HU), and subsequently lead to a significant improvement in the PET reconstruction error (4.74% ± 1.52% compared to baseline 13.72% ± 2.48% and multi-atlas propagation 6.68% ± 2.06%).

Changes in symptomatology, reinfection, and transmissibility associated with the SARS-CoV-2 variant B.1.1.7: an ecological study.

BACKGROUND: The SARS-CoV-2 variant B.1.1.7 was first identified in December, 2020, in England. We aimed to investigate whether increases in the proportion of infections with this variant are associated with differences in symptoms or disease course, reinfection rates, or transmissibility. METHODS: We did an ecological study to examine the association between the regional proportion of infections with the SARS-CoV-2 B.1.1.7 variant and reported symptoms, disease course, rates of reinfection, and transmissibility. Data on types and duration of symptoms were obtained from longitudinal reports from users of the COVID Symptom Study app who reported a positive test for COVID-19 between Sept 28 and Dec 27, 2020 (during which the prevalence of B.1.1.7 increased most notably in parts of the UK). From this dataset, we also estimated the frequency of possible reinfection, defined as the presence of two reported positive tests separated by more than 90 days with a period of reporting no symptoms for more than 7 days before the second positive test. The proportion of SARS-CoV-2 infections with the B.1.1.7 variant across the UK was estimated with use of genomic data from the COVID-19 Genomics UK Consortium and data from Public Health England on spike-gene target failure (a non-specific indicator of the B.1.1.7 variant) in community cases in England. We used linear regression to examine the association between reported symptoms and proportion of B.1.1.7. We assessed the Spearman correlation between the proportion of B.1.1.7 cases and number of reinfections over time, and between the number of positive tests and reinfections. We estimated incidence for B.1.1.7 and previous variants, and compared the effective reproduction number, Rt, for the two incidence estimates. FINDINGS: From Sept 28 to Dec 27, 2020, positive COVID-19 tests were reported by 36 920 COVID Symptom Study app users whose region was known and who reported as healthy on app sign-up. We found no changes in reported symptoms or disease duration associated with B.1.1.7. For the same period, possible reinfections were identified in 249 (0·7% [95% CI 0·6-0·8]) of 36 509 app users who reported a positive swab test before Oct 1, 2020, but there was no evidence that the frequency of reinfections was higher for the B.1.1.7 variant than for pre-existing variants. Reinfection occurrences were more positively correlated with the overall regional rise in cases (Spearman correlation 0·56-0·69 for South East, London, and East of England) than with the regional increase in the proportion of infections with the B.1.1.7 variant (Spearman correlation 0·38-0·56 in the same regions), suggesting B.1.1.7 does not substantially alter the risk of reinfection. We found a multiplicative increase in the Rt of B.1.1.7 by a factor of 1·35 (95% CI 1·02-1·69) relative to pre-existing variants. However, Rt fell below 1 during regional and national lockdowns, even in regions with high proportions of infections with the B.1.1.7 variant. INTERPRETATION: The lack of change in symptoms identified in this study indicates that existing testing and surveillance infrastructure do not need to change specifically for the B.1.1.7 variant. In addition, given that there was no apparent increase in the reinfection rate, vaccines are likely to remain effective against the B.1.1.7 variant. FUNDING: Zoe Global, Department of Health (UK), Wellcome Trust, Engineering and Physical Sciences Research Council (UK), National Institute for Health Research (UK), Medical Research Council (UK), Alzheimer's Society.

Attributes and predictors of long COVID.

Reports of long-lasting coronavirus disease 2019 (COVID-19) symptoms, the so-called 'long COVID', are rising but little is known about prevalence, risk factors or whether it is possible to predict a protracted course early in the disease. We analyzed data from 4,182 incident cases of COVID-19 in which individuals self-reported their symptoms prospectively in the COVID Symptom Study app1. A total of 558 (13.3%) participants reported symptoms lasting ≥28 days, 189 (4.5%) for ≥8 weeks and 95 (2.3%) for ≥12 weeks. Long COVID was characterized by symptoms of fatigue, headache, dyspnea and anosmia and was more likely with increasing age and body mass index and female sex. Experiencing more than five symptoms during the first week of illness was associated with long COVID (odds ratio = 3.53 (2.76-4.50)). A simple model to distinguish between short COVID and long COVID at 7 days (total sample size, n = 2,149) showed an area under the curve of the receiver operating characteristic curve of 76%, with replication in an independent sample of 2,472 individuals who were positive for severe acute respiratory syndrome coronavirus 2. This model could be used to identify individuals at risk of long COVID for trials of prevention or treatment and to plan education and rehabilitation services.

Symptom clusters in COVID-19: A potential clinical prediction tool from the COVID Symptom Study app.

As no one symptom can predict disease severity or the need for dedicated medical support in coronavirus disease 2019 (COVID-19), we asked whether documenting symptom time series over the first few days informs outcome. Unsupervised time series clustering over symptom presentation was performed on data collected from a training dataset of completed cases enlisted early from the COVID Symptom Study Smartphone application, yielding six distinct symptom presentations. Clustering was validated on an independent replication dataset between 1 and 28 May 2020. Using the first 5 days of symptom logging, the ROC-AUC (receiver operating characteristic - area under the curve) of need for respiratory support was 78.8%, substantially outperforming personal characteristics alone (ROC-AUC 69.5%). Such an approach could be used to monitor at-risk patients and predict medical resource requirements days before they are required.

Geo-social gradients in predicted COVID-19 prevalence in Great Britain: results from 1 960 242 users of the COVID-19 Symptoms Study app.

Understanding the geographical distribution of COVID-19 through the general population is key to the provision of adequate healthcare services. Using self-reported data from 1 960 242 unique users in Great Britain (GB) of the COVID-19 Symptom Study app, we estimated that, concurrent to the GB government sanctioning lockdown, COVID-19 was distributed across GB, with evidence of 'urban hotspots'. We found a geo-social gradient associated with predicted disease prevalence suggesting urban areas and areas of higher deprivation are most affected. Our results demonstrate use of self-reported symptoms data to provide focus on geographical areas with identified risk factors.

Cancer and Risk of COVID-19 Through a General Community Survey.

Individuals with cancer may be at high risk for coronavirus disease 2019 (COVID-19) and adverse outcomes. However, evidence from large population-based studies examining whether cancer and cancer-related therapy exacerbates the risk of COVID-19 infection is still limited. Data were collected from the COVID Symptom Study smartphone application since March 29 through May 8, 2020. Among 23,266 participants with cancer and 1,784,293 without cancer, we documented 10,404 reports of a positive COVID-19 test. Compared with participants without cancer, those living with cancer had a 60% increased risk of a positive COVID-19 test. Among patients with cancer, current treatment with chemotherapy or immunotherapy was associated with a 2.2-fold increased risk of a positive test. The association between cancer and COVID-19 infection was stronger among participants >65 years and males. Future studies are needed to identify subgroups by tumor types and treatment regimens who are particularly at risk for COVID-19 infection and adverse outcomes.

Detecting COVID-19 infection hotspots in England using large-scale self-reported data from a mobile application: a prospective, observational study.

BACKGROUND: As many countries seek to slow the spread of COVID-19 without reimposing national restrictions, it has become important to track the disease at a local level to identify areas in need of targeted intervention. METHODS: In this prospective, observational study, we did modelling using longitudinal, self-reported data from users of the COVID Symptom Study app in England between March 24, and Sept 29, 2020. Beginning on April 28, in England, the Department of Health and Social Care allocated RT-PCR tests for COVID-19 to app users who logged themselves as healthy at least once in 9 days and then reported any symptom. We calculated incidence of COVID-19 using the invited swab (RT-PCR) tests reported in the app, and we estimated prevalence using a symptom-based method (using logistic regression) and a method based on both symptoms and swab test results. We used incidence rates to estimate the effective reproduction number, R(t), modelling the system as a Poisson process and using Markov Chain Monte-Carlo. We used three datasets to validate our models: the Office for National Statistics (ONS) Community Infection Survey, the Real-time Assessment of Community Transmission (REACT-1) study, and UK Government testing data. We used geographically granular estimates to highlight regions with rapidly increasing case numbers, or hotspots. FINDINGS: From March 24 to Sept 29, 2020, a total of 2 873 726 users living in England signed up to use the app, of whom 2 842 732 (98·9%) provided valid age information and daily assessments. These users provided a total of 120 192 306 daily reports of their symptoms, and recorded the results of 169 682 invited swab tests. On a national level, our estimates of incidence and prevalence showed a similar sensitivity to changes to those reported in the ONS and REACT-1 studies. On Sept 28, 2020, we estimated an incidence of 15 841 (95% CI 14 023-17 885) daily cases, a prevalence of 0·53% (0·45-0·60), and R(t) of 1·17 (1·15-1·19) in England. On a geographically granular level, on Sept 28, 2020, we detected 15 (75%) of the 20 regions with highest incidence according to government test data. INTERPRETATION: Our method could help to detect rapid case increases in regions where government testing provision is lower. Self-reported data from mobile applications can provide an agile resource to inform policy makers during a quickly moving pandemic, serving as a complementary resource to more traditional instruments for disease surveillance. FUNDING: Zoe Global, UK Government Department of Health and Social Care, Wellcome Trust, UK Engineering and Physical Sciences Research Council, UK National Institute for Health Research, UK Medical Research Council and British Heart Foundation, Alzheimer's Society, Chronic Disease Research Foundation.

Detecting COVID-19 infection hotspots in England using large-scale self-reported data from a mobile application: a prospective, observational study.

BACKGROUND: As many countries seek to slow the spread of COVID-19 without reimposing national restrictions, it has become important to track the disease at a local level to identify areas in need of targeted intervention. METHODS: We performed modelling on longitudinal, self-reported data from users of the COVID Symptom Study app in England between 24 March and 29 September, 2020. Combining a symptom-based predictive model for COVID-19 positivity and RT-PCR tests provided by the Department of Health we were able to estimate disease incidence, prevalence and effective reproduction number. Geographically granular estimates were used to highlight regions with rapidly increasing case numbers, or hotspots. FINDINGS: More than 2.8 million app users in England provided 120 million daily reports of their symptoms, and recorded the results of 170,000 PCR tests. On a national level our estimates of incidence and prevalence showed similar sensitivity to changes as two national community surveys: the ONS and REACT-1 studies. On 28 September 2020 we estimated 15,841 (95% CI 14,023-17,885) daily cases, a prevalence of 0.53% (95% CI 0.45-0.60), and R(t) of 1.17 (95% credible interval 1.15-1.19) in England. On a geographically granular level, on 28 September 2020 we detected 15 of the 20 regions with highest incidence according to Government test data, with indications that our method may be able to detect rapid case increases in regions where Government testing provision is more limited. INTERPRETATION: Self-reported data from mobile applications can provide an agile resource to inform policymakers during a fast-moving pandemic, serving as an independent and complementary resource to more traditional instruments for disease surveillance. FUNDING: Zoe Global Limited, Department of Health, Wellcome Trust, EPSRC, NIHR, MRC, Alzheimer's Society.

Risk of COVID-19 among front-line health-care workers and the general community: a prospective cohort study.

BACKGROUND: Data for front-line health-care workers and risk of COVID-19 are limited. We sought to assess risk of COVID-19 among front-line health-care workers compared with the general community and the effect of personal protective equipment (PPE) on risk. METHODS: We did a prospective, observational cohort study in the UK and the USA of the general community, including front-line health-care workers, using self-reported data from the COVID Symptom Study smartphone application (app) from March 24 (UK) and March 29 (USA) to April 23, 2020. Participants were voluntary users of the app and at first use provided information on demographic factors (including age, sex, race or ethnic background, height and weight, and occupation) and medical history, and subsequently reported any COVID-19 symptoms. We used Cox proportional hazards modelling to estimate multivariate-adjusted hazard ratios (HRs) of our primary outcome, which was a positive COVID-19 test. The COVID Symptom Study app is registered with ClinicalTrials.gov, NCT04331509. FINDINGS: Among 2 035 395 community individuals and 99 795 front-line health-care workers, we recorded 5545 incident reports of a positive COVID-19 test over 34 435 272 person-days. Compared with the general community, front-line health-care workers were at increased risk for reporting a positive COVID-19 test (adjusted HR 11·61, 95% CI 10·93-12·33). To account for differences in testing frequency between front-line health-care workers and the general community and possible selection bias, an inverse probability-weighted model was used to adjust for the likelihood of receiving a COVID-19 test (adjusted HR 3·40, 95% CI 3·37-3·43). Secondary and post-hoc analyses suggested adequacy of PPE, clinical setting, and ethnic background were also important factors. INTERPRETATION: In the UK and the USA, risk of reporting a positive test for COVID-19 was increased among front-line health-care workers. Health-care systems should ensure adequate availability of PPE and develop additional strategies to protect health-care workers from COVID-19, particularly those from Black, Asian, and minority ethnic backgrounds. Additional follow-up of these observational findings is needed. FUNDING: Zoe Global, Wellcome Trust, Engineering and Physical Sciences Research Council, National Institutes of Health Research, UK Research and Innovation, Alzheimer's Society, National Institutes of Health, National Institute for Occupational Safety and Health, and Massachusetts Consortium on Pathogen Readiness.

Risk of COVID-19 among frontline healthcare workers and the general community: a prospective cohort study.

Background: Data for frontline healthcare workers (HCWs) and risk of SARS-CoV-2 infection are limited and whether personal protective equipment (PPE) mitigates this risk is unknown. We evaluated risk for COVID-19 among frontline HCWs compared to the general community and the influence of PPE. Methods: We performed a prospective cohort study of the general community, including frontline HCWs, who reported information through the COVID Symptom Study smartphone application beginning on March 24 (United Kingdom, U.K.) and March 29 (United States, U.S.) through April 23, 2020. We used Cox proportional hazards modeling to estimate multivariate-adjusted hazard ratios (aHRs) of a positive COVID-19 test. Findings: Among 2,035,395 community individuals and 99,795 frontline HCWs, we documented 5,545 incident reports of a positive COVID-19 test over 34,435,272 person-days. Compared with the general community, frontline HCWs had an aHR of 11·6 (95% CI: 10·9 to 12·3) for reporting a positive test. The corresponding aHR was 3·40 (95% CI: 3·37 to 3·43) using an inverse probability weighted Cox model adjusting for the likelihood of receiving a test. A symptom-based classifier of predicted COVID-19 yielded similar risk estimates. Compared with HCWs reporting adequate PPE, the aHRs for reporting a positive test were 1·46 (95% CI: 1·21 to 1·76) for those reporting PPE reuse and 1·31 (95% CI: 1·10 to 1·56) for reporting inadequate PPE. Compared with HCWs reporting adequate PPE who did not care for COVID-19 patients, HCWs caring for patients with documented COVID-19 had aHRs for a positive test of 4·83 (95% CI: 3·99 to 5·85) if they had adequate PPE, 5·06 (95% CI: 3·90 to 6·57) for reused PPE, and 5·91 (95% CI: 4·53 to 7·71) for inadequate PPE. Interpretation: Frontline HCWs had a significantly increased risk of COVID-19 infection, highest among HCWs who reused PPE or had inadequate access to PPE. However, adequate supplies of PPE did not completely mitigate high-risk exposures. Funding: Zoe Global Ltd., Wellcome Trust, EPSRC, NIHR, UK Research and Innovation, Alzheimer's Society, NIH, NIOSH, Massachusetts Consortium on Pathogen Readiness.

Real-time tracking of self-reported symptoms to predict potential COVID-19.

A total of 2,618,862 participants reported their potential symptoms of COVID-19 on a smartphone-based app. Among the 18,401 who had undergone a SARS-CoV-2 test, the proportion of participants who reported loss of smell and taste was higher in those with a positive test result (4,668 of 7,178 individuals; 65.03%) than in those with a negative test result (2,436 of 11,223 participants; 21.71%) (odds ratio = 6.74; 95% confidence interval = 6.31-7.21). A model combining symptoms to predict probable infection was applied to the data from all app users who reported symptoms (805,753) and predicted that 140,312 (17.42%) participants are likely to have COVID-19.

Rapid implementation of mobile technology for real-time epidemiology of COVID-19.

The rapid pace of the coronavirus disease 2019 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) presents challenges to the robust collection of population-scale data to address this global health crisis. We established the COronavirus Pandemic Epidemiology (COPE) Consortium to unite scientists with expertise in big data research and epidemiology to develop the COVID Symptom Study, previously known as the COVID Symptom Tracker, mobile application. This application-which offers data on risk factors, predictive symptoms, clinical outcomes, and geographical hotspots-was launched in the United Kingdom on 24 March 2020 and the United States on 29 March 2020 and has garnered more than 2.8 million users as of 2 May 2020. Our initiative offers a proof of concept for the repurposing of existing approaches to enable rapidly scalable epidemiologic data collection and analysis, which is critical for a data-driven response to this public health challenge.

A k-Space Model of Movement Artefacts: Application to Segmentation Augmentation and Artefact Removal.

Patient movement during the acquisition of magnetic resonance images (MRI) can cause unwanted image artefacts. These artefacts may affect the quality of clinical diagnosis and cause errors in automated image analysis. In this work, we present a method for generating realistic motion artefacts from artefact-free magnitude MRI data to be used in deep learning frameworks, increasing training appearance variability and ultimately making machine learning algorithms such as convolutional neural networks (CNNs) more robust to the presence of motion artefacts. By modelling patient movement as a sequence of randomly-generated, 'demeaned', rigid 3D affine transforms, we resample artefact-free volumes and combine these in k-space to generate motion artefact data. We show that by augmenting the training of semantic segmentation CNNs with artefacts, we can train models that generalise better and perform more reliably in the presence of artefact data, with negligible cost to their performance on clean data. We show that the performance of models trained using artefact data on segmentation tasks on real-world test-retest image pairs is more robust. We also demonstrate that our augmentation model can be used to learn to retrospectively remove certain types of motion artefacts from real MRI scans. Finally, we show that measures of uncertainty obtained from motion augmented CNN models reflect the presence of artefacts and can thus provide relevant information to ensure the safe usage of deep learning extracted biomarkers in a clinical pipeline.

Multi-domain Adaptation in Brain MRI Through Paired Consistency and Adversarial Learning.

Supervised learning algorithms trained on medical images will often fail to generalize across changes in acquisition parameters. Recent work in domain adaptation addresses this challenge and successfully leverages labeled data in a source domain to perform well on an unlabeled target domain. Inspired by recent work in semi-supervised learning we introduce a novel method to adapt from one source domain to n target domains (as long as there is paired data covering all domains). Our multi-domain adaptation method utilises a consistency loss combined with adversarial learning. We provide results on white matter lesion hyperintensity segmentation from brain MRIs using the MICCAI 2017 challenge data as the source domain and two target domains. The proposed method significantly outperforms other domain adaptation baselines.

As Easy as 1, 2 4? Uncertainty in Counting Tasks for Medical Imaging.

Counting is a fundamental task in biomedical imaging and count is an important biomarker in a number of conditions. Estimating the uncertainty in the measurement is thus vital to making definite, informed conclusions. In this paper, we first compare a range of existing methods to perform counting in medical imaging and suggest ways of deriving predictive intervals from these. We then propose and test a method for calculating intervals as an output of a multi-task network. These predictive intervals are optimised to be as narrow as possible, while also enclosing a desired percentage of the data. We demonstrate the effectiveness of this technique on histopathological cell counting and white matter hyperintensity counting. Finally, we offer insight into other areas where this technique may apply.