Explore the vast range of titles available.

In this work we present a three-stage Machine Learning strategy to country-level risk classification based on countries that are reporting COVID-19 information. A K% binning discretisation (K = 25) is used to create four risk groups of countries based on the risk of transmission (coronavirus cases per million population), risk of mortality (coronavirus deaths per million population), and risk of inability to test (coronavirus tests per million population). The four risk groups produced by K% binning are labelled as 'low', 'medium-low', 'medium-high', and 'high'. Coronavirus-related data are then removed and the attributes for prediction of the three types of risk are given as the geopolitical and demographic data describing each country. Thus, the calculation of class label is based on coronavirus data but the input attributes are country-level information regardless of coronavirus data. The three four-class classification problems are then explored and benchmarked through leave-one-country-out cross validation to find the strongest model, producing a Stack of Gradient Boosting and Decision Tree algorithms for risk of transmission, a Stack of Support Vector Machine and Extra Trees for risk of mortality, and a Gradient Boosting algorithm for the risk of inability to test. It is noted that high risk for inability to test is often coupled with low risks for transmission and mortality, therefore the risk of inability to test should be interpreted first, before consideration is given to the predicted transmission and mortality risks. Finally, the approach is applied to more recent risk levels to data from September 2020 and weaker results are noted due to the growth of international collaboration detracting useful knowledge from country-level attributes which suggests that similar machine learning approaches are more useful prior to situations later unfolding.

Contemporary Artificial Intelligence technologies allow for the employment of Computer Vision to discern good crops from bad, providing a step in the pipeline of selecting healthy fruit from undesirable fruit, such as those which are mouldy or gangrenous. State-of-the-art works in the field report high accuracy results on small datasets (<1000 images), which are not representative of the population regarding real-world usage. The goals of this study are to further enable real-world usage by improving generalisation with data augmentation as well as to reduce overfitting and energy usage through model pruning. In this work, we suggest a machine learning pipeline that combines the ideas of fine-tuning, transfer learning, and generative model-based training data augmentation towards improving fruit quality image classification. A linear network topology search is performed to tune a VGG16 lemon quality classification model using a publicly-available dataset of 2690 images. We find that appending a 4096 neuron fully connected layer to the convolutional layers leads to an image classification accuracy of 83.77%. We then train a Conditional Generative Adversarial Network on the training data for 2000 epochs, and it learns to generate relatively realistic images. Grad-CAM analysis of the model trained on real photographs shows that the synthetic images can exhibit classifiable characteristics such as shape, mould, and gangrene. A higher image classification accuracy of 88.75% is then attained by augmenting the training with synthetic images, arguing that Conditional Generative Adversarial Networks have the ability to produce new data to alleviate issues of data scarcity. Finally, model pruning is performed via polynomial decay, where we find that the Conditional GAN-augmented classification network can retain 81.16% classification accuracy when compressed to 50% of its original size.

The great majority of globally circulating pathogens go undetected, undermining patient care and hindering outbreak preparedness and response. To enable routine surveillance and comprehensive diagnostic applications, there is a need for detection technologies that can scale to test many samples 1 , 2 – 3 while simultaneously testing for many pathogens 4 , 5 – 6 . Here, we develop Combinatorial Arrayed Reactions for Multiplexed Evaluation of Nucleic acids (CARMEN), a platform for scalable, multiplexed pathogen detection. In the CARMEN platform, nanolitre droplets containing CRISPR-based nucleic acid detection reagents 7 self-organize in a microwell array 8 to pair with droplets of amplified samples, testing each sample against each CRISPR RNA (crRNA) in replicate. The combination of CARMEN and Cas13 detection (CARMEN–Cas13) enables robust testing of more than 4,500 crRNA–target pairs on a single array. Using CARMEN–Cas13, we developed a multiplexed assay that simultaneously differentiates all 169 human-associated viruses with at least 10 published genome sequences and rapidly incorporated an additional crRNA to detect the causative agent of the 2020 COVID-19 pandemic. CARMEN–Cas13 further enables comprehensive subtyping of influenza A strains and multiplexed identification of dozens of HIV drug-resistance mutations. The intrinsic multiplexing and throughput capabilities of CARMEN make it practical to scale, as miniaturization decreases reagent cost per test by more than 300-fold. Scalable, highly multiplexed CRISPR-based nucleic acid detection shifts diagnostic and surveillance efforts from targeted testing of high-priority samples to comprehensive testing of large sample sets, greatly benefiting patients and public health 9 , 10 – 11 . CRISPR-based nucleic acid detection is used in a platform that can simultaneously detect 169 human-associated viruses in multiple samples, providing scalable, multiplexed pathogen detection aimed at routine surveillance for public health.

The development of consumer sleep-tracking technologies has outpaced the scientific evaluation of their accuracy. In this study, five consumer sleep-tracking devices, research-grade actigraphy, and polysomnography were used simultaneously to monitor the overnight sleep of fifty-three young adults in the lab for one night. Biases and limits of agreement were assessed to determine how sleep stage estimates for each device and research-grade actigraphy differed from polysomnography-derived measures. Every device, except the Garmin Vivosmart, was able to estimate total sleep time comparably to research-grade actigraphy. All devices overestimated nights with shorter wake times and underestimated nights with longer wake times. For light sleep, absolute bias was low for the Fitbit Inspire and Fitbit Versa. The Withings Mat and Garmin Vivosmart overestimated shorter light sleep and underestimated longer light sleep. The Oura Ring underestimated light sleep of any duration. For deep sleep, bias was low for the Withings Mat and Garmin Vivosmart while other devices overestimated shorter and underestimated longer times. For REM sleep, bias was low for all devices. Taken together, these results suggest that proportional bias patterns in consumer sleep-tracking technologies are prevalent and could have important implications for their overall accuracy.

Background Choosing to have dialysis or conservative kidney management is often challenging for older people with advanced kidney disease. While we know that clinical communication has a major impact on patients’ treatment decision-making, little is known about how this occurs in practice. The OSCAR study (Optimising Staff-Patient Communication in Advanced Renal disease) aimed to identify how clinicians present kidney failure treatment options in consultations with older patients and the implications of this for patient engagement. Methods An observational, multi-method study design was adopted. Outpatient consultations at four UK renal units were video-recorded, and patients completed a post-consultation measure of shared decision-making (SDM-Q-9). Units were sampled according to variable rates of conservative management. Eligible patients were ≥ 65 years old with an eGFR of ≤ 20 mls/min/1.73m 2 within the last 6 months. Video-recordings were screened to identify instances where clinicians presented both dialysis and conservative management. These instances were transcribed in fine-grained detail and recurrent practices identified using conversation-analytic methods, an empirical, observational approach to studying language and social interaction. Results 110 outpatient consultations were recorded (105 video, 5 audio only), involving 38 clinicians (doctors and nurses) and 94 patients: mean age 77 (65–97); 61 males/33 females; mean eGFR 15 (range 4–23). There were 21 instances where clinicians presented both dialysis and conservative management. Two main practices were identified: (1) Conservative management and dialysis both presented as the main treatment options; (2) Conservative management presented as a subordinate option to dialysis. The first practice was less commonly used (6 vs. 15 cases), but associated with more opportunities in the conversation for patients to ask questions and share their perspective, through which they tended to evaluate conservative management as an option that was potentially personally relevant. This practice was also associated with significantly higher post-consultation ratings of shared decision-making among patients (SDM-Q-9 median total score 24 vs. 37, p  = 0.041). Conclusions Presenting conservative management and dialysis as on an equal footing enables patient to take a more active role in decision-making. Findings should inform clinical communication skills training and education. Clinical trial number No trial number as this is not a clinical trial.

Preventing SARS-CoV-2 infection by modulating viral host receptors, such as angiotensin-converting enzyme 2 (ACE2) 1 , could represent a new chemoprophylactic approach for COVID-19 that complements vaccination 2 , 3 . However, the mechanisms that control the expression of ACE2 remain unclear. Here we show that the farnesoid X receptor (FXR) is a direct regulator of ACE2 transcription in several tissues affected by COVID-19, including the gastrointestinal and respiratory systems. We then use the over-the-counter compound z-guggulsterone and the off-patent drug ursodeoxycholic acid (UDCA) to reduce FXR signalling and downregulate ACE2 in human lung, cholangiocyte and intestinal organoids and in the corresponding tissues in mice and hamsters. We show that the UDCA-mediated downregulation of ACE2 reduces susceptibility to SARS-CoV-2 infection in vitro, in vivo and in human lungs and livers perfused ex situ. Furthermore, we reveal that UDCA reduces the expression of ACE2 in the nasal epithelium in humans. Finally, we identify a correlation between UDCA treatment and positive clinical outcomes after SARS-CoV-2 infection using retrospective registry data, and confirm these findings in an independent validation cohort of recipients of liver transplants. In conclusion, we show that FXR has a role in controlling ACE2 expression and provide evidence that modulation of this pathway could be beneficial for reducing SARS-CoV-2 infection, paving the way for future clinical trials. FXR regulates the levels of ACE2 in tissues of the respiratory and gastrointestinal systems that are affected by COVID-19, and inhibiting FXR with ursodeoxycholic acid downregulates ACE2 and reduces susceptibility to SARS-CoV-2 infection.

Ice shelves cover ~1.6 million km2 of the Antarctic continental shelf and are sensitive indicators of climate change. With ice-shelf retreat, aphotic marine environments transform into new open-water spaces of photo-induced primary production and associated organic matter export to the benthos. Predicting how Antarctic seafloor assemblages may develop following ice-shelf loss requires knowledge of assemblages bordering the ice-shelf margins, which are relatively undocumented. This study investigated seafloor assemblages, by taxa and functional groups, in a coastal polynya adjacent to the Larsen C Ice Shelf front, western Weddell Sea. The study area is rarely accessed, at the frontline of climate change, and located within a CCAMLR-proposed international marine protected area. Four sites, ~1 to 16 km from the ice-shelf front, were explored for megabenthic assemblages, and potential environmental drivers of assemblage structures were assessed. Faunal density increased with distance from the ice shelf, with epifaunal deposit-feeders a surrogate for overall density trends. Faunal richness did not exhibit a significant pattern with distance from the ice shelf and was most variable at sites closest to the ice-shelf front. Faunal assemblages significantly differed in composition among sites, and those nearest to the ice shelf were the most dissimilar; however, ice-shelf proximity did not emerge as a significant driver of assemblage structure. Overall, the study found a biologically-diverse and complex seafloor environment close to an ice-shelf front and provides ecological baselines for monitoring benthic ecosystem responses to environmental change, supporting marine management.

PurposeThe optimal ventilatory settings in patients after cardiac arrest and their association with outcome remain unclear. The aim of this study was to describe the ventilatory settings applied in the first 72 h of mechanical ventilation in patients after out-of-hospital cardiac arrest and their association with 6-month outcomes.MethodsPreplanned sub-analysis of the Target Temperature Management-2 trial. Clinical outcomes were mortality and functional status (assessed by the Modified Rankin Scale) 6 months after randomization.ResultsA total of 1848 patients were included (mean age 64 [Standard Deviation, SD = 14] years). At 6 months, 950 (51%) patients were alive and 898 (49%) were dead. Median tidal volume (VT) was 7 (Interquartile range, IQR = 6.2–8.5) mL per Predicted Body Weight (PBW), positive end expiratory pressure (PEEP) was 7 (IQR = 5–9) cmH20, plateau pressure was 20 cmH20 (IQR = 17–23), driving pressure was 12 cmH20 (IQR = 10–15), mechanical power 16.2 J/min (IQR = 12.1–21.8), ventilatory ratio was 1.27 (IQR = 1.04–1.6), and respiratory rate was 17 breaths/minute (IQR = 14–20). Median partial pressure of oxygen was 87 mmHg (IQR = 75–105), and partial pressure of carbon dioxide was 40.5 mmHg (IQR = 36–45.7). Respiratory rate, driving pressure, and mechanical power were independently associated with 6-month mortality (omnibus p-values for their non-linear trajectories: p < 0.0001, p = 0.026, and p = 0.029, respectively). Respiratory rate and driving pressure were also independently associated with poor neurological outcome (odds ratio, OR = 1.035, 95% confidence interval, CI = 1.003–1.068, p = 0.030, and OR = 1.005, 95% CI = 1.001–1.036, p = 0.048). A composite formula calculated as [(4*driving pressure) + respiratory rate] was independently associated with mortality and poor neurological outcome.ConclusionsProtective ventilation strategies are commonly applied in patients after cardiac arrest. Ventilator settings in the first 72 h after hospital admission, in particular driving pressure and respiratory rate, may influence 6-month outcomes.

Background Early‐life education quality has been associated with dementia risk and late‐life cognitive functioning. However, the association between education quality and neuroimaging outcomes remains unclear. Methods These analyses utilized data from 450 participants in two harmonized cohorts of racially and ethnically diverse adults aged 50 years and older (KHANDLE and STAR) who completed brain Magnetic Resonance Imaging and whose self‐reported school location at 9th grade could be linked to historical educational quality data from the National Center for Education Statistics. Three county‐level indicators of 9th grade education quality were categorized into tertiles based on the pooled dataset and treated as ordinal variables with the lowest tertile serving as the reference: percent attendance, teacher‐student ratio (number of teachers per 30 students), and term length. Measures of gray matter volume, hippocampal volume, white matter hyperintensity volume (log‐transformed), and ventricular volume (log‐transformed) were residualized on total cranial volume. Linear regression models with robust standard errors estimated associations between education quality measures and each imaging marker, adjusting for age at imaging, demographics, childhood socioeconomic status, study cohort, and history of racial segregation in state of 9th grade education. Sensitivity analyses examined each education quality measures separately, adjusting for covariates. Results Higher teacher‐student ratio was associated with greater gray matter volume (β = 6.25, 95% CI: 1.57, 10.92) but was not associated hippocampal volume (β = ‐0.01, 95% CI: ‐0.14, 0.11), ventricular volume (β = ‐0.04, 95% CI: ‐0.13, 0.04), or white matter hyperintensity (β = ‐0.02, 95% CI: ‐0.31, 0.26) (Table 2). Higher percent attendance was associated with greater ventricular volume (β = 0.10, 95% CI: 0.03, 0.18) but was not associated with gray matter volume (β = ‐3.12, 95% CI: ‐7.07, ‐0.82), hippocampal volume (β = ‐0.06, 95% CI: (‐0.17, 0.05), or white matter hyperintensity (β = ‐0.02, 95% CI: ‐0.26, 0.22). Term length was not associated with any marker of neurodegeneration or vascular brain injury. Results were similar when examining each education quality measure separately. Conclusions County‐level teacher‐student ratio and percent attendance, markers of education quality, had qualitatively different associations with imaging markers of neurodegeneration. Additional studies in larger samples are needed to confirm and better understand underlying reasons for these potential differences.

PurposeThe 2021 guidelines endorsed by the European Resuscitation Council (ERC) and the European Society of Intensive Care Medicine (ESICM) recommend using highly malignant electroencephalogram (EEG) patterns (HMEP; suppression or burst-suppression) at > 24 h after cardiac arrest (CA) in combination with at least one other concordant predictor to prognosticate poor neurological outcome. We evaluated the prognostic accuracy of HMEP in a large multicentre cohort and investigated the added value of absent EEG reactivity.MethodsThis is a pre-planned prognostic substudy of the Targeted Temperature Management trial 2. The presence of HMEP and background reactivity to external stimuli on EEG recorded > 24 h after CA was prospectively reported. Poor outcome was measured at 6 months and defined as a modified Rankin Scale score of 4–6. Prognostication was multimodal, and withdrawal of life-sustaining therapy (WLST) was not allowed before 96 h after CA.Results845 patients at 59 sites were included. Of these, 579 (69%) had poor outcome, including 304 (36%) with WLST due to poor neurological prognosis. EEG was recorded at a median of 71 h (interquartile range [IQR] 52–93) after CA. HMEP at > 24 h from CA had 50% [95% confidence interval [CI] 46–54] sensitivity and 93% [90–96] specificity to predict poor outcome. Specificity was similar (93%) in 541 patients without WLST. When HMEP were unreactive, specificity improved to 97% [94–99] (p = 0.008).ConclusionThe specificity of the ERC-ESICM-recommended EEG patterns for predicting poor outcome after CA exceeds 90% but is lower than in previous studies, suggesting that large-scale implementation may reduce their accuracy. Combining HMEP with an unreactive EEG background significantly improved specificity. As in other prognostication studies, a self-fulfilling prophecy bias may have contributed to observed results.