Search

Abstract Potentially blinding corneal diseases alter the morphology of the human cornea. At the early stages of disease development, these changes occur at the cellular level. The ability to visualize and quantify such changes can lead to early diagnostics of corneal pathologies, which is pivotal for the long-term preservation of vision. Here we present a Powell Lens-based Line-Field Optical Coherence Tomography system that combines high spatial resolution (2.4 μm × 2.2 μm × 1.7 μm (x × y × z)) in biological tissue, sufficient to resolve individual cells, high sensitivity (90.5 dB), sufficient to image the semi-transparent human cornea, and fast image acquisition rate (~ 2,400 fps), sufficient to suppress most involuntary eye motion artifacts and allow for contactless, in-vivo imaging of the cellular structure of the human cornea. Volumetric images acquired in-vivo from corneas of healthy subjects show epithelial, endothelial and keratocytes cells, as well as sub-basal and stromal nerves. The system’s high axial resolution also allows for volumetric morphometry of the corneal endothelium, Descemet’s membrane and the pre-Descemet’s (Dua) layer.

Abstract Self-supervised pretraining has been observed to be effective at improving feature representations for transfer learning, leveraging large amounts of unlabelled data. This review summarizes recent research into its usage in X-ray, computed tomography, magnetic resonance, and ultrasound imaging, concentrating on studies that compare self-supervised pretraining to fully supervised learning for diagnostic tasks such as classification and segmentation. The most pertinent finding is that self-supervised pretraining generally improves downstream task performance compared to full supervision, most prominently when unlabelled examples greatly outnumber labelled examples. Based on the aggregate evidence, recommendations are provided for practitioners considering using self-supervised learning. Motivated by limitations identified in current research, directions and practices for future study are suggested, such as integrating clinical knowledge with theoretically justified self-supervised learning methods, evaluating on public datasets, growing the modest body of evidence for ultrasound, and characterizing the impact of self-supervised pretraining on generalization.

The core idea of Multi-Task Learning (MTL) is to develop neural networks with a shared feature extraction backbone and multiple prediction heads, each capable of inferring a different task simultaneously. Parameters in the backbone contribute to all tasks while those in the prediction heads contribute to only one or fewer tasks. Challenges arise when multiple tasks compete for resource. Existing methods focus on resource competition in shared parameters and proposed explanatory factors of task conflicts, task dominance, and gradient stability. However the fundamental nature of MTL is still understudied. In this paper, instead of following the existing methodology research directions, we carry out large-scale empirical study and provide deeper insight on understanding MTL. In particular, instead of focusing on resource competition in the shared parameters in the backbone, we shift our attention to resource competition in the backbone output, which is the embedded representation that is shared by all prediction heads. We show that the existing explanatory factors display weak causal relationship with model performance. We propose a novel measurement, which we term Feature Disentanglement, and show that understanding MTL problems from the perspective of how the shared representation is leveraged by different prediction heads, is a more faithful and reliable way than that from the perspective of how supervision signals from different tasks are interfering in the shared parameters. Additionally, it has been a commonly employed technique to replace gradients w.r.t. shared parameters with gradients w.r.t. shared representation for reduced computation. We conduct a comprehensive study and show that unless a theoretical analysis could be developed, there is not general guarantee that this fast approximation technique would work in practice.

Significant advances have been made in recent years in improving the robustness of deep neural networks, particularly under adversarial machine learning scenarios where the data has been contaminated to fool networks into making undesirable predictions. However, such improvements in adversarial robustness has often come at a significant cost in model accuracy when dealing with uncontaminated data (i.e., clean data), making such defense mechanisms challenging to adapt for real-world practical scenarios where data is primarily clean and accuracy needs to be high. Motivated to find a better balance between adversarial robustness and clean data accuracy, we propose a new model-agnostic adversarial defense mechanism named Dual-model Bounded Divergence (DBD), driven by a theoretical and empirical analysis of the bias-variance trade-off within an adversarial machine learning context. More specifically, the proposed DBD mechanism is premised on the observation that the variance in deep neural networks tends to increase in the presence of adversarial perturbations in the input data. As such, DBD employs a gating mechanism to decide on the final model prediction output based on a novel dual-model variance measure (coined DBD Variance), which is a bounded version of KL-Divergence between models. Not only is the proposed DBD mechanism itself training-free, but it can be combined with existing adversarial defense mechanisms to boost the balance between clean data accuracy and adversarial robustness. Comprehensive experimental results across over 10 different state-of-the-art adversarial defense mechanisms using both CIFAR-10 and ImageNet benchmark datasets across different adversarial attacks (e.g., APGD, AutoAttack) demonstrates that the integration of DBD can lead to as much as a 6% improvement on clean data accuracy without compromising much on adversarial robustness.

Skin cancer is the most common type of cancer in the United States and is estimated to affect one in five Americans. Recent advances have demonstrated strong performance on skin cancer detection, as exemplified by state of the art performance in the SIIM-ISIC Melanoma Classification Challenge; however, these solutions leverage ensembles of complex deep neural architectures requiring immense storage and computation costs, and therefore may not be tractable. A recent movement for TinyML applications is integrating Double-Condensing Attention Condensers (DC-AC) into a self-attention neural network backbone architecture to allow for faster and more efficient computation. This paper explores leveraging an efficient self-attention structure to detect skin cancer in skin lesion images and introduces a deep neural network design with DC-AC customized for skin cancer detection from skin lesion images. We demonstrate that our approach with only 1.6 million parameters and 0.32 GFLOPs achieves better performance compared to traditional architecture designs as it obtains an area under the ROC curve of 0.90 on the public ISIC 2020 test set and 0.89 on the private ISIC test set, over 0.13 above the best Cancer-Net SCa network architecture design. The final model is publicly available as a part of a global open-source initiative dedicated to accelerating advancement in machine learning to aid clinicians in the fight against cancer. Future work of this research includes iterating on the design of the selected network architecture and refining the approach to generalize to other forms of cancer.

IntroductionSelf-supervised learning (SSL) is a strategy for addressing the paucity of labelled data in medical imaging by learning representations from unlabelled images. Contrastive and non-contrastive SSL methods produce learned representations that are similar for pairs of related images. Such pairs are commonly constructed by randomly distorting the same image twice. The videographic nature of ultrasound offers flexibility for defining the similarity relationship between pairs of images.MethodsWe investigated the effect of utilizing proximal, distinct images from the same B-mode ultrasound video as pairs for SSL. Additionally, we introduced a sample weighting scheme that increases the weight of closer image pairs and demonstrated how it can be integrated into SSL objectives.ResultsNamed Intra-Video Positive Pairs (IVPP), the method surpassed previous ultrasound-specific contrastive learning methods' average test accuracy on COVID-19 classification with the POCUS dataset by ≥ 1.3%. Detailed investigations of IVPP's hyperparameters revealed that some combinations of IVPP hyperparameters can lead to improved or worsened performance, depending on the downstream task.DiscussionGuidelines for practitioners were synthesized based on the results, such as the merit of IVPP with task-specific hyperparameters, and the improved performance of contrastive methods for ultrasound compared to non-contrastive counterparts.

Unidentified Anomalous Phenomena (UAP) have become a serious topic in the US Congress, and new legislation has been released outlining a plan for declassification for the public. There are numerous factors that could lead an individual to mistakenly think they saw a UAP, and one of those factors might be the proclivities of the personalities that observe what they think to be a UAP. This study examined the big five personality traits: extraversion, neuroticism, openness, agreeableness, and conscientiousness, as well as schizotypy traits, to see if UAP experiencers could be distinguished from people who had not seen a UAP. The study included 206 participants, with 103 people who self-reported to have seen a UAP. Latent profile analysis was conducted on the personality variables to explore the grouping of participants. Group one was average on the traits, a second cluster was labeled as the Neurotic/Schizotypy group, which was high in neuroticism and schizotypy traits, and a third group was labeled as O-ACE, which were high on openness, agreeableness, conscientiousness, and extraversion but low on neuroticism and schizotypy traits. The findings indicated that the O-ACE group was more likely to see UAP, but this effect was not strong. A presumptuous stereotype exists in the general public that people who see UAP are probably people who are emotionally reactive (neurotic) and vulnerable to perceptual and cognitive abnormalities, but this was not evident in our data. We also found that the descriptive UAP accounts by the general public were similar to the descriptions provided by military witnesses. It was also of note that only 28% of participants reported their sightings anywhere, and 14% used a UFO reporting organization, which suggests that events are vastly underreported. Stigma and a lack of places to legitimately report sightings appeared to be primary barriers. The conclusion of this study is that personality factors are an insufficient explanation for most UAP sightings.

Abstract Since the World Health Organization declared COVID-19 a pandemic in 2020, the global community has faced ongoing challenges in controlling and mitigating the transmission of the SARS-CoV-2 virus, as well as its evolving subvariants and recombinants. A significant challenge during the pandemic has not only been the accurate detection of positive cases but also the efficient prediction of risks associated with complications and patient survival probabilities. These tasks entail considerable clinical resource allocation and attention. In this study, we introduce COVID-Net Biochem, a versatile and explainable framework for constructing machine learning models. We apply this framework to predict COVID-19 patient survival and the likelihood of developing Acute Kidney Injury during hospitalization, utilizing clinical and biochemical data in a transparent, systematic approach. The proposed approach advances machine learning model design by seamlessly integrating domain expertise with explainability tools, enabling model decisions to be based on key biomarkers. This fosters a more transparent and interpretable decision-making process made by machines specifically for medical applications. More specifically, the framework comprises two phases: In the first phase, referred to as the “clinician-guided design” phase, the dataset is preprocessed using explainable AI and domain expert input. To better demonstrate this phase, we prepared a benchmark dataset of carefully curated clinical and biochemical markers based on clinician assessments for survival and kidney injury prediction in COVID-19 patients. This dataset was selected from a patient cohort of 1366 individuals at Stony Brook University. Moreover, we designed and trained a diverse collection of machine learning models, encompassing gradient-based boosting tree architectures and deep transformer architectures, specifically for survival and kidney injury prediction based on the selected markers. In the second phase, called the “explainability-driven design refinement” phase, the proposed framework employs explainability methods to not only gain a deeper understanding of each model’s decision-making process but also to identify the overall impact of individual clinical and biochemical markers for bias identification. In this context, we used the models constructed in the previous phase for the prediction task and analyzed the explainability outcomes alongside a clinician with over 8 years of experience to gain a deeper understanding of the clinical validity of the decisions made. The explainability-driven insights obtained, in conjunction with the associated clinical feedback, are then utilized to guide and refine the training policies and architectural design iteratively. This process aims to enhance not only the prediction performance but also the clinical validity and trustworthiness of the final machine learning models. Employing the proposed explainability-driven framework, we attained 93.55% accuracy in survival prediction and 88.05% accuracy in predicting kidney injury complications. The models have been made available through an open-source platform. Although not a production-ready solution, this study aims to serve as a catalyst for clinical scientists, machine learning researchers, and citizen scientists to develop innovative and trustworthy clinical decision support solutions, ultimately assisting clinicians worldwide in managing pandemic outcomes.

Abstract Patients with Waldenström macroglobulinaemia (WM) are at increased risk of severe COVID‐19 infection and have poor immune responses to COVID‐19 vaccination. This study assessed whether a closely monitored pause in Bruton's Tyrosine Kinase inhibitor (BTKi) therapy might result in an improved humoral response to a 3rd COVID‐19 vaccine dose. Improved response was observed in WM patients who paused their BTKi, compared to a group who did not pause their BTKi. However, the response was attenuated after BTKi recommencement. This data contributes to our understanding of vaccination strategies in this patient group and may help inform consensus approaches in the future.

Abstract Background Nutrient dense food that supports health is a goal of food service in long-term care (LTC). The objective of this work was to characterize the “healthfulness” of foods in Canadian LTC and inflammatory potential of the LTC diet and how this varied by key covariates. Here, we define foods to have higher “healthfulness” if the are in accordance with the evidence-based 2019 Canada’s Food Guide, or with comparatively lower inflammatory potential. Methods We conducted a secondary analysis of the Making the Most of Mealtimes dataset (32 LTC homes; four provinces). A novel computational algorithm categorized food items from 3-day weighed food records into 68 expert-informed categories and Canada’s Food Guide (CFG) food groups. The dietary inflammatory potential of these food sources was assessed using the Dietary Inflammatory Index (DII). Comparisons were made by sex, diet texture, and nutritional status. Results Consumption patterns using expert-informed categories indicated no single protein or vegetable source was among the top 5 most commonly consumed foods. In terms of CFG’s groups, protein food sources (i.e., foods with a high protein content) represented the highest proportion of daily calorie intake (33.4%; animal-based: 31.6%, plant-based: 1.8%), followed by other foods (31.3%) including juice (9.8%), grains (25.0%; refined: 15.0%, whole: 10.0%), and vegetables/fruits (10.3%; plain: 4.9%, with additions: 5.4%). The overall DII score (mean, IQR) was positive (0.93, 0.23 to 1.75) indicating foods consumed tend towards a pro-inflammatory response. DII was significantly associated with sex (female higher; p

Ocular surface temperature (OST) is affected by changes in eye physiology caused by normal homeostasis, environmental changes, or systemic and local disease. OST can help a physician diagnose eye disease with improved accuracy and provide useful information for eye research. This paper presents a novel system, including novel hardware design and novel algorithms, capable of automatically measuring and tracking OST from the cornea over any period of time. The system uses an infrared (IR) camera and a visible (VIS) camera to capture synchronous thermal and visible videos, respectively, from the eye surface. The frames for each camera video sequence are then registered together (video registration) using two sets of control points. The points are manually selected on the first pair of timestamped IR/VIS frames and tracked over all subsequent frames using the Lucas–Kanade (LK) optical flow algorithm (point tracking). A mean square error (MSE) of 5.43±2.01 pixels was reported for salient point tracking of the IR video and 6.81±2.32 pixels for tracking of the VIS video. Overall MSE for registration was 5.03 ±1.82 pixels. The corneal area was segmented in the VIS images and localized on the IR images using the semantic segmentation method (corneal segmentation). A mean Intersection over Union (IoU) of 94.6% was found, representing the accuracy of corneal segmentation. A system for measuring and tracking eye surface temperature over time was developed. The system is able to localize the cornea on both VIS and IR images, and report temperature profiles of the cornea over the period of measurement. Experimental results show that the system can work as a tool for measuring and tracking OST over time.

In this study, we investigated whether self-supervised pretraining could produce a neural network feature extractor applicable to multiple classification tasks in B-mode lung ultrasound analysis. When fine-tuning on three lung ultrasound tasks, pretrained models resulted in an improvement of the average across-task area under the receiver operating characteristic curve (AUC) by 0.032 and 0.061 on local and external test sets respectively. Compact nonlinear classifiers trained on features outputted by a single pretrained model did not improve performance across all tasks; however, they reduced inference time by 49% compared to the serial execution of separate fine-tuned models. When training using 1% of the available labels, pretrained models consistently outperformed fully supervised models, with a maximum observed test AUC increase of 0.396 for the task of view classification. Overall, the results indicate that self-supervised pretraining is a useful strategy for producing initial weights for lung ultrasound classifiers.

While significant research advances have been made in the field of deep reinforcement learning, there have been no concrete adversarial attack strategies in literature tailored for studying the vulnerability of deep reinforcement learning algorithms to membership inference attacks. In such attacking systems, the adversary targets the set of collected input data on which the deep reinforcement learning algorithm has been trained. To address this gap, we propose an adversarial attack framework designed for testing the vulnerability of a state-of-the-art deep reinforcement learning algorithm to a membership inference attack. In particular, we design a series of experiments to investigate the impact of temporal correlation, which naturally exists in reinforcement learning training data, on the probability of information leakage. Moreover, we compare the performance of collective and individual membership attacks against the deep reinforcement learning algorithm. Experimental results show that the proposed adversarial attack framework is surprisingly effective at inferring data with an accuracy exceeding 84% in individual and 97% in collective modes in three different continuous control Mujoco tasks, which raises serious privacy concerns in this regard. Finally, we show that the learning state of the reinforcement learning algorithm influences the level of privacy breaches significantly.

While no longer a public health emergency of international concern, COVID-19 remains an established and ongoing global health threat. As the global population continues to face significant negative impacts of the pandemic, there has been an increased usage of point-of-care ultrasound (POCUS) imaging as a low-cost, portable, and effective modality of choice in the COVID-19 clinical workflow. A major barrier to the widespread adoption of POCUS in the COVID-19 clinical workflow is the scarcity of expert clinicians who can interpret POCUS examinations, leading to considerable interest in artificial intelligence-driven clinical decision support systems to tackle this challenge. A major challenge to building deep neural networks for COVID-19 screening using POCUS is the heterogeneity in the types of probes used to capture ultrasound images (e.g., convex vs. linear probes), which can lead to very different visual appearances. In this study, we propose an analytic framework for COVID-19 assessment able to consume ultrasound images captured by linear and convex probes. We analyze the impact of leveraging extended linear-convex ultrasound augmentation learning on producing enhanced deep neural networks for COVID-19 assessment, where we conduct data augmentation on convex probe data alongside linear probe data that have been transformed to better resemble convex probe data. The proposed explainable framework, called COVID-Net L2C-ULTRA, employs an efficient deep columnar anti-aliased convolutional neural network designed via a machine-driven design exploration strategy. Our experimental results confirm that the proposed extended linear–convex ultrasound augmentation learning significantly increases performance, with a gain of 3.9% in test accuracy and 3.2% in AUC, 10.9% in recall, and 4.4% in precision. The proposed method also demonstrates a much more effective utilization of linear probe images through a 5.1% performance improvement in recall when such images are added to the training dataset, while all other methods show a decrease in recall when trained on the combined linear–convex dataset. We further verify the validity of the model by assessing what the network considers to be the critical regions of an image with our contribution clinician.

With an aging population, numerous assistive and monitoring technologies are under development to enable older adults to age in place. To facilitate aging in place, predicting risk factors such as falls and hospitalization and providing early interventions are important. Much of the work on ambient monitoring for risk prediction has centered on gait speed analysis, utilizing privacy-preserving sensors like radar. Despite compelling evidence that monitoring step length in addition to gait speed is crucial for predicting risk, radar-based methods have not explored step length measurement in the home. Furthermore, laboratory experiments on step length measurement using radars are limited to proof-of-concept studies with few healthy subjects. To address this gap, a radar-based step length measurement system for the home is proposed based on detection and tracking using a radar point cloud followed by Doppler speed profiling of the torso to obtain step lengths in the home. The proposed method was evaluated in a clinical environment involving 35 frail older adults to establish its validity. Additionally, the method was assessed in people’s homes, with 21 frail older adults who had participated in the clinical assessment. The proposed radar-based step length measurement method was compared to the gold-standard Zeno Walkway Gait Analysis System, revealing a 4.5 cm/8.3% error in a clinical setting. Furthermore, it exhibited excellent reliability (ICC(2,k) = 0.91, 95% CI 0.82 to 0.96) in uncontrolled home settings. The method also proved accurate in uncontrolled home settings, as indicated by a strong consistency (ICC(3,k) = 0.81 (95% CI 0.53 to 0.92)) between home measurements and in-clinic assessments.

Abstract Background Skin cancer continues to be the most frequently diagnosed form of cancer in the U.S., with not only significant effects on health and well-being but also significant economic costs associated with treatment. A crucial step to the treatment and management of skin cancer is effective early detection with key screening approaches such as dermoscopy examinations, leading to stronger recovery prognoses. Motivated by the advances of deep learning and inspired by the open source initiatives in the research community, in this study we introduce Cancer-Net SCa, a suite of deep neural network designs tailored for the detection of skin cancer from dermoscopy images that is open source and available to the general public. To the best of the authors’ knowledge, Cancer-Net SCa comprises the first machine-driven design of deep neural network architectures tailored specifically for skin cancer detection, one of which leverages attention condensers for an efficient self-attention design. Results We investigate and audit the behaviour of Cancer-Net SCa in a responsible and transparent manner through explainability-driven performance validation. All the proposed designs achieved improved accuracy when compared to the ResNet-50 architecture while also achieving significantly reduced architectural and computational complexity. In addition, when evaluating the decision making process of the networks, it can be seen that diagnostically relevant critical factors are leveraged rather than irrelevant visual indicators and imaging artifacts. Conclusion The proposed Cancer-Net SCa designs achieve strong skin cancer detection performance on the International Skin Imaging Collaboration (ISIC) dataset, while providing a strong balance between computation and architectural efficiency and accuracy. While Cancer-Net SCa is not a production-ready screening solution, the hope is that the release of Cancer-Net SCa in open source, open access form will encourage researchers, clinicians, and citizen data scientists alike to leverage and build upon them.

Abstract Background Depression is common in the human immunodeficiency virus (HIV)-hepatitis C virus (HCV) co-infected population. Demographic, behavioural, and clinical data collected in research settings may be of help in identifying those at risk for clinical depression. We aimed to predict the presence of depressive symptoms indicative of a risk of depression and identify important classification predictors using supervised machine learning. Methods We used data from the Canadian Co-infection Cohort, a multicentre prospective cohort, and its associated sub-study on Food Security (FS). The Center for Epidemiologic Studies Depression Scale-10 (CES-D-10) was administered in the FS sub-study; participants were classified as being at risk for clinical depression if scores ≥ 10. We developed two random forest algorithms using the training data (80%) and tenfold cross validation to predict the CES-D-10 classes—1. Full algorithm with all candidate predictors (137 predictors) and 2. Reduced algorithm using a subset of predictors based on expert opinion (46 predictors). We evaluated the algorithm performances in the testing data using area under the receiver operating characteristic curves (AUC) and generated predictor importance plots. Results We included 1,934 FS sub-study visits from 717 participants who were predominantly male (73%), white (76%), unemployed (73%), and high school educated (52%). At the first visit, median age was 49 years (IQR:43–54) and 53% reported presence of depressive symptoms with CES-D-10 scores ≥ 10. The full algorithm had an AUC of 0.82 (95% CI:0.78–0.86) and the reduced algorithm of 0.76 (95% CI:0.71–0.81). Employment, HIV clinical stage, revenue source, body mass index, and education were the five most important predictors. Conclusion We developed a prediction algorithm that could be instrumental in identifying individuals at risk for depression in the HIV-HCV co-infected population in research settings. Development of such machine learning algorithms using research data with rich predictor information can be useful for retrospective analyses of unanswered questions regarding impact of depressive symptoms on clinical and patient-centred outcomes among vulnerable populations.

Abstract Prostate cancer (PCa) is the second most common cancer in men worldwide and the most frequently diagnosed cancer among men in more developed countries. The prognosis of PCa is excellent if detected at an early stage, making early screening crucial for detection and treatment. In recent years, a new form of diffusion magnetic resonance imaging called correlated diffusion imaging (CDI) was introduced, and preliminary results show promise as a screening tool for PCa. In the largest study of its kind, we investigate the relationship between PCa presence and a new variant of CDI we term synthetic correlated diffusion imaging (CDI $$^s$$ s ), as well as its performance for PCa delineation compared to current standard MRI techniques [T2-weighted (T2w) imaging, diffusion-weighted imaging (DWI), and dynamic contrast-enhanced (DCE) imaging] across a cohort of 200 patient cases. Statistical analyses reveal that hyperintensity in CDI $$^s$$ s is a strong indicator of PCa presence and achieves strong delineation of clinically significant cancerous tissue compared to T2w, DWI, and DCE. These results suggest that CDI $$^s$$ s hyperintensity may be a powerful biomarker for the presence of PCa, and may have a clinical impact as a diagnostic aid for improving PCa screening.

Abstract Malnutrition is a multidomain problem affecting 54% of older adults in long-term care (LTC). Monitoring nutritional intake in LTC is laborious and subjective, limiting clinical inference capabilities. Recent advances in automatic image-based food estimation have not yet been evaluated in LTC settings. Here, we describe a fully automatic imaging system for quantifying food intake. We propose a novel deep convolutional encoder-decoder food network with depth-refinement (EDFN-D) using an RGB-D camera for quantifying a plate’s remaining food volume relative to reference portions in whole and modified texture foods. We trained and validated the network on the pre-labelled UNIMIB2016 food dataset and tested on our two novel LTC-inspired plate datasets (689 plate images, 36 unique foods). EDFN-D performed comparably to depth-refined graph cut on IOU (0.879 vs. 0.887), with intake errors well below typical 50% (mean percent intake error: $$-4.2$$ - 4.2 %). We identify how standard segmentation metrics are insufficient due to visual-volume discordance, and include volume disparity analysis to facilitate system trust. This system provides improved transparency, approximates human assessors with enhanced objectivity, accuracy, and precision while avoiding hefty semi-automatic method time requirements. This may help address short-comings currently limiting utility of automated early malnutrition detection in resource-constrained LTC and hospital settings.

Soil moisture (SM) estimation is a critical part of environmental and agricultural monitoring, with satellite-based microwave remote sensing being the main SM source. However, the limited spatial resolution of most current remote sensing SM products reduces their utility for many applications, such as evapotranspiration modeling and agriculture management. In this article, to address this issue, we propose a Bayesian deep image prior (BDIP) downscaling approach to estimate the high-resolution SM from satellite products. More specifically, the high-resolution SM estimation problem is formulated as a maximum a posteriori problem, and solved via a neural network comprising of a deep fully convolutional neural network (FCNN) for modeling the prior spatial correlation distribution of the underlying high-resolution SM variables, and a forward model characterizing the SM map degeneration process for modeling the data likelihood. As such, the proposed BDIP approach provides a statistical framework that integrates deep learning with forward modeling in a coherent manner for combining different sources of information, i.e., the knowledge in the forward model, the spatial correlation prior in FCNN architecture, and the remote sensing data and products. Experiments on the downscaling of SM active passive SM products using the moderate resolution imaging spectroradiometer products show that SM maps estimated using the proposed method provide greater spatial detail information than other downscaling methods, with the SM estimates very close to in situ measurements.

Background and Aims: The obesity epidemic has increased the risk of nonalcoholic fatty liver disease (NAFLD) in both the general and chronic hepatitis B (CHB) populations. Our study aims to determine the prevalence of NAFLD in patients with CHB based on controlled attenuation parameter (CAP) and the epidemiological, clinical, and virological factors associated with severe hepatic steatosis. Methods: The Canadian Hepatitis B Network cohort was utilized to provide a cross-sectional description of demographics, comorbidities, antiviral treatment, and hepatits B virus (HBV) tests. Liver fibrosis and steatosis were measured by transient elastography and CAP, respectively. Any grade and severe steatosis were defined as CAP >248 and >280 dB/m, respectively. Advanced liver fibrosis was defined as transient elastography measurement >10.7 kPa. Results: In 1178 patients with CHB (median age: 47.4%, 57.7% males, 75.7% Asian, 13% African, 6.5% White, 86% HBV e antigen negative, median HBV DNA of 2.44 log10IU/mL, 42.7% receiving treatment), the prevalence of any grade and severe steatosis was 53% and 36%, respectively. In the multivariate analysis, obesity was a significant predictor for severe steatosis (adjusted odds ratio: 5.046, 95% confidence interval: 1.22–20.93). Severe steatosis was a determinant associated with viral load (adjusted odds ratio: 0.385, 95% confidence interval: 0.20–0.75, P

Summary: Background: Understanding what factors lead to youth polysubstance use (PSU) patterns and how the transitions between use patterns can inform the design and implementation of PSU prevention programs. We explore the dynamics of PSU patterns from a large cohort of Canadian secondary school students using machine learning techniques. Methods: We employed a multivariate latent Markov model (LMM) on COMPASS data, with a linked sample (N = 8824) of three-annual waves, Wave I (WI, 2016–17, as baseline), Wave II (WII, 2017–18), and Wave III (WIII, 2018–19). Substance use indicators, i.e., cigarette, e-cigarette, alcohol and marijuana, were self-reported and were categorized into never/occasional/current use. Outcomes: Four distinct use patterns were identified: no-use (S1), single-use of alcohol (S2), dual-use of e-cigarettes and alcohol (S3), and multi-use (S4). S1 had the highest prevalence (60.5%) at WI, however, S3 became the prominent use pattern (32.5%) by WIII. Most students remained in the same subgroup over time, particularly S4 had the highest transition probability (0.87) across the three-wave. With time, those who transitioned typically moved towards a higher use pattern, with the most and least likely transition occurring S2→S3 (0.45) and S3→S2 (

BackgroundHalf of long-term care (LTC) residents are malnourished, leading to increased hospitalization, mortality, and morbidity, with low quality of life. Current tracking methods are subjective and time-consuming. ObjectiveThis paper presented the automated food imaging and nutrient intake tracking technology designed for LTC. MethodsA needs assessment was conducted with 21 participating staff across 12 LTC and retirement homes. We created 2 simulated LTC intake data sets comprising modified (664/1039, 63.91% plates) and regular (375/1039, 36.09% plates) texture foods. Overhead red-green-blue-depth images of plated foods were acquired, and foods were segmented using a pretrained food segmentation network. We trained a novel convolutional autoencoder food feature extractor network using an augmented UNIMIB2016 food data set. A meal-specific food classifier was appended to the feature extractor and tested on our simulated LTC food intake data sets. Food intake (percentage) was estimated as the differential volume between classified full portion and leftover plates. ResultsThe needs assessment yielded 13 nutrients of interest, requirement for objectivity and repeatability, and account for real-world environmental constraints. For 12 meal scenarios with up to 15 classes each, the top-1 classification accuracy was 88.9%, with mean intake error of −0.4 (SD 36.7) mL. Nutrient intake estimation by volume was strongly linearly correlated with nutrient estimates from mass (r2=0.92-0.99), with good agreement between methods (σ=−2.7 to −0.01; 0 within each of the limits of agreement). ConclusionsThe automated food imaging and nutrient intake tracking approach is a deep learning–powered computational nutrient sensing system that appears to be feasible (validated accuracy against gold-standard weighed food method, positive end user engagement) and may provide a novel means for more accurate and objective tracking of LTC residents’ food intake to support and prevent malnutrition tracking strategies.

Computer vision and deep learning have the potential to improve medical artificial intelligence (AI) by assisting in diagnosis, prediction, and prognosis. However, the application of deep learning to medical image analysis is challenging due to limited data availability and imbalanced data. While model performance is undoubtedly essential for medical image analysis, model trust is equally important. To address these challenges, we propose TRUDLMIA, a trustworthy deep learning framework for medical image analysis, which leverages image features learned through self-supervised learning and utilizes a novel surrogate loss function to build trustworthy models with optimal performance. The framework is validated on three benchmark data sets for detecting pneumonia, COVID-19, and melanoma, and the created models prove to be highly competitive, even outperforming those designed specifically for the tasks. Furthermore, we conduct ablation studies, cross-validation, and result visualization and demonstrate the contribution of proposed modules to both model performance (up to 21%) and model trust (up to 5%). We expect that the proposed framework will support researchers and clinicians in advancing the use of deep learning for dealing with public health crises, improving patient outcomes, increasing diagnostic accuracy, and enhancing the overall quality of healthcare delivery.

Those experiencing homelessness in Canada are impacted inequitably by COVID-19 due to their increased exposure, vulnerability of environment and medical comorbidities, and their lack of access to preventive care and treatment in the context of the pandemic. In shelter environments one is unable to effectively physically distance, maintain hygiene, obtain a test, or isolate. As a result, unique strategies are required for this population to protect them and those who serve them. Recommendations are provided to reduce or prevent further negative consequences from the COVID-19 pandemic for people experiencing homelessness. These recommendations were informed by a systematic review of the literature, as well as a jurisdictional scan. Where evidence did not exist, expert consensus from key providers and those experiencing homelessness throughout Canada was included. These recommendations recognize the need for short-term interventions to mitigate the immediate risk to this community, including coordination of response, appropriate precautions and protective equipment, reducing congestion, cohorting, testing, case and contact management strategies, dealing with outbreaks, isolation centres, and immunization. Longer-term recommendations are also provided with a view to ending homelessness by addressing the root causes of homelessness and by the provision of adequate subsidized and supportive housing through a Housing First strategy. It is imperative that meaningful changes take place now in how we serve those experiencing homelessness and how we mitigate specific vulnerabilities. These recommendations call for intersectoral, collaborative engagement to work for solutions targeted towards protecting the most vulnerable within our community through both immediate actions and long-term planning to eliminate homelessness.

Neural architecture search has proven to be highly effective in the design of efficient convolutional neural networks that are better suited for mobile deployment than hand-designed networks. Hypothesizing that neural architecture search holds great potential for human pose estimation, we explore the application of neuroevolution, a form of neural architecture search inspired by biological evolution, in the design of 2D human pose networks for the first time. Additionally, we propose a new weight transfer scheme that enables us to accelerate neuroevolution in a flexible manner. Our method produces network designs that are more efficient and more accurate than state-of-the-art hand-designed networks. In fact, the generated networks process images at higher resolutions using less computation than previous hand-designed networks at lower resolutions, allowing us to push the boundaries of 2D human pose estimation. Our base network designed via neuroevolution, which we refer to as EvoPose2D-S, achieves comparable accuracy to SimpleBaseline while being 50% faster and $12.7\times $ smaller in terms of file size. Our largest network, EvoPose2D-L, achieves new state-of-the-art accuracy on the Microsoft COCO Keypoints benchmark, is $4.3\times $ smaller than its nearest competitor, and has similar inference speed. The code is publicly available at https://github.com/wmcnally/evopose2d.

As the COVID-19 pandemic devastates globally, the use of chest X-ray (CXR) imaging as a complimentary screening strategy to RT-PCR testing continues to grow given its routine clinical use for respiratory complaint. As part of the COVID-Net open source initiative, we introduce COVID-Net CXR-2, an enhanced deep convolutional neural network design for COVID-19 detection from CXR images built using a greater quantity and diversity of patients than the original COVID-Net. We also introduce a new benchmark dataset composed of 19,203 CXR images from a multinational cohort of 16,656 patients from at least 51 countries, making it the largest, most diverse COVID-19 CXR dataset in open access form. The COVID-Net CXR-2 network achieves sensitivity and positive predictive value of 95.5 and 97.0%, respectively, and was audited in a transparent and responsible manner. Explainability-driven performance validation was used during auditing to gain deeper insights in its decision-making behavior and to ensure clinically relevant factors are leveraged for improving trust in its usage. Radiologist validation was also conducted, where select cases were reviewed and reported on by two board-certified radiologists with over 10 and 19 years of experience, respectively, and showed that the critical factors leveraged by COVID-Net CXR-2 are consistent with radiologist interpretations.

Tuberculosis (TB) remains a global health problem, and is the leading cause of death from an infectious disease. A crucial step in the treatment of tuberculosis is screening high risk populations and the early detection of the disease, with chest x-ray (CXR) imaging being the most widely-used imaging modality. As such, there has been significant recent interest in artificial intelligence-based TB screening solutions for use in resource-limited scenarios where there is a lack of trained healthcare workers with expertise in CXR interpretation. Motivated by this pressing need and the recent recommendation by the World Health Organization (WHO) for the use of computer-aided diagnosis of TB in place of a human reader, we introduce TB-Net, a self-attention deep convolutional neural network tailored for TB case screening. We used CXR data from a multi-national patient cohort to train and test our models. A machine-driven design exploration approach leveraging generative synthesis was used to build a highly customized deep neural network architecture with attention condensers. We conducted an explainability-driven performance validation process to validate TB-Net's decision-making behavior. Experiments on CXR data from a multi-national patient cohort showed that the proposed TB-Net is able to achieve accuracy/sensitivity/specificity of 99.86/100.0/99.71%. Radiologist validation was conducted on select cases by two board-certified radiologists with over 10 and 19 years of experience, respectively, and showed consistency between radiologist interpretation and critical factors leveraged by TB-Net for TB case detection for the case where radiologists identified anomalies. The proposed TB-Net not only achieves high tuberculosis case detection performance in terms of sensitivity and specificity, but also leverages clinically relevant critical factors in its decision making process. While not a production-ready solution, we hope that the open-source release of TB-Net as part of the COVID-Net initiative will support researchers, clinicians, and citizen data scientists in advancing this field in the fight against this global public health crisis.

Background. Lung transplant patients are vulnerable to various forms of allograft injury, whether from acute rejection (AR) (encompassing acute cellular rejection [ACR] and antibody-mediated rejection [AMR]), chronic lung allograft dysfunction (CLAD), or infection (INFXN). Previous research indicates that donor-derived cell-free DNA (dd-cfDNA) is a promising noninvasive biomarker for the detection of AR and allograft injury. Our aim was to validate a clinical plasma dd-cfDNA assay for detection of AR and other allograft injury and to confirm and expand on dd-cfDNA and allograft injury associations observed in previous studies. Methods. We measured dd-cfDNA fraction using a novel single-nucleotide polymorphism-based assay in prospectively collected plasma samples paired with clinical-pathologic diagnoses. dd-cfDNA fraction was compared across clinical-pathologic cohorts: stable, ACR, AMR, isolated lymphocytic bronchiolitis, CLAD/neutrophilic-responsive allograft dysfunction (NRAD), and INFXN. Performance characteristics were calculated for AR and combined allograft injury (AR + CLAD/NRAD + INFXN) versus the stable cohort. Results. The study included 195 samples from 103 patients. Median dd-cfDNA fraction was significantly higher for ACR (1.43%, interquartile range [IQR]: 0.67%–2.32%, P = 5 × 10−6), AMR (2.50%, IQR: 2.06%–3.79%, P = 2 × 10−5), INFXN (0.74%, IQR: 0.46%–1.38%, P = 0.02), and CLAD/NRAD (1.60%, IQR: 0.57%–2.60%, P = 1.4 × 10−4) versus the stable cohort. Area under the receiver operator characteristic curve for AR versus stable was 0.91 (95% confidence interval [CI]: 0.83-0.98). Using a ≥1% dd-cfDNA fraction threshold, sensitivity for AR was 89.1% (95% CI: 76.2%-100.0%), specificity 82.9% (95% CI: 73.3%-92.4%), positive predictive value, 51.9% (95% CI: 37.5%-66.3%), and negative predictive value, 97.3% (95% CI: 94.3%-100%). For combined allograft injury area under the receiver operator characteristic curve was 0.76 (95% CI: 0.66-0.85), sensitivity 59.9% (95% CI: 46.0%-73.9%), specificity 83.9% (95% CI: 74.1%-93.7%), positive predictive value, 43.6% (95% CI: 27.6%-59.6%), and negative predictive value, 91.0% (95% CI: 87.9%-94.0%). Conclusions. These results indicate that our dd-cfDNA assay detects AR and other allograft injury. dd-cfDNA monitoring, accompanied by standard clinical assessments, represents a valuable precision tool to support lung transplant health and is appropriate for further assessment in a prospective randomized-controlled study.

Abstract Background There are multiple obstacles encountered by immigrants attempting to engage hepatitis C virus (HCV) care and treatment. We evaluated the diversity and treatment outcomes of HCV-infected immigrants evaluated for Direct Acting Antiviral (DAA) therapy in Canada. Methods The Canadian Network Undertaking against Hepatitis C (CANUHC) Cohort contains demographic information and DAA treatment information prospectively collected at 10 Canadian sites. Information on country of origin and race are collected. Characteristics and outcomes (sustained virological response; SVR) were compared by immigration status and race. Results Between January 2016 and May 2018, 725 HCV-infected patients assessed for DAA therapy were enrolled in CANUHC (mean age: 52.66 ± 12.68 years); 65.66% male; 82.08% White, 5.28% Indigenous, 4.64% South East Asian, 4.64% East Indian, 3.36% Black). 18.48% were born outside of Canada. Mean age was similar [immigrants: 54.36 ± 13.95 years), Canadian-born: 52.27 ± 12.35 years); (p = 0.085)]. The overall baseline fibrosis score (in kPa measured by transient elastography) was similar among Canadian and foreign-born patients. Fibrosis score was not predicted by race or genotype. The proportion initiating DAA therapy was similar by immigrant status (56.72% vs 49.92%). SVR rates by intent-to-treat analysis were similar (immigrants-89.47%, Canadian-born-92.52%; p = 0.575). Conclusion A diverse immigrant population is engaging care in Canada, initiating HCV antiviral therapy in an equitable fashion and achieving SVR proportions similar to Canada-born patients. Our Canadian experience may be of value in informing HCV elimination efforts in economically developed regions.

As part of a combined antiretroviral regimen, doravirine is safe and effective at suppressing viral replication in both treatment-naive and treatment-experienced adults living with human immunodeficiency virus (HIV)-1 who have no history of drug resistance against doravirine. In virologically suppressed individuals switching to a combination of doravirine, lamivudine, and tenofovir disoproxil fumarate, no resistance was found after 48 weeks. In treatment-naive individuals, rare cases (

Abstract Background An individual is considered HIV positive when a confirmatory HIV-1/HIV-2 differentiation test returns positive following an initial reactive antigen/antibody combination screen. Falsely reactive HIV screens have been reported in patients with various concomitant infectious and autoimmune conditions. Falsely positive confirmatory HIV differentiation assays are seen less frequently, but have been observed in cases of pregnancy, pulmonary embolism, and malaria. Case presentation A healthy 27 year-old man was referred after a reactive ADVIA Centaur® HIV Ag/Ab screen and positive Bio-Rad Geenius™ HIV 1/2 Confirmatory assay, suggesting HIV-1 infection. The patient’s HIV viral load was undetectable prior to initiation of antiretroviral therapy, and remained undetectable on subsequent testing after initiation of antiretroviral therapy. Both Centaur® and Geenius™ tests were repeated and returned reactive. As this patient was believed to be at low risk of acquiring HIV infection, samples were additionally run on Genscreen™ HIV-1 Ag assay and Fujirebio Inno-LIA™ HIV-1/2 score, with both returning non-reactive. For confirmation, the patient’s proviral HIV DNA testing was negative, confirming the initial results as being falsely positive. The patient disclosed that he had been using a variety of anabolic steroids before and during the time of HIV testing. Discussion and conclusions The erroneous diagnosis of HIV can result in decreased quality of life and adverse effects of antiretroviral therapy if initiated, hence the importance of interpreting the results of HIV testing in the context of an individual patient. This reports suggests a potential association between the use of anabolic steroids and falsely-reactive HIV testing.

Advance knowledge of soil water content (SWC) in the soil wetting layer of crop irrigation can help develop more reasonable irrigation plans and improve the efficiency of agricultural irrigation water use. To improve the accuracy of predicting SWC at multiple depths, the ResBiLSTM model was proposed, in which continuous meteorological and SWC data were gridded and transformed as model inputs, and then high-dimensional spatial and time series features were extracted by ResNet and BiLSTM, respectively, and integrated by a meta-learner. Meteorological, SWC and growth stage records data from seven typical maize monitoring stations in Hebei Province, China, during the 2016-2018 summer maize planting process were utilized for the training, evaluation and testing of the ResBiLSTM model, with model prediction targets set at 20cm, 30cm, 40cm and 50cm depths. Experimental results showed that: 1) ResBiLSTM model could achieve better model fit and prediction of meteorological and SWC data at all growth stages, with R2 within [0.818, 0.991], average MAE within [0.79%, 2.00%], and the overall prediction accuracy ranked as follows: anthesis maturity stage > seedling stage > tassel stage; 2) The average MSE of the ResBiLSTM model for the prediction of SWC in the next 1-6 days was within [3.91%, 15.82%], and the prediction accuracy decreased with the extension of the prediction time; 3) Compared with the classical machine learning model and related deep learning models, the ResBiLSTM model was able to obtain better prediction accuracy performance.

The COVID-19 pandemic continues to rage on, with multiple waves causing substantial harm to health and economies around the world. Motivated by the use of computed tomography (CT) imaging at clinical institutes around the world as an effective complementary screening method to RT-PCR testing, we introduced COVID-Net CT, a deep neural network tailored for detection of COVID-19 cases from chest CT images, along with a large curated benchmark dataset comprising 1,489 patient cases as part of the open-source COVID-Net initiative. However, one potential limiting factor is restricted data quantity and diversity given the single nation patient cohort used in the study. To address this limitation, in this study we introduce enhanced deep neural networks for COVID-19 detection from chest CT images which are trained using a large, diverse, multinational patient cohort. We accomplish this through the introduction of two new CT benchmark datasets, the largest of which comprises a multinational cohort of 4,501 patients from at least 16 countries. To the best of our knowledge, this represents the largest, most diverse multinational cohort for COVID-19 CT images in open-access form. Additionally, we introduce a novel lightweight neural network architecture called COVID-Net CT S, which is significantly smaller and faster than the previously introduced COVID-Net CT architecture. We leverage explainability to investigate the decision-making behavior of the trained models and ensure that decisions are based on relevant indicators, with the results for select cases reviewed and reported on by two board-certified radiologists with over 10 and 30 years of experience, respectively. The best-performing deep neural network in this study achieved accuracy, COVID-19 sensitivity, positive predictive value, specificity, and negative predictive value of 99.0%/99.1%/98.0%/99.4%/99.7%, respectively. Moreover, explainability-driven performance validation shows consistency with radiologist interpretation by leveraging correct, clinically relevant critical factors. The results are promising and suggest the strong potential of deep neural networks as an effective tool for computer-aided COVID-19 assessment. While not a production-ready solution, we hope the open-source, open-access release of COVID-Net CT-2 and the associated benchmark datasets will continue to enable researchers, clinicians, and citizen data scientists alike to build upon them.

Robotic leg prostheses and exoskeletons can provide powered locomotor assistance to older adults and/or persons with physical disabilities. However, the current locomotion mode recognition systems being developed for automated high-level control and decision-making rely on mechanical, inertial, and/or neuromuscular sensors, which inherently have limited prediction horizons (i.e., analogous to walking blindfolded). Inspired by the human vision-locomotor control system, we developed an environment classification system powered by computer vision and deep learning to predict the oncoming walking environments prior to physical interaction, therein allowing for more accurate and robust high-level control decisions. In this study, we first reviewed the development of our “ExoNet” database—the largest and most diverse open-source dataset of wearable camera images of indoor and outdoor real-world walking environments, which were annotated using a hierarchical labeling architecture. We then trained and tested over a dozen state-of-the-art deep convolutional neural networks (CNNs) on the ExoNet database for image classification and automatic feature engineering, including: EfficientNetB0, InceptionV3, MobileNet, MobileNetV2, VGG16, VGG19, Xception, ResNet50, ResNet101, ResNet152, DenseNet121, DenseNet169, and DenseNet201. Finally, we quantitatively compared the benchmarked CNN architectures and their environment classification predictions using an operational metric called “NetScore,” which balances the image classification accuracy with the computational and memory storage requirements (i.e., important for onboard real-time inference with mobile computing devices). Our comparative analyses showed that the EfficientNetB0 network achieves the highest test accuracy; VGG16 the fastest inference time; and MobileNetV2 the best NetScore, which can inform the optimal architecture design or selection depending on the desired performance. Overall, this study provides a large-scale benchmark and reference for next-generation environment classification systems for robotic leg prostheses and exoskeletons.

Medical image analysis continues to hold interesting challenges given the subtle characteristics of certain diseases and the significant overlap in appearance between diseases. In this study, we explore the concept of self-attention for tackling such subtleties in and between diseases. To this end, we introduce, a multi-scale encoder-decoder self-attention (MEDUSA) mechanism tailored for medical image analysis. While self-attention deep convolutional neural network architectures in existing literature center around the notion of multiple isolated lightweight attention mechanisms with limited individual capacities being incorporated at different points in the network architecture, MEDUSA takes a significant departure from this notion by possessing a single, unified self-attention mechanism with significantly higher capacity with multiple attention heads feeding into different scales in the network architecture. To the best of the authors' knowledge, this is the first “single body, multi-scale heads” realization of self-attention and enables explicit global context among selective attention at different levels of representational abstractions while still enabling differing local attention context at individual levels of abstractions. With MEDUSA, we obtain state-of-the-art performance on multiple challenging medical image analysis benchmarks including COVIDx, Radiological Society of North America (RSNA) RICORD, and RSNA Pneumonia Challenge when compared to previous work. Our MEDUSA model is publicly available.

As the Coronavirus Disease 2019 (COVID-19) continues to impact many aspects of life and the global healthcare systems, the adoption of rapid and effective screening methods to prevent the further spread of the virus and lessen the burden on healthcare providers is a necessity. As a cheap and widely accessible medical image modality, point-of-care ultrasound (POCUS) imaging allows radiologists to identify symptoms and assess severity through visual inspection of the chest ultrasound images. Combined with the recent advancements in computer science, applications of deep learning techniques in medical image analysis have shown promising results, demonstrating that artificial intelligence-based solutions can accelerate the diagnosis of COVID-19 and lower the burden on healthcare professionals. However, the lack of large, well annotated datasets poses a challenge in developing effective deep neural networks, especially in the case of rare diseases and new pandemics. To address this issue, we present COVID-Net USPro, an explainable few-shot deep prototypical network that is designed to detect COVID-19 cases from very few ultrasound images. Through intensive quantitative and qualitative assessments, the network not only demonstrates high performance in identifying COVID-19 positive cases, using an explainability component, but it is also shown that the network makes decisions based on the actual representative patterns of the disease. Specifically, COVID-Net USPro achieves 99.55% overall accuracy, 99.93% recall, and 99.83% precision for COVID-19-positive cases when trained with only five shots. In addition to the quantitative performance assessment, our contributing clinician with extensive experience in POCUS interpretation verified the analytic pipeline and results, ensuring that the network’s decisions are based on clinically relevant image patterns integral to COVID-19 diagnosis. We believe that network explainability and clinical validation are integral components for the successful adoption of deep learning in the medical field. As part of the COVID-Net initiative, and to promote reproducibility and foster further innovation, the network is open-sourced and available to the public.

Pulmonary fibrosis is a devastating chronic lung disease that causes irreparable lung tissue scarring and damage, resulting in progressive loss in lung capacity and has no known cure. A critical step in the treatment and management of pulmonary fibrosis is the assessment of lung function decline, with computed tomography (CT) imaging being a particularly effective method for determining the extent of lung damage caused by pulmonary fibrosis. Motivated by this, we introduce Fibrosis-Net, a deep convolutional neural network design tailored for the prediction of pulmonary fibrosis progression from chest CT images. More specifically, machine-driven design exploration was leveraged to determine a strong architectural design for CT lung analysis, upon which we build a customized network design tailored for predicting forced vital capacity (FVC) based on a patient’s CT scan, initial spirometry measurement, and clinical metadata. Finally, we leverage an explainability-driven performance validation strategy to study the decision-making behavior of Fibrosis-Net as to verify that predictions are based on relevant visual indicators in CT images. Experiments using a patient cohort from the OSIC Pulmonary Fibrosis Progression Challenge showed that the proposed Fibrosis-Net is able to achieve a significantly higher modified Laplace Log Likelihood score than the winning solutions on the challenge. Furthermore, explainability-driven performance validation demonstrated that the proposed Fibrosis-Net exhibits correct decision-making behavior by leveraging clinically-relevant visual indicators in CT images when making predictions on pulmonary fibrosis progress. Fibrosis-Net is able to achieve a significantly higher modified Laplace Log Likelihood score than the winning solutions on the OSIC Pulmonary Fibrosis Progression Challenge, and has been shown to exhibit correct decision-making behavior when making predictions. Fibrosis-Net is available to the general public in an open-source and open access manner as part of the OpenMedAI initiative. While Fibrosis-Net is not yet a production-ready clinical assessment solution, we hope that its release will encourage researchers, clinicians, and citizen data scientists alike to leverage and build upon it.

This paper presents a novel hallway gait extraction system that enables an individual’s spatiotemporal gait parameter extraction at each gait cycle using a single FMCW (Frequency Modulated Continuous Wave) radar. The purpose of the proposed system is to detect changes in gait that may be the signs of changes in mobility, cognition, and frailty, particularly for older adults in retirement homes. We believe that one of the straightforward applications for gait monitoring using radars is in corridors and hallways, which are commonly available in most retirement and long-term care homes. To achieve in-corridor coverage, we designed an in-package hyperbola-based lens antenna integrated with a radar module package empowered by our fast and easy-to-implement gait extraction method. We validated system functionality by capturing spatiotemporal gait values (e.g., speed, step points, step time, step length, and step count) of people walking in a hallway. The results achieved in this work pave the way to explore the use of stand-alone radar-based sensors in long hallways in retirement apartment buildings or individual’s homes for use in day-to-day long-term monitoring of gait parameters of older adults.

Background: Hepatitis B surface antigen (HBsAg) loss is associated with improved clinical outcomes for individuals with chronic hepatitis B (CHB); however, the effects of varying HBsAg levels on clinical outcomes in diverse cohorts are understudied. Methods: In this cross-sectional, multicentre, retrospective study, the data on adult subjects enrolled in the Canadian HBV Network with CHB seen from 1 January 2012 to 30 January 2021 with the treatment and virologic data within 1 year of HBsAg testing were analyzed. Patients were tested for HBsAg using qualitative (for HBsAg-negative samples) and/or commercial quantitative assays. Fibrosis or hepatic necroinflammation was determined by the liver stiffness measurement (LSM). The baseline data were summarized using descriptive statistics and compared by using univariable/multivariable analyses. Results: This study included 844 CHB patients, with a median age of 49.6 years (IQR 40.1–60.5), and 37% were female. In total, 751 patients (78.6%) had known ethnicity data, and 76.7% self-reported as Asian, 11.4% as Black, 6.8% as White, and 4.8% as other. Among the 844 patients, 237 (28.0%) were HBsAg (−) (1000 IU/mL. Overall, 80% (682) had known HBeAg status at the last follow-up, and the majority (87.0%) were HBeAg-negative. In addition, 54% (461/844) had prior antiviral therapy, 19.7% of which (16.3, 23.7, n = 91) were HBsAg (−). The treated patients had a lower risk of cirrhosis (16.46, 95% CI 1.89–143.39, p = 0.01) or HCC (8.23, 95% CI 1.01–67.39, p = 0.05) than the untreated patients. A lower proportion of the HBsAg-loss group had cirrhosis (5.7% vs. 10.9%, p = 0.021) and HCC (0.9% vs. 6.2%, p = 0.001). Conclusion: In this retrospective, ethnically diverse cohort study, CHB patients who received antiviral therapy and/or had HBsAg loss were less likely to develop cirrhosis and HCC, confirming the results of the studies in less diverse cohorts. No association was found between the qHBsAg level and fibrosis determined with LSM. Individuals who achieved HBsAg loss had low-level qHBsAg within 1 year of seroclearance.

The tremendous potential exhibited by deep learning is often offset by architectural and computational complexity, making widespread deployment a challenge for edge scenarios such as mobile and other consumer devices. To tackle this challenge, we explore the following idea: Can we learn generative machines to automatically generate deep neural networks with efficient network architectures? In this study, we introduce the idea of generative synthesis, which is premised on the intricate interplay between a generator‐inquisitor pair that work in tandem to garner insights and learn to generate highly efficient deep neural networks that best satisfies operational requirements. Experimental results for image classification, semantic segmentation, and object detection tasks illustrate the efficacy of generative synthesis (GenSynth) in producing generators that automatically generate highly efficient deep neural networks (which we nickname FermiNets with higher model efficiency and lower computational costs (reaching >10× more efficient and fewer multiply‐accumulate operations than several tested state‐of‐the‐art networks), as well as higher energy efficiency (reaching >4× improvements in image inferences per joule consumed on a Nvidia Tegra X2 mobile processor). As such, GenSynth can be a powerful, generalised approach for accelerating and improving the building of deep neural networks for on‐device edge scenarios.

Radiomics-driven computer aided diagnosis (CAD) has shown considerable promise in recent years as a potential tool for improving clinical decision support in medical oncology, particularly those based around the concept of discovery radiomics, where radiomic sequencers are discovered through the analysis of medical imaging data. One of the main limitations, with current CAD approaches, is that it is very difficult to gain insight or rationale as to how decisions are made, thus limiting their utility to clinicians. In this paper, we propose CLEAR-DR, a novel interpretable CAD system based on the notion of CLass-Enhanced Attentive Response Discovery Radiomics for the purpose of clinical decision support for diabetic retinopathy. In addition to disease grading via the discovered deep radiomic sequencer, the CLEAR-DR system also produces a visual interpretation of the decision-making process to provide better insight and understanding of the decision-making process of the system. We demonstrate the effectiveness and utility of the proposed CLEAR-DR system of enhancing the interpretability of diagnostic grading results for the application of diabetic retinopathy grading. CLEAR-DR can act as a potentially powerful tool to address the uninterpretability issue of current CAD systems, thus improving their utility to clinicians.

Harmful algae blooms are a growing global concern since they negatively affect the quality of drinking water. The gold-standard process to identify and enumerate algae requires highly trained professionals to manually observe algae under a microscope. Therefore, an automated approach to identify and enumerate these micro-organisms is needed. This research investigates the feasibility of leveraging machine learning and fluorescence-based spectral-morphological features to enable the identification of six different algae types in an automated fashion. A custom multi-band fluorescence imaging microscope is used to capture fluorescence data of water samples at six different excitation wavelengths ranging from 405 to 530 nm. Automated data processing and segmentation were performed on the captured data to isolate different micro-organisms from the water sample. Different morphological and spectral fluorescence features are then extracted from the isolated micro-organism imaging data and is used to train neural network classification models. The experimental results using three different neural network classification models (one trained on morphological features, one trained on fluorescence-based spectral features, and one trained on fluorescence-based spectral-morphological features) showed that the use of either fluorescence-based spectral features or fluorescence-based spectral-morphological features to train neural network classification models led to statistically significant improvements in identification accuracy when compared to the use of morphological features (with average identification accuracies of 95.7% ± 3.5% and 96.1% ± 1.5%, respectively). These preliminary results are promising and illustrate the feasibility of leveraging machine learning and fluorescence-based spectral-morphological features as a viable method for automated identification of different algae types.

Lung cancer is the leading cause of cancer-related death worldwide. Computer-aided diagnosis (CAD) systems have shown significant promise in recent years for facilitating the effective detection and classification of abnormal lung nodules in computed tomography (CT) scans. While hand-engineered radiomic features have been traditionally used for lung cancer prediction, there have been significant recent successes achieving state-of-the-art results in the area of discovery radiomics. Here, radiomic sequencers comprising of highly discriminative radiomic features are discovered directly from archival medical data. However, the interpretation of predictions made using such radiomic sequencers remains a challenge. A novel end-to-end interpretable discovery radiomics-driven lung cancer prediction pipeline has been designed, build, and tested. The radiomic sequencer being discovered possesses a deep architecture comprised of stacked interpretable sequencing cells (SISC). The SISC architecture is shown to outperform previous approaches while providing more insight in to its decision making process. The SISC radiomic sequencer is able to achieve state-of-the-art results in lung cancer prediction, and also offers prediction interpretability in the form of critical response maps. The critical response maps are useful for not only validating the predictions of the proposed SISC radiomic sequencer, but also provide improved radiologist-machine collaboration for effective diagnosis.

While recent advances in deep learning have led to significant improvements in facial expression classification (FEC), a major challenge that remains a bottleneck for the widespread deployment of such systems is their high architectural and computational complexities. This is especially challenging given the operational requirements of various FEC applications, such as safety, marketing, learning, and assistive living, where real-time requirements on low-cost embedded devices is desired. Motivated by this need for a compact, low latency, yet accurate system capable of performing FEC in real-time on low-cost embedded devices, this study proposes EmotionNet Nano, an efficient deep convolutional neural network created through a human-machine collaborative design strategy, where human experience is combined with machine meticulousness and speed in order to craft a deep neural network design catered toward real-time embedded usage. To the best of the author’s knowledge, this is the very first deep neural network architecture for facial expression recognition leveraging machine-driven design exploration in its design process, and exhibits unique architectural characteristics such as high architectural heterogeneity and selective long-range connectivity not seen in previous FEC network architectures. Two different variants of EmotionNet Nano are presented, each with a different trade-off between architectural and computational complexity and accuracy. Experimental results using the CK + facial expression benchmark dataset demonstrate that the proposed EmotionNet Nano networks achieved accuracy comparable to state-of-the-art FEC networks, while requiring significantly fewer parameters. Furthermore, we demonstrate that the proposed EmotionNet Nano networks achieved real-time inference speeds (e.g., >25 FPS and >70 FPS at 15 and 30 W, respectively) and high energy efficiency (e.g., >1.7 images/sec/watt at 15 W) on an ARM embedded processor, thus further illustrating the efficacy of EmotionNet Nano for deployment on embedded devices.

BackgroundHIV-HCV coinfected individuals are often more deprived than the general population. However, deprivation is difficult to measure, often relying on aggregate data which does not capture individual heterogeneity. We developed an individual-level deprivation index for HIV-HCV co-infected persons that encapsulated social, material, and lifestyle factors.MethodsWe estimated an individual-level deprivation index with data from the Canadian Coinfection Cohort, a national prospective cohort study. We used a predetermined process to select 9 out of 19 dichotomous variables at baseline visit to include in the deprivation model: income >$1500/month; education >high school; employment; identifying as gay or bisexual; Indigenous status; injection drug use in last 6 months; injection drug use ever; past incarceration, and past psychiatric hospitalization. We fitted an item response theory model with: severity parameters (how likely an item was reported), discriminatory parameters, (how well a variable distinguished index levels), and an individual parameter (the index). We considered two models: a simple one with no provincial variation and a hierarchical model by province. The Widely Applicable Information Criterion (WAIC) was used to compare the fitted models. To showcase a potential utility of the proposed index, we evaluated with logistic regression the association of the index with non-attendance to a second clinic visit (as a proxy for disengagement) and using WAIC compared it to a model containing all the individual parameters that compose the index as covariates.ResultsWe analyzed 1547 complete cases of 1842 enrolled participants. According to the WAIC the hierarchical model provided a better fit when compared to the model that does not consider the individual's province. Values of the index were similarly distributed across the provinces. Overall, past incarceration, education, and unemployment had the highest discriminatory parameters. However, in each province different components of the index were associated with being deprived reflecting local epidemiology. For example, Saskatchewan had the highest severity parameter for Indigenous status while Quebec the lowest. For the secondary analysis, 457 (30%) failed to attend a second visit. A one-unit increase in the index was associated with 17% increased odds (95% credible interval, 2% to 34%) of not attending a second visit. The model with just the index performed better than the model with all the components as covariates in terms of WAIC.ConclusionWe estimated an individual-level deprivation index in the Canadian Coinfection cohort. The index identified deprivation profiles across different provinces. This index and the methodology used may be useful in studying health and treatment outcomes that are influenced by social disparities in co-infected Canadians. The methodological approach described can be used in other studies with similar characteristics.

Background: The Coronavirus Disease 2019 (COVID-19) pandemic continues to have a devastating effect on the health and well-being of the global population. Apart from the global health crises, the pandemic has also caused significant economic and financial difficulties and socio-physiological implications. Effective screening, triage, treatment planning, and prognostication of outcome play a key role in controlling the pandemic. Recent studies have highlighted the role of point-of-care ultrasound imaging for COVID-19 screening and prognosis, particularly given that it is non-invasive, globally available, and easy-to-sanitize. COVIDx-US Dataset: Motivated by these attributes and the promise of artificial intelligence tools to aid clinicians, we introduce COVIDx-US, an open-access benchmark dataset of COVID-19 related ultrasound imaging data. The COVIDx-US dataset was curated from multiple data sources and its current version, i.e., v1.5., consists of 173 ultrasound videos and 21,570 processed images across 147 patients with COVID-19 infection, non-COVID-19 infection, other lung diseases/conditions, as well as normal control cases. Conclusions: The COVIDx-US dataset was released as part of a large open-source initiative, the COVID-Net initiative, and will be continuously growing, as more data sources become available. To the best of the authors’ knowledge, COVIDx-US is the first and largest open-access fully-curated benchmark lung ultrasound imaging dataset that contains a standardized and unified lung ultrasound score per video file, providing better interpretation while enabling other research avenues such as severity assessment. In addition, the dataset is reproducible, easy-to-use, and easy-to-scale thanks to the well-documented modular design.

Mapping soil heavy metal concentration using machine learning models based on readily available satellite remote sensing images is highly desirable. Accurate mapping relies on appropriate data, feature extraction, and model selection. To this end, a data processing pipeline for soil copper (Cu) concentration estimation has been designed. First, instead of using single Landsat scenes, the utilization of multiple Landsat scenes of the same location over time is considered. Second, to generate a preferred feature set as input to a regression model, a number of feature extraction methods are motivated and compared. Third, to find a preferred regression model, a variety of approaches are implemented and compared for accuracy. In this research, 11 Landsat-8 images from 2013 to 2017 of Gulin County, Sichuan China, and 138 soil samples with lab-measured Cu concentrations collected from the area in 2015 are used. A variety a metrics under cross-validation are used for comparison. The results indicate that multi-temporal images increase accuracy compared to single Landsat images. The preferred feature extraction varies based on the regression model used; however, the best results are obtained using support vector regression and the original data. The final soil Cu map generated using the recommended data processing pipeline shows a consistent spatial pattern with a ground-truth land cover classification map. These results indicate that machine learning has the ability to perform large-scale soil heavy metal mapping from widely available satellite remote sensing images.

The coronavirus disease 2019 (COVID-19) pandemic continues to have a tremendous impact on patients and healthcare systems around the world. In the fight against this novel disease, there is a pressing need for rapid and effective screening tools to identify patients infected with COVID-19, and to this end CT imaging has been proposed as one of the key screening methods which may be used as a complement to RT-PCR testing, particularly in situations where patients undergo routine CT scans for non-COVID-19 related reasons, patients have worsening respiratory status or developing complications that require expedited care, or patients are suspected to be COVID-19-positive but have negative RT-PCR test results. Early studies on CT-based screening have reported abnormalities in chest CT images which are characteristic of COVID-19 infection, but these abnormalities may be difficult to distinguish from abnormalities caused by other lung conditions. Motivated by this, in this study we introduce COVIDNet-CT, a deep convolutional neural network architecture that is tailored for detection of COVID-19 cases from chest CT images via a machine-driven design exploration approach. Additionally, we introduce COVIDx-CT, a benchmark CT image dataset derived from CT imaging data collected by the China National Center for Bioinformation comprising 104,009 images across 1,489 patient cases. Furthermore, in the interest of reliability and transparency, we leverage an explainability-driven performance validation strategy to investigate the decision-making behavior of COVIDNet-CT, and in doing so ensure that COVIDNet-CT makes predictions based on relevant indicators in CT images. Both COVIDNet-CT and the COVIDx-CT dataset are available to the general public in an open-source and open access manner as part of the COVID-Net initiative. While COVIDNet-CT is not yet a production-ready screening solution, we hope that releasing the model and dataset will encourage researchers, clinicians, and citizen data scientists alike to leverage and build upon them.