Separating signal from noise in wastewater data: An algorithm to identify community-level COVID-19 surges in real time.

Wastewater monitoring has provided health officials with early warnings for new COVID-19 outbreaks, but to date, no approach has been validated to distinguish signal (sustained surges) from noise (background variability) in wastewater data to alert officials to the need for heightened public health response. We analyzed 62 wk of data from 19 sites participating in the North Carolina Wastewater Monitoring Network to characterize wastewater metrics around the Delta and Omicron surges. We found that wastewater data identified outbreaks 4 to 5 d before case data (reported on the earlier of the symptom start date or test collection date), on average. At most sites, correlations between wastewater and case data were similar regardless of how wastewater concentrations were normalized and whether calculated with county-level or sewershed-level cases, suggesting that officials may not need to geospatially align case data with sewershed boundaries to gain insights into disease transmission. Although wastewater trend lines captured clear differences in the Delta versus Omicron surge trajectories, no single wastewater metric (detectability, percent change, or flow-population normalized viral concentrations) reliably signaled when these surges started. After iteratively examining different combinations of these three metrics, we developed the Covid-SURGE (Signaling Unprecedented Rises in Groupwide Exposure) algorithm, which identifies unprecedented signals in the wastewater data. With a true positive rate of 82%, a false positive rate of 7%, and strong performance during both surges and in small and large sites, our algorithm provides public health officials with an automated way to flag community-level COVID-19 surges in real time.

Deep learning based on carotid transverse B-mode scan videos for the diagnosis of carotid plaque: a prospective multicenter study.

OBJECTIVES: Accurate detection of carotid plaque using ultrasound (US) is essential for preventing stroke. However, the diagnostic performance of junior radiologists (with approximately 1 year of experience in carotid US evaluation) is relatively poor. We thus aim to develop a deep learning (DL) model based on US videos to improve junior radiologists' performance in plaque detection. METHODS: This multicenter prospective study was conducted at five hospitals. CaroNet-Dynamic automatically detected carotid plaque from carotid transverse US videos allowing clinical detection. Model performance was evaluated using expert annotations (with more than 10 years of experience in carotid US evaluation) as the ground truth. Model robustness was investigated on different plaque characteristics and US scanning systems. Furthermore, its clinical applicability was evaluated by comparing the junior radiologists' diagnoses with and without DL-model assistance. RESULTS: A total of 1647 videos from 825 patients were evaluated. The DL model yielded high performance with sensitivities of 87.03% and 94.17%, specificities of 82.07% and 74.04%, and areas under the receiver operating characteristic curve of 0.845 and 0.841 on the internal and multicenter external test sets, respectively. Moreover, no significant difference in performance was noted among different plaque characteristics and scanning systems. Using the DL model, the performance of the junior radiologists improved significantly, especially in terms of sensitivity (largest increase from 46.3 to 94.44%). CONCLUSIONS: The DL model based on US videos corresponding to real examinations showed robust performance for plaque detection and significantly improved the diagnostic performance of junior radiologists. KEY POINTS: • The deep learning model based on US videos conforming to real examinations showed robust performance for plaque detection. • Computer-aided diagnosis can significantly improve the diagnostic performance of junior radiologists in clinical practice.

Isolating the AFFF Signature in Coastal Watersheds Using Oxidizable PFAS Precursors and Unexplained Organofluorine.

Water supplies for millions of U.S. individuals exceed maximum contaminant levels for per- and polyfluoroalkyl substances (PFAS). Contemporary and legacy use of aqueous film forming foams (AFFF) is a major contamination source. However, diverse PFAS sources are present within watersheds, making it difficult to isolate their predominant origins. Here we examine PFAS source signatures among six adjacent coastal watersheds on Cape Cod, MA, U.S.A. using multivariate clustering techniques. A distinct signature of AFFF contamination enriched in precursors with six perfluorinated carbons (C6) was identified in watersheds with an AFFF source, while others were enriched in C4 precursors. Principal component analysis of PFAS composition in impacted watersheds showed a decline in precursor composition relative to AFFF stocks and a corresponding increase in terminal perfluoroalkyl sulfonates with < C6 but not those with ≥ C6. Prior work shows that in AFFF stocks, all extractable organofluorine (EOF) can be explained by targeted PFAS and precursors inferred using Bayesian inference on the total oxidizable precursor assay. Using the same techniques for the first time in impacted watersheds, we find that only 24%-63% of the EOF can be explained by targeted PFAS and oxidizable precursors. Our work thus indicates the presence of large non-AFFF organofluorine sources in these coastal watersheds.

Lower Intent to Comply with COVID-19 Public Health Recommendations Correlates to Higher Disease Burden in Following 30 Days.

OBJECTIVES: We sought to determine whether self-reported intent to comply with public health recommendations correlates with future coronavirus disease 2019 (COVID-19) disease burden. METHODS: A cross-sectional, online survey of US adults, recruited by snowball sampling, from April 9 to July 12, 2020. Primary measurements were participant survey responses about their intent to comply with public health recommendations. Each participant's intent to comply was compared with his or her local COVID-19 case trajectory, measured as the 7-day rolling median percentage change in COVID-19 confirmed cases within participants' 3-digit ZIP code area, using public county-level data, 30 days after participants completed the survey. RESULTS: After applying raking techniques, the 10,650-participant sample was representative of US adults with respect to age, sex, race, and ethnicity. Intent to comply varied significantly by state and sex. Lower reported intent to comply was associated with higher COVID-19 case increases during the following 30 days. For every 3% increase in intent to comply with public health recommendations, which could be achieved by improving average compliance by a single point for a single item, we estimate a 9% reduction in new COVID-19 cases during the subsequent 30 days. CONCLUSIONS: Self-reported intent to comply with public health recommendations may be used to predict COVID-19 disease burden. Measuring compliance intention offers an inexpensive, readily available method of predicting disease burden that can also identify populations most in need of public health education aimed at behavior change.

Selenium ameliorates cadmium-induced mouse leydig TM3 cell apoptosis via inhibiting the ROS/JNK /c-jun signaling pathway.

Despite the well-known acknowledgement of both the toxicity of cadmium (Cd) and the ameliorative effect of selenium (Se), the mechanism of the protective effect of selenium on cadmium-induced Mouse Leydig (TM3) cell apoptosis remains unknown. In this study, we hypothesized that the reactive oxygen species (ROS)-mediated c-jun N-terminal kinase (JNK) signaling pathway is involved in anti-apoptosis of selenium against cadmium in TM3 cells. We found that exposure to cadmium caused evident cytotoxicity, in which cell viability was inhibited, followed by inducement of apoptosis. Moreover, the level of ROS generation was elevated, leading to the phosphorylation of JNK. In addition, following cadmium exposure, the nuclear transcription factor c-jun was significantly activated, which led to increased expression of downstream gene c-jun, resulting in downstream activation of the apoptosis-related protein Caspase3 and upregulation of Cleaved-PARP, as well as inhibition of the anti-apoptosis protein Bcl-2. However, pretreatment with selenium remarkably suppressed cadmium-induced TM3 cell apoptosis. Furthermore, the level of ROS declined, and the JNK signaling pathway was blocked. Following this, the gene expression of c-jun decreased while Bcl-2 increased, which was consistent with the effects on proteins, that Caspase3 activity and Cleaved-PARP were inhibited while Bcl-2 level was restored. In order to explain the relationship between molecules of the signaling pathway, N-acetyl-L-cysteine (NAC), the ROS inhibitor, and JNK1/2 siRNA were administered, which further indicated the mediatory role of the ROS/JNK/c-jun signaling pathway in regulating anti-apoptosis of selenium against cadmium-induced TM3 cell apoptosis.

Cadmium-induced apoptosis through reactive oxygen species-mediated mitochondrial oxidative stress and the JNK signaling pathway in TM3 cells, a model of mouse Leydig cells.

Cadmium (Cd) is a heavy metal that widely exists in the environment and industry, and which causes serious damages to reproductive system. Recent studies have reported that cadmium induces apoptosis of various germ cells in testes, resulting in male infertility. However, the exact mechanism of cadmium-induced apoptosis remains unclear. In this study, we hypothesized that reactive oxygen species (ROS)-mediated c-jun N-terminal kinase (JNK) signaling pathway was involved in cadmium-induced apoptosis in TM3 cells, a model of mouse Leydig cells. TM3 cells were exposed for various times to a range of cadmium concentrations. We found that cadmium reduced TM3 cell viability and increased apoptosis in a time- and dose- dependent manner. Moreover, the levels of ROS generation and the phosphorylation of JNK were elevated by cadmium treatment. In addition, the nuclear transcription factor c-jun was significantly activated, which led to increased expression of downstream c-jun targets and Bcl-2 was decreased, accompanied with downstream activation of apoptosis-related proteins such as Cleaved-Caspase3 and Cleaved-PARP. However, pretreatment with the ROS inhibitor N-acetyl-L-cysteine (NAC) and JNK inhibitor JNK-IN-8, ROS, JNK and cadmium-induced TM3 cell apoptosis were remarkably suppressed. Based on above-mentioned results, this study provides a mechanistic understanding of cadmium induced TM3 cell apoptosis through the ROS/JNK signaling pathways.

Shifting Global Exposures to Poly- and Perfluoroalkyl Substances (PFASs) Evident in Longitudinal Birth Cohorts from a Seafood-Consuming Population.

Rapid declines in legacy poly- and perfluoroalkyl substances (PFASs) have been reported in human populations globally following changes in production since 2000. However, changes in exposure sources are not well understood. Here, we report serum concentrations of 19 PFASs (∑19PFAS) measured in children between 1993 and 2012 from a North Atlantic fishing community (Faroe Islands). Median ∑19PFAS concentrations in children (ages 5-13 years) peaked in 2000 (47.7 ng mL-1) and declined significantly by 14.4% year-1 until 2012. Principal component analysis (PCA) identified two groups of PFASs that likely reflect exposures from diverse consumer products and a third group that consisted of perfluorocarboxylic acids (PFCAs) with nine or more carbons (C ≥ 9). These C ≥ 9 PFASs are strongly associated with mercury in children's hair, a well-established proxy for seafood consumption, especially perfluoroundecanoic acid (PFUnDA, r = 0.72). Toxicokinetic modeling shows PFAS exposures from seafood have become increasingly important (53% of perfluorooctanesulfonate, PFOS, in 2012), despite a decline in whale consumption in recent years. We infer that even in a major seafood-consuming population, declines in legacy PFAS exposure after 2000 were achieved by the rapid phase out of PFOS and its precursors in consumer products. These results emphasize the importance of better understanding exposures to replacement PFASs in these sources.

Can profiles of poly- and Perfluoroalkyl substances (PFASs) in human serum provide information on major exposure sources?

BACKGROUND: Humans are exposed to poly- and perfluoroalkyl substances (PFASs) from diverse sources and this has been associated with negative health impacts. Advances in analytical methods have enabled routine detection of more than 15 PFASs in human sera, allowing better profiling of PFAS exposures. The composition of PFASs in human sera reflects the complexity of exposure sources but source identification can be confounded by differences in toxicokinetics affecting uptake, distribution, and elimination. Common PFASs, such as perfluorooctanoic acid (PFOA), perfluorooctane sulfonic acid (PFOS) and their precursors are ubiquitous in multiple exposure sources. However, their composition varies among sources, which may impact associated adverse health effects. METHODS: We use available PFAS concentrations from several demographic groups in a North Atlantic seafood consuming population (Faroe Islands) to explore whether chemical fingerprints in human sera provide insights into predominant exposure sources. We compare serum PFAS profiles from Faroese individuals to other North American populations to investigate commonalities in potential exposure sources. We compare individuals with similar demographic and physiological characteristics and samples from the same years to reduce confounding by toxicokinetic differences and changing environmental releases. RESULTS: Using principal components analysis (PCA) confirmed by hierarchical clustering, we assess variability in serum PFAS concentrations across three Faroese groups. The first principal component (PC)/cluster consists of C9-C12 perfluoroalkyl carboxylates (PFCAs) and is consistent with measured PFAS profiles in consumed seafood. The second PC/cluster includes perfluorohexanesulfonic acid (PFHxS) and the PFOS precursor N-ethyl perfluorooctane sulfonamidoacetate (N-EtFOSAA), which are directly used or metabolized from fluorochemicals in consumer products such as carpet and food packaging. We find that the same compounds are associated with the same exposure sources in two North American populations, suggesting generalizability of results from the Faroese population. CONCLUSIONS: We conclude that PFAS homologue profiles in serum provide valuable information on major exposure sources. It is essential to compare samples collected at similar time periods and to correct for demographic groups that are highly affected by differences in physiological processes (e.g., pregnancy). Information on PFAS homologue profiles is crucial for attributing adverse health effects to the proper mixtures or individual PFASs.

Temporal Shifts in Poly- and Perfluoroalkyl Substances (PFASs) in North Atlantic Pilot Whales Indicate Large Contribution of Atmospheric Precursors.

Poly- and perfluoroalkyl substances (PFASs) are persistent, bioaccumulative anthropogenic compounds associated with adverse health impacts on humans and wildlife. PFAS production changed in North America and Europe around the year 2000, but impacts on wildlife appear to vary across species and location. Unlike other mammal species, cetaceans lack the enzyme for transforming an important intermediate precursor (perfluorooctane sulfonamide: FOSA), into a prevalent compound in most wildlife (perfluorooctanesulfonate: PFOS). Thus, their tissue burden differentiates these two compounds while other mammals contain PFOS from both direct exposure and precursor degradation. Here we report temporal trends in 15 PFASs measured in muscle from juvenile male North Atlantic pilot whales (Globicephala melas) harvested between 1986 and 2013. FOSA accounted for a peak of 84% of the 15 PFASs around 2000 but declined to 34% in recent years. PFOS and long-chained PFCAs (C9-C13) increased significantly over the whole period (2.8% yr-1 to 8.3% yr-1), but FOSA declined by 13% yr-1 after 2006. Results from FOSA partitioning and bioaccumulation modeling forced by changes in atmospheric inputs reasonably capture magnitudes and temporal patterns in FOSA concentrations measured in pilot whales. Rapid changes in atmospheric FOSA in polar and subpolar regions around 2000 helps to explain large declines in PFOS exposure for species that metabolize FOSA, including seafood consuming human populations. This work reinforces the importance of accounting for biological exposures to PFAS precursors.

Segment anything model for medical images?

The Segment Anything Model (SAM) is the first foundation model for general image segmentation. It has achieved impressive results on various natural image segmentation tasks. However, medical image segmentation (MIS) is more challenging because of the complex modalities, fine anatomical structures, uncertain and complex object boundaries, and wide-range object scales. To fully validate SAM's performance on medical data, we collected and sorted 53 open-source datasets and built a large medical segmentation dataset with 18 modalities, 84 objects, 125 object-modality paired targets, 1050K 2D images, and 6033K masks. We comprehensively analyzed different models and strategies on the so-called COSMOS 1050K dataset. Our findings mainly include the following: (1) SAM showed remarkable performance in some specific objects but was unstable, imperfect, or even totally failed in other situations. (2) SAM with the large ViT-H showed better overall performance than that with the small ViT-B. (3) SAM performed better with manual hints, especially box, than the Everything mode. (4) SAM could help human annotation with high labeling quality and less time. (5) SAM was sensitive to the randomness in the center point and tight box prompts, and may suffer from a serious performance drop. (6) SAM performed better than interactive methods with one or a few points, but will be outpaced as the number of points increases. (7) SAM's performance correlated to different factors, including boundary complexity, intensity differences, etc. (8) Finetuning the SAM on specific medical tasks could improve its average DICE performance by 4.39% and 6.68% for ViT-B and ViT-H, respectively. Codes and models are available at: https://github.com/yuhoo0302/Segment-Anything-Model-for-Medical-Images. We hope that this comprehensive report can help researchers explore the potential of SAM applications in MIS, and guide how to appropriately use and develop SAM.

Hierarchical online contrastive anomaly detection for fetal arrhythmia diagnosis in ultrasound.

Arrhythmia is a major cardiac abnormality in fetuses. Therefore, early diagnosis of arrhythmia is clinically crucial. Pulsed-wave Doppler ultrasound is a commonly used diagnostic tool for fetal arrhythmia. Its key step for diagnosis involves identifying adjacent measurable cardiac cycles (MCCs). As cardiac activity is complex and the experience of sonographers is often varied, automation can improve user-independence and diagnostic-validity. However, arrhythmias pose several challenges for automation because of complex waveform variations, which can cause major localization bias and missed or false detection of MCCs. Filtering out non-MCC anomalies is difficult because of large intra-class and small inter-class variations between MCCs and non-MCCs caused by agnostic morphological waveform variations. Moreover, rare arrhythmia cases are insufficient for classification algorithms to adequately learn discriminative features. Using only normal cases for training, we propose a novel hierarchical online contrastive anomaly detection (HOCAD) framework for arrhythmia diagnosis during test time. The contribution of this study is three-fold. First, we develop a coarse-to-fine framework inspired by hierarchical diagnostic logic, which can refine localization and avoid missed detection of MCCs. Second, we propose an online learning-based contrastive anomaly detection with two new anomaly scores, which can adaptively filter out non-MCC anomalies on a single image during testing. With these complementary efforts, we precisely determine MCCs for correct measurements and diagnosis. Third, to the best of our knowledge, this is the first reported study investigating intelligent diagnosis of fetal arrhythmia on a large-scale and multi-center ultrasound dataset. Extensive experiments on 3850 cases, including 266 cases covering three typical types of arrhythmias, demonstrate the effectiveness of the proposed framework.

The Utility of Machine Learning Models for Predicting Chemical Contaminants in Drinking Water: Promise, Challenges, and Opportunities.

PURPOSE OF REVIEW: This review aims to better understand the utility of machine learning algorithms for predicting spatial patterns of contaminants in the United States (U.S.) drinking water. RECENT FINDINGS: We found 27 U.S. drinking water studies in the past ten years that used machine learning algorithms to predict water quality. Most studies (42%) developed random forest classification models for groundwater. Continuous models show low predictive power, suggesting that larger datasets and additional predictors are needed. Categorical/classification models for arsenic and nitrate that predict exceedances of pollution thresholds are most common in the literature because of good national scale data coverage and priority as environmental health concerns. Most groundwater data used to develop models were obtained from the United States Geological Survey (USGS) National Water Information System (NWIS). Predictors were similar across contaminants but challenges are posed by the lack of a standard methodology for imputation, pre-processing, and differing availability of data across regions. We reviewed 27 articles that focused on seven drinking water contaminants. Good performance metrics were reported for binary models that classified chemical concentrations above a threshold value by finding significant predictors. Classification models are especially useful for assisting in the design of sampling efforts by identifying high-risk areas. Only a few studies have developed continuous models and obtaining good predictive performance for such models is still challenging. Improving continuous models is important for potential future use in epidemiological studies to supplement data gaps in exposure assessments for drinking water contaminants. While significant progress has been made over the past decade, methodological advances are still needed for selecting appropriate model performance metrics and accounting for spatial autocorrelations in data. Finally, improved infrastructure for code and data sharing would spearhead more rapid advances in machine-learning models for drinking water quality.

Sociodemographic Disparities in Mercury Exposure from United States Coal-Fired Power Plants.

Hazardous air pollutants emitted by United States (U.S) coal-fired power plants have been controlled by the Mercury and Air Toxics Standards (MATS) since 2012. Sociodemographic disparities in traditional air pollutant exposures from U.S. power plants are known to occur but have not been evaluated for mercury (Hg), a neurotoxicant that bioaccumulates in food webs. Atmospheric Hg deposition from domestic power plants decreased by 91% across the contiguous U.S. from 6.4 Mg in 2010 to 0.55 Mg in 2020. Prior to MATS, populations living within 5 km of power plants (n = 507) included greater proportions of frequent fish consumers, individuals with low annual income and less than a high school education, and limited English-proficiency households compared to the US general population. These results reinforce a lack of distributional justice in plant siting found in prior work. Significantly greater proportions of low-income individuals lived within 5 km of active facilities in 2020 (n = 277) compared to plants that retired after 2010, suggesting that socioeconomic status may have played a role in retirement. Despite large deposition declines, an end-member scenario for remaining exposures from the largest active power plants for individuals consuming self-caught fish suggests they could still exceed the U.S. Environmental Protection Agency reference dose for methylmercury.

Nano-selenium Alleviates Cadmium-Induced Mouse Leydig Cell Injury, via the Inhibition of Reactive Oxygen Species and the Restoration of Autophagic Flux.

Cadmium (Cd) is a well-known environmental pollutant that can contribute to male reproductive toxicity through oxidative stress. Nano-selenium (Nano-se) is an active single body of selenium with strong antioxidant properties and low toxicity. Some studies have addressed the potential ameliorative effect of Nano-se against Cd-induced testicular toxicity; however, the underlying mechanisms remain to be investigated. This study aimed to explore the protective effect of Nano-se on Cd-induced mouse testicular TM3 cell toxicity by regulating autophagy process. We showed that cadmium exposure to TM3 cells inhibited cell viability and elevated the level of reactive oxygen species (ROS) generation. Morphology observation by transmission electron microscope and the presence of mRFP-GFP-LC3 fluorescence puncta demonstrated that cadmium increased autophagosome formation and accumulation in TM3 cells, resulting in blocking the autophagic flux of TM3 cells. Meanwhile, cadmium remarkably increased the ratio of LC3-II to LC3-I protein expression (2.07 ± 0.31) and the Beclin-1 protein expression (1.97 ± 0.40) in TM3 cells (P < 0.01). Pretreatment with Nano-se significantly reduced Cd-induced TM3 cell toxicity (P < 0.01). Furthermore, Nano-se treatment reversed Cd-induced ROS production and autophagosome accumulation, and autophagy as evidenced by the ratio of LC3-II to LC3-I and Beclin-1 expression. In addition, ROS scavenger, N-acetyl-L-cysteine (NAC) or autophagy inhibitor, 3-methyladenine (3-MA) reversed cadmium-induced ROS generation, autophagosome accumulation, and autophagy-related protein expression levels, which confirmed that cadmium induced TM3 cell injury via ROS signal pathway and blockage of autophagic flux. Collectively, our results reveal that Nano-se attenuates Cd-induced TM3 cell toxicity through the inhibition of ROS production and the amelioration of autophagy disruption.

Locating Multiple Standard Planes in First-Trimester Ultrasound Videos via the Detection and Scoring of Key Anatomical Structures.

OBJECTIVE: This study was aimed at developing a first-trimester standard plane detection (FTSPD) system that can automatically locate nine standard planes in ultrasound videos and investigating its utility in clinical practice. METHODS: The FTSPD system, based on the YOLOv3 network, was developed to detect structures and evaluate the quality of plane images by using a pre-defined scoring system. A total of 220 videos from two different ultrasound scanners were collected to compare detection performance between our FTSPD system and sonographers with different levels of experience. The quality of the detected standard planes was quantitatively rated by an expert according to a scoring protocol. Kolmogorov-Smirnov analysis was used to compare the distributions of scores across all nine standard planes. RESULTS: The expert-rated scores indicated that the quality of the standard planes detected by the FTSPD system was on par with that of the planes detected by senior sonographers. There were no significant differences in the distributions of the scores across all nine standard planes. The FTSPD system performed significantly better than junior sonographers in five standard plane types. CONCLUSION: The results of this study suggest that our FTSPD system has significant potential for detecting standard planes in first-trimester ultrasound screening, which may help to improve the accuracy of fetal ultrasound screening and facilitate early diagnosis of abnormalities. The quality of the standard planes selected by junior sonographers can be significantly improved with the assistance of our FTSPD system.

RecON: Online learning for sensorless freehand 3D ultrasound reconstruction.

Sensorless freehand 3D ultrasound (US) reconstruction based on deep networks shows promising advantages, such as large field of view, relatively high resolution, low cost, and ease of use. However, existing methods mainly consider vanilla scan strategies with limited inter-frame variations. These methods thus are degraded on complex but routine scan sequences in clinics. In this context, we propose a novel online learning framework for freehand 3D US reconstruction under complex scan strategies with diverse scanning velocities and poses. First, we devise a motion-weighted training loss in training phase to regularize the scan variation frame-by-frame and better mitigate the negative effects of uneven inter-frame velocity. Second, we effectively drive online learning with local-to-global pseudo supervisions. It mines both the frame-level contextual consistency and the path-level similarity constraint to improve the inter-frame transformation estimation. We explore a global adversarial shape before transferring the latent anatomical prior as supervision. Third, we build a feasible differentiable reconstruction approximation to enable the end-to-end optimization of our online learning. Experimental results illustrate that our freehand 3D US reconstruction framework outperformed current methods on two large, simulated datasets and one real dataset. In addition, we applied the proposed framework to clinical scan videos to further validate its effectiveness and generalizability.

Test-time bi-directional adaptation between image and model for robust segmentation.

BACKGROUND AND OBJECTIVE: Deep learning models often suffer from performance degradations when deployed in real clinical environments due to appearance shifts between training and testing images. Most extant methods use training-time adaptation, which almost require target domain samples in the training phase. However, these solutions are limited by the training process and cannot guarantee the accurate prediction of test samples with unforeseen appearance shifts. Further, it is impractical to collect target samples in advance. In this paper, we provide a general method of making existing segmentation models robust to samples with unknown appearance shifts when deployed in daily clinical practice. METHODS: Our proposed test-time bi-directional adaptation framework combines two complementary strategies. First, our image-to-model (I2M) adaptation strategy adapts appearance-agnostic test images to the learned segmentation model using a novel plug-and-play statistical alignment style transfer module during testing. Second, our model-to-image (M2I) adaptation strategy adapts the learned segmentation model to test images with unknown appearance shifts. This strategy applies an augmented self-supervised learning module to fine-tune the learned model with proxy labels that it generates. This innovative procedure can be adaptively constrained using our novel proxy consistency criterion. This complementary I2M and M2I framework demonstrably achieves robust segmentation against unknown appearance shifts using existing deep-learning models. RESULTS: Extensive experiments on 10 datasets containing fetal ultrasound, chest X-ray, and retinal fundus images demonstrate that our proposed method achieves promising robustness and efficiency in segmenting images with unknown appearance shifts. CONCLUSIONS: To address the appearance shift problem in clinically acquired medical images, we provide robust segmentation by using two complementary strategies. Our solution is general and amenable for deployment in clinical settings.

Sketch guided and progressive growing GAN for realistic and editable ultrasound image synthesis.

Ultrasound (US) imaging is widely used for anatomical structure inspection in clinical diagnosis. The training of new sonographers and deep learning based algorithms for US image analysis usually requires a large amount of data. However, obtaining and labeling large-scale US imaging data are not easy tasks, especially for diseases with low incidence. Realistic US image synthesis can alleviate this problem to a great extent. In this paper, we propose a generative adversarial network (GAN) based image synthesis framework. Our main contributions include: (1) we present the first work that can synthesize realistic B-mode US images with high-resolution and customized texture editing features; (2) to enhance structural details of generated images, we propose to introduce auxiliary sketch guidance into a conditional GAN. We superpose the edge sketch onto the object mask and use the composite mask as the network input; (3) to generate high-resolution US images, we adopt a progressive training strategy to gradually generate high-resolution images from low-resolution images. In addition, a feature loss is proposed to minimize the difference of high-level features between the generated and real images, which further improves the quality of generated images; (4) the proposed US image synthesis method is quite universal and can also be generalized to the US images of other anatomical structures besides the three ones tested in our study (lung, hip joint, and ovary); (5) extensive experiments on three large US image datasets are conducted to validate our method. Ablation studies, customized texture editing, user studies, and segmentation tests demonstrate promising results of our method in synthesizing realistic US images.

Joint Landmark and Structure Learning for Automatic Evaluation of Developmental Dysplasia of the Hip.

The ultrasound (US) screening of the infant hip is vital for the early diagnosis of developmental dysplasia of the hip (DDH). The US diagnosis of DDH refers to measuring alpha and beta angles that quantify hip joint development. These two angles are calculated from key anatomical landmarks and structures of the hip. However, this measurement process is not trivial for sonographers and usually requires a thorough understanding of complex anatomical structures. In this study, we propose a multi-task framework to learn the relationships among landmarks and structures jointly and automatically evaluate DDH. Our multi-task networks are equipped with three novel modules. Firstly, we adopt Mask R-CNN as the basic framework to detect and segment key anatomical structures and add one landmark detection branch to form a new multi-task framework. Secondly, we propose a novel shape similarity loss to refine the incomplete anatomical structure prediction robustly and accurately. Thirdly, we further incorporate the landmark-structure consistent prior to ensure the consistency of the bony rim estimated from the segmented structure and the detected landmark. In our experiments, 1231 US images of the infant hip from 632 patients are collected, of which 247 images from 126 patients are tested. The average errors in alpha and beta angles are 2.221 ° and 2.899 °. About 93% and 85% estimates of alpha and beta angles have errors less than 5 degrees, respectively. Experimental results demonstrate that the proposed method can accurately and robustly realize the automatic evaluation of DDH, showing great potential for clinical application.

Developing a Flexible National Wastewater Surveillance System for COVID-19 and Beyond.

BACKGROUND: Wastewater testing offers a cost-effective strategy for measuring population disease prevalence and health behaviors. For COVID-19, wastewater surveillance addresses testing gaps and provides an early warning for outbreaks. As U.S. federal agencies build a National Wastewater Surveillance System around the pandemic, thinking through ways to develop flexible frameworks for wastewater sampling, testing, and reporting can avoid unnecessary system overhauls for future infectious disease, chronic disease, and drug epidemics. OBJECTIVES: We discuss ways to transform a historically academic exercise into a tool for epidemic response. We generalize lessons learned by a global network of wastewater researchers around validation and implementation for COVID-19 and opioids while also drawing on our experience with wastewater-based epidemiology in the United States. DISCUSSION: Sustainable wastewater surveillance requires coordination between health and safety officials, utilities, labs, and researchers. Adapting sampling frequency, type, and location to threat level, community vulnerability, biomarker properties, and decisions that wastewater data will inform can increase the practical value of the data. Marketplace instabilities, coupled with a fragmented testing landscape due to specialization, may require officials to engage multiple labs to test for known and unknown threats. Government funding can stabilize the market, balancing commercial pressures with public good, and incentivize data sharing. When reporting results, standardizing metrics and contextualizing wastewater data with health resource data can provide insights into a community's vulnerability and identify strategies to prevent health care systems from being overwhelmed. If wastewater data will inform policy decisions for an entire community, comparing characteristics of the wastewater treatment plant's service population to those of the larger community can help determine whether the wastewater data are generalizable. Ethical protocols may be needed to protect privacy and avoid stigmatization. With data-driven approaches to sample collection, analysis, and interpretation, officials can use wastewater surveillance for adaptive resource allocation, pandemic management, and program evaluation. https://doi.org/10.1289/EHP8572.