RESUMO
Deep learning is a subfield of artificial intelligence and machine learning based mostly on neural networks and often combined with attention algorithms that has been used to detect and identify objects in text, audio, images, and video. Serghiou and Rough (Am J Epidemiol. 0000;000(00):0000-0000) present a primer for epidemiologists on deep learning models. These models provide substantial opportunities for epidemiologists to expand and amplify their research in both data collection and analyses by increasing the geographic reach of studies, including more research subjects, and working with large or high dimensional data. The tools for implementing deep learning methods are not quite yet as straightforward or ubiquitous for epidemiologists as traditional regression methods found in standard statistical software, but there are exciting opportunities for interdisciplinary collaboration with deep learning experts, just as epidemiologists have with statisticians, healthcare providers, urban planners, and other professionals. Despite the novelty of these methods, epidemiological principles of assessing bias, study design, interpretation and others still apply when implementing deep learning methods or assessing the findings of studies that have used them.
RESUMO
Pathologists have, over several decades, developed criteria for diagnosing and grading prostate cancer. However, this knowledge has not, so far, been included in the design of convolutional neural networks (CNN) for prostate cancer detection and grading. Further, it is not known whether the features learned by machine-learning algorithms coincide with diagnostic features used by pathologists. We propose a framework that enforces algorithms to learn the cellular and subcellular differences between benign and cancerous prostate glands in digital slides from hematoxylin and eosin-stained tissue sections. After accurate gland segmentation and exclusion of the stroma, the central component of the pipeline, named HistoEM, utilizes a histogram embedding of features from the latent space of the CNN encoder. Each gland is represented by 128 feature-wise histograms that provide the input into a second network for benign vs cancer classification of the whole gland. Cancer glands are further processed by a U-Net structured network to separate low-grade from high-grade cancer. Our model demonstrates similar performance compared with other state-of-the-art prostate cancer grading models with gland-level resolution. To understand the features learned by HistoEM, we first rank features based on the distance between benign and cancer histograms and visualize the tissue origins of the 2 most important features. A heatmap of pixel activation by each feature is generated using Grad-CAM and overlaid on nuclear segmentation outlines. We conclude that HistoEM, similar to pathologists, uses nuclear features for the detection of prostate cancer. Altogether, this novel approach can be broadly deployed to visualize computer-learned features in histopathology images.
Assuntos
Patologistas , Neoplasias da Próstata , Masculino , Humanos , Fluxo de Trabalho , Redes Neurais de Computação , Algoritmos , Neoplasias da Próstata/patologiaRESUMO
Nine in 10 road traffic deaths occur in low- and middle-income countries (LMICs). Despite this disproportionate burden, few studies have examined built environment correlates of road traffic injury in these settings, including in Latin America. We examined road traffic collisions in Bogotá, Colombia, occurring between 2015 and 2019, and assessed the association between neighborhood-level built environment features and pedestrian injury and death. We used descriptive statistics to characterize all police-reported road traffic collisions that occurred in Bogotá between 2015 and 2019. Cluster detection was used to identify spatial clustering of pedestrian collisions. Adjusted multivariate Poisson regression models were fit to examine associations between several neighborhood-built environment features and rate of pedestrian road traffic injury and death. A total of 173,443 police-reported traffic collisions occurred in Bogotá between 2015 and 2019. Pedestrians made up about 25% of road traffic injuries and 50% of road traffic deaths in Bogotá between 2015 and 2019. Pedestrian collisions were spatially clustered in the southwestern region of Bogotá. Neighborhoods with more street trees (RR, 0.90; 95% CI, 0.82-0.98), traffic signals (0.89, 0.81-0.99), and bus stops (0.89, 0.82-0.97) were associated with lower pedestrian road traffic deaths. Neighborhoods with greater density of large roads were associated with higher pedestrian injury. Our findings highlight the potential for pedestrian-friendly infrastructure to promote safer interactions between pedestrians and motorists in Bogotá and in similar urban contexts globally.
Assuntos
Acidentes de Trânsito , Ambiente Construído , Pedestres , Características de Residência , Ferimentos e Lesões , Humanos , Colômbia/epidemiologia , Acidentes de Trânsito/estatística & dados numéricos , Acidentes de Trânsito/mortalidade , Estudos Transversais , Adulto , Masculino , Feminino , Pedestres/estatística & dados numéricos , Adulto Jovem , Pessoa de Meia-Idade , Ferimentos e Lesões/epidemiologia , Adolescente , Características de Residência/estatística & dados numéricos , Criança , Pré-Escolar , Idoso , Planejamento AmbientalRESUMO
OBJECTIVE: The USA has higher rates of fatal motor vehicle collisions than most high-income countries. Previous studies examining the role of the built environment were generally limited to small geographic areas or single cities. This study aims to quantify associations between built environment characteristics and traffic collisions in the USA. METHODS: Built environment characteristics were derived from Google Street View images and summarised at the census tract level. Fatal traffic collisions were obtained from the 2019-2021 Fatality Analysis Reporting System. Fatal and non-fatal traffic collisions in Washington DC were obtained from the District Department of Transportation. Adjusted Poisson regression models examined whether built environment characteristics are related to motor vehicle collisions in the USA, controlling for census tract sociodemographic characteristics. RESULTS: Census tracts in the highest tertile of sidewalks, single-lane roads, streetlights and street greenness had 70%, 50%, 30% and 26% fewer fatal vehicle collisions compared with those in the lowest tertile. Street greenness and single-lane roads were associated with 37% and 38% fewer pedestrian-involved and cyclist-involved fatal collisions. Analyses with fatal and non-fatal collisions in Washington DC found streetlights and stop signs were associated with fewer pedestrians and cyclists-involved vehicle collisions while road construction had an adverse association. CONCLUSION: This study demonstrates the utility of using data algorithms that can automatically analyse street segments to create indicators of the built environment to enhance understanding of large-scale patterns and inform interventions to decrease road traffic injuries and fatalities.
RESUMO
INTRODUCTION: Manual motor problems have been reported in mild cognitive impairment (MCI) and Alzheimer's disease (AD), but the specific aspects that are affected, their neuropathology, and potential value for classification modeling is unknown. The current study examined if multiple measures of motor strength, dexterity, and speed are affected in MCI and AD, related to AD biomarkers, and are able to classify MCI or AD. METHODS: Fifty-three cognitively normal (CN), 33 amnestic MCI, and 28 AD subjects completed five manual motor measures: grip force, Trail Making Test A, spiral tracing, finger tapping, and a simulated feeding task. Analyses included (1) group differences in manual performance; (2) associations between manual function and AD biomarkers (PET amyloid ß, hippocampal volume, and APOE ε4 alleles); and (3) group classification accuracy of manual motor function using machine learning. RESULTS: Amnestic MCI and AD subjects exhibited slower psychomotor speed and AD subjects had weaker dominant hand grip strength than CN subjects. Performance on these measures was related to amyloid ß deposition (both) and hippocampal volume (psychomotor speed only). Support vector classification well-discriminated control and AD subjects (area under the curve of 0.73 and 0.77, respectively) but poorly discriminated MCI from controls or AD. CONCLUSION: Grip strength and spiral tracing appear preserved, while psychomotor speed is affected in amnestic MCI and AD. The association of motor performance with amyloid ß deposition and atrophy could indicate that this is due to amyloid deposition in and atrophy of motor brain regions, which generally occurs later in the disease process. The promising discriminatory abilities of manual motor measures for AD emphasize their value alongside other cognitive and motor assessment outcomes in classification and prediction models, as well as potential enrichment of outcome variables in AD clinical trials.
RESUMO
Passive design characteristics (PDC) play a pivotal role in reducing the energy burden on households without imposing additional financial constraints on project stakeholders. However, the scarcity of PDC data has posed a challenge in previous studies when assessing their energy-saving impact. To tackle this issue, this research introduces an innovative approach that combines deep learning-powered computer vision with machine learning techniques to examine the relationship between PDC and energy burden in residential buildings. In this study, we employ a convolutional neural network computer vision model to identify and measure key indicators, including window-to-wall ratio (WWR), external shading, and operable window types, using Google Street View images within the Chicago metropolitan area as our case study. Subsequently, we utilize the derived passive design features in conjunction with demographic characteristics to train and compare various machine learning methods. These methods encompass Decision Tree Regression, Random Forest Regression, and Support Vector Regression, culminating in the development of a comprehensive model for energy burden prediction. Our framework achieves a 74.2% accuracy in forecasting the average energy burden. These results yield invaluable insights for policymakers and urban planners, paving the way toward the realization of smart and sustainable cities.
RESUMO
Adversarial training, especially projected gradient descent (PGD), has proven to be a successful approach for improving robustness against adversarial attacks. After adversarial training, gradients of models with respect to their inputs have a preferential direction. However, the direction of alignment is not mathematically well established, making it difficult to evaluate quantitatively. We propose a novel definition of this direction as the direction of the vector pointing toward the closest point of the support of the closest inaccurate class in decision space. To evaluate the alignment with this direction after adversarial training, we apply a metric that uses generative adversarial networks to produce the smallest residual needed to change the class present in the image. We show that PGD-trained models have a higher alignment than the baseline according to our definition, that our metric presents higher alignment values than a competing metric formulation, and that enforcing this alignment increases the robustness of models.
RESUMO
BACKGROUND: The built environment is a structural determinant of health and has been shown to influence health expenditures, behaviors, and outcomes. Traditional methods of assessing built environment characteristics are time-consuming and difficult to combine or compare. Google Street View (GSV) images represent a large, publicly available data source that can be used to create indicators of characteristics of the physical environment with machine learning techniques. The aim of this study is to use GSV images to measure the association of built environment features with health-related behaviors and outcomes at the census tract level. METHODS: We used computer vision techniques to derive built environment indicators from approximately 31 million GSV images at 7.8 million intersections. Associations between derived indicators and health-related behaviors and outcomes on the census-tract level were assessed using multivariate regression models, controlling for demographic factors and socioeconomic position. Statistical significance was assessed at the α = 0.05 level. RESULTS: Single lane roads were associated with increased diabetes and obesity, while non-single-family home buildings were associated with decreased obesity, diabetes and inactivity. Street greenness was associated with decreased prevalence of physical and mental distress, as well as decreased binge drinking, but with increased obesity. Socioeconomic disadvantage was negatively associated with binge drinking prevalence and positively associated with all other health-related behaviors and outcomes. CONCLUSIONS: Structural determinants of health such as the built environment can influence population health. Our study suggests that higher levels of urban development have mixed effects on health and adds further evidence that socioeconomic distress has adverse impacts on multiple physical and mental health outcomes.
Assuntos
Ambiente Construído/estatística & dados numéricos , Saúde da População Urbana/estatística & dados numéricos , Cidades , Sistemas de Informação Geográfica , Humanos , Estados UnidosRESUMO
The morphological effect of impurities on α-U3O8 has been investigated. This study provides the first evidence that the presence of impurities can alter nuclear material morphology, and these changes can be quantified to aid in revealing processing history. Four elements: Ca, Mg, V, and Zr were implemented in the uranyl peroxide synthesis route and studied individually within the α-U3O8. Six total replicates were synthesized, and replicates 1-3 were filtered and washed with Millipore water (18.2 MΩ) to remove any residual nitrates. Replicates 4-6 were filtered but not washed to determine the amount of impurities removed during washing. Inductively coupled plasma mass spectrometry (ICP-MS) was employed at key points during the synthesis to quantify incorporation of the impurity. Each sample was characterized using powder X-ray diffraction (p-XRD), high-resolution scanning electron microscopy (HRSEM), and SEM with energy dispersive X-ray spectroscopy (SEM-EDS). p-XRD was utilized to evaluate any crystallographic changes due to the impurities; HRSEM imagery was analyzed with Morphological Analysis for MAterials (MAMA) software and machine learning classification for quantification of the morphology; and SEM-EDS was utilized to locate the impurity within the α-U3O8. All samples were found to be quantifiably distinguishable, further demonstrating the utility of quantitative morphology as a signature for the processing history of nuclear material.
RESUMO
The local field potential (LFP) is of growing importance in neurophysiology as a metric of network activity and as a readout signal for use in brain-machine interfaces. However, there are uncertainties regarding the kind and visual field extent of information carried by LFP signals, as well as the specific features of the LFP signal conveying such information, especially under naturalistic conditions. To address these questions, we recorded LFP responses to natural images in V1 of awake and anesthetized macaques using Utah multielectrode arrays. First, we have shown that it is possible to identify presented natural images from the LFP responses they evoke using trained Gabor wavelet (GW) models. Because GW models were devised to explain the spiking responses of V1 cells, this finding suggests that local spiking activity and LFPs (thought to reflect primarily local synaptic activity) carry similar visual information. Second, models trained on scalar metrics, such as the evoked LFP response range, provide robust image identification, supporting the informative nature of even simple LFP features. Third, image identification is robust only for the first 300 ms following image presentation, and image information is not restricted to any of the spectral bands. This suggests that the short-latency broadband LFP response carries most information during natural scene viewing. Finally, best image identification was achieved by GW models incorporating information at the scale of â¼ 0.5° in size and trained using four different orientations. This suggests that during natural image viewing, LFPs carry stimulus-specific information at spatial scales corresponding to few orientation columns in macaque V1.
Assuntos
Potenciais Evocados Visuais , Córtex Visual/fisiologia , Percepção Visual , Animais , Macaca fascicularis , Masculino , Estimulação LuminosaRESUMO
Background: This study utilizes innovative computer vision methods alongside Google Street View images to characterize neighborhood built environments across Utah. Methods: Convolutional Neural Networks were used to create indicators of street greenness, crosswalks, and building type on 1.4 million Google Street View images. The demographic and medical profiles of Utah residents came from the Utah Population Database (UPDB). We implemented hierarchical linear models with individuals nested within zip codes to estimate associations between neighborhood built environment features and individual-level obesity and diabetes, controlling for individual- and zip code-level characteristics (n = 1,899,175 adults living in Utah in 2015). Sibling random effects models were implemented to account for shared family attributes among siblings (n = 972,150) and twins (n = 14,122). Results: Consistent with prior neighborhood research, the variance partition coefficients (VPC) of our unadjusted models nesting individuals within zip codes were relatively small (0.5%-5.3%), except for HbA1c (VPC = 23%), suggesting a small percentage of the outcome variance is at the zip code-level. However, proportional change in variance (PCV) attributable to zip codes after the inclusion of neighborhood built environment variables and covariates ranged between 11% and 67%, suggesting that these characteristics account for a substantial portion of the zip code-level effects. Non-single-family homes (indicator of mixed land use), sidewalks (indicator of walkability), and green streets (indicator of neighborhood aesthetics) were associated with reduced diabetes and obesity. Zip codes in the third tertile for non-single-family homes were associated with a 15% reduction (PR: 0.85; 95% CI: 0.79, 0.91) in obesity and a 20% reduction (PR: 0.80; 95% CI: 0.70, 0.91) in diabetes. This tertile was also associated with a BMI reduction of -0.68 kg/m2 (95% CI: -0.95, -0.40). Conclusion: We observe associations between neighborhood characteristics and chronic diseases, accounting for biological, social, and cultural factors shared among siblings in this large population-based study.
RESUMO
Convolutional neural networks (CNNs) have been successfully applied to chest x-ray (CXR) images. Moreover, annotated bounding boxes have been shown to improve the interpretability of a CNN in terms of localizing abnormalities. However, only a few relatively small CXR datasets containing bounding boxes are available, and collecting them is very costly. Opportunely, eye-tracking (ET) data can be collected during the clinical workflow of a radiologist. We use ET data recorded from radiologists while dictating CXR reports to train CNNs. We extract snippets from the ET data by associating them with the dictation of keywords and use them to supervise the localization of specific abnormalities. We show that this method can improve a model's interpretability without impacting its image-level classification.
RESUMO
This paper aims to address the challenges associated with evaluating the impact of neighborhood environments on health outcomes. Google street view (GSV) images provide a valuable tool for assessing neighborhood environments on a large scale. By annotating the GSV images with labels indicating the presence or absence of specific neighborhood features, we can develop classifiers capable of automatically analyzing and evaluating the environment. However, the process of labeling GSV images to analyze and evaluate the environment is a time-consuming and labor-intensive task. To overcome these challenges, we propose using a multi-task classifier to enhance the training of classifiers with limited supervised GSV data. Our multi-task classifier utilizes readily available, inexpensive online images collected from Flickr as a related classification task. The hypothesis is that a classifier trained on multiple related tasks is less likely to overfit to small amounts of training data and generalizes better to unseen data. We leverage the power of multiple related tasks to improve the classifier's overall performance and generalization capability. Here we show that, with the proposed learning paradigm, predicted labels for GSV test images are more accurate. Across different environment indicators, the accuracy, F1 score and balanced accuracy increase up to 6 % in the multi-task learning framework compared to its single-task learning counterpart. The enhanced accuracy of the predicted labels obtained through the multi-task classifier contributes to a more reliable and precise regression analysis determining the correlation between predicted built environment indicators and health outcomes. The R2 values calculated for different health outcomes improve by up to 4 % using multi-task learning detected indicators.
RESUMO
BACKGROUND: Despite reports of gross motor problems in mild cognitive impairment (MCI) and Alzheimer's disease (AD), fine motor function has been relatively understudied. OBJECTIVE: We examined if finger tapping is affected in AD, related to AD biomarkers, and able to classify MCI or AD. METHODS: Forty-seven cognitively normal, 27 amnestic MCI, and 26 AD subjects completed unimanual and bimanual computerized tapping tests. We tested 1) group differences in tapping with permutation models; 2) associations between tapping and biomarkers (PET amyloid-ß, hippocampal volume, and APOEÉ4 alleles) with linear regression; and 3) the predictive value of tapping for group classification using machine learning. RESULTS: AD subjects had slower reaction time and larger speed variability than controls during all tapping conditions, except for dual tapping. MCI subjects performed worse than controls on reaction time and speed variability for dual and non-dominant hand tapping. Tapping speed and variability were related to hippocampal volume, but not to amyloid-ß deposition or APOEÉ4 alleles. Random forest classification (overall accuracyâ=â70%) discriminated control and AD subjects, but poorly discriminated MCI from controls or AD. CONCLUSIONS: MCI and AD are linked to more variable finger tapping with slower reaction time. Associations between finger tapping and hippocampal volume, but not amyloidosis, suggest that tapping deficits are related to neuropathology that presents later during the disease. Considering that tapping performance is able to differentiate between control and AD subjects, it can offer a cost-efficient tool for augmenting existing AD biomarkers.
Assuntos
Doença de Alzheimer , Amiloidose , Disfunção Cognitiva , Humanos , Doença de Alzheimer/psicologia , Peptídeos beta-Amiloides , Disfunção Cognitiva/psicologia , BiomarcadoresRESUMO
"Drivers" are theorized mechanisms for persistent atrial fibrillation. Machine learning algorithms have been used to identify drivers, but the small size of current driver datasets limits their performance. We hypothesized that pretraining with unsupervised learning on a large dataset of unlabeled electrograms would improve classifier accuracy on a smaller driver dataset. In this study, we used a SimCLR-based framework to pretrain a residual neural network on a dataset of 113K unlabeled 64-electrode measurements and found weighted testing accuracy to improve over a non-pretrained network (78.6±3.9% vs 71.9±3.3%). This lays ground for development of superior driver detection algorithms and supports use of transfer learning for other datasets of endocardial electrograms.
RESUMO
Artificial intelligence - machine learning (AI-ML) is a computational technique that has been demonstrated to be able to extract meaningful clinical information from diagnostic data that are not available using either human interpretation or more simple analysis methods. Recent developments have shown that AI-ML approaches applied to ECGs can accurately predict different patient characteristics and pathologies not detectable by expert physician readers. There is an extensive body of literature surrounding the use of AI-ML in other fields, which has given rise to an array of predefined open-source AI-ML architectures which can be translated to new problems in an "off-the-shelf" manner. Applying "off-the-shelf" AI-ML architectures to ECG-based datasets opens the door for rapid development and identification of previously unknown disease biomarkers. Despite the excellent opportunity, the ideal open-source AI-ML architecture for ECG related problems is not known. Furthermore, there has been limited investigation on how and when these AI-ML approaches fail and possible bias or disparities associated with particular network architectures. In this study, we aimed to: (1) determine if open-source, "off-the-shelf" AI-ML architectures could be trained to classify low LVEF from ECGs, (2) assess the accuracy of different AI-ML architectures compared to each other, and (3) to identify which, if any, patient characteristics are associated with poor AI-ML performance.
RESUMO
Circuitry mapping of metazoan neural systems is difficult because canonical neural regions (regions containing one or more copies of all components) are large, regional borders are uncertain, neuronal diversity is high, and potential network topologies so numerous that only anatomical ground truth can resolve them. Complete mapping of a specific network requires synaptic resolution, canonical region coverage, and robust neuronal classification. Though transmission electron microscopy (TEM) remains the optimal tool for network mapping, the process of building large serial section TEM (ssTEM) image volumes is rendered difficult by the need to precisely mosaic distorted image tiles and register distorted mosaics. Moreover, most molecular neuronal class markers are poorly compatible with optimal TEM imaging. Our objective was to build a complete framework for ultrastructural circuitry mapping. This framework combines strong TEM-compliant small molecule profiling with automated image tile mosaicking, automated slice-to-slice image registration, and gigabyte-scale image browsing for volume annotation. Specifically we show how ultrathin molecular profiling datasets and their resultant classification maps can be embedded into ssTEM datasets and how scripted acquisition tools (SerialEM), mosaicking and registration (ir-tools), and large slice viewers (MosaicBuilder, Viking) can be used to manage terabyte-scale volumes. These methods enable large-scale connectivity analyses of new and legacy data. In well-posed tasks (e.g., complete network mapping in retina), terabyte-scale image volumes that previously would require decades of assembly can now be completed in months. Perhaps more importantly, the fusion of molecular profiling, image acquisition by SerialEM, ir-tools volume assembly, and data viewers/annotators also allow ssTEM to be used as a prospective tool for discovery in nonneural systems and a practical screening methodology for neurogenetics. Finally, this framework provides a mechanism for parallelization of ssTEM imaging, volume assembly, and data analysis across an international user base, enhancing the productivity of a large cohort of electron microscopists.
Assuntos
Processamento de Imagem Assistida por Computador , Rede Nervosa/ultraestrutura , Neurônios Retinianos/ultraestrutura , Animais , Mapeamento Encefálico , Simulação por Computador , Feminino , Armazenamento e Recuperação da Informação , Masculino , Camundongos , Microscopia Eletrônica de Transmissão , Modelos Neurológicos , Coelhos , Neurônios Retinianos/fisiologiaRESUMO
As climate change increases the frequency and intensity of devastating and unpredictable extreme heat events, developments to the built environment should consider instigating practices that minimize the likelihood of indoor overheating during hot weather. Heatwaves are the leading cause of death among weather-related causes worldwide, including in developed and developing countries. In this empirical study, a four-step approach was used to collect, extract and analyze data from twenty-seven states in the United States. Three housing characteristic categories (i.e., general housing conditions, living conditions, and housing thermal inertia) and eight variables were extracted from the American Housing Survey database, ResStock database and CDC's National Environmental Public Health Tracking Network. Multivariable regression models were used to understand the influential variables, a multicollinearity test was used to determine the dependence of those variables, and then a logistic model was used to verify the results. Three variables-housing age (HA), housing crowding ratio (HCR), and roof condition (RC)-were found to be correlated with the risk of heat-related illness (HRI) indexes. Then, a logistic regression model was generated using the three variables to predict the risk of heat-related emergency department visits (EDV) and heat-related mortality (MORD) on a state level. The results indicate that the proposed logistic regression model correctly predicted 100% of the high-risk states for MORD for the eight states tested. Overall, this analysis provides additional evidence about the housing character variables that influence HRI. The outcomes also reinforce the concept of the built environment determined health and demonstrate that the built environment, especially housing, should be considered in techniques for mitigating climate change-exacerbated health conditions.
Assuntos
Calor Extremo , Transtornos de Estresse por Calor , Humanos , Estados Unidos/epidemiologia , Temperatura Alta , Transtornos de Estresse por Calor/epidemiologia , Habitação , Calor Extremo/efeitos adversos , Mudança ClimáticaRESUMO
Deep learning has shown recent success in classifying anomalies in chest x-rays, but datasets are still small compared to natural image datasets. Supervision of abnormality localization has been shown to improve trained models, partially compensating for dataset sizes. However, explicitly labeling these anomalies requires an expert and is very time-consuming. We propose a potentially scalable method for collecting implicit localization data using an eye tracker to capture gaze locations and a microphone to capture a dictation of a report, imitating the setup of a reading room. The resulting REFLACX (Reports and Eye-Tracking Data for Localization of Abnormalities in Chest X-rays) dataset was labeled across five radiologists and contains 3,032 synchronized sets of eye-tracking data and timestamped report transcriptions for 2,616 chest x-rays from the MIMIC-CXR dataset. We also provide auxiliary annotations, including bounding boxes around lungs and heart and validation labels consisting of ellipses localizing abnormalities and image-level labels. Furthermore, a small subset of the data contains readings from all radiologists, allowing for the calculation of inter-rater scores.
Assuntos
Tecnologia de Rastreamento Ocular , Radiografia Torácica , Aprendizado Profundo , Humanos , Radiografia , Raios XRESUMO
Built environment neighborhood characteristics are difficult to measure and assess on a large scale. Consequently, there is a lack of sufficient data that can help us investigate neighborhood characteristics as structural determinants of health on a national level. The objective of this study is to utilize publicly available Google Street View images as a data source for characterizing built environments and to examine the influence of built environments on chronic diseases and health behaviors in the United States. Data were collected by processing 164 million Google Street View images from November 2019 across the United States. Convolutional Neural Networks, a class of multi-layer deep neural networks, were used to extract features of the built environment. Validation analyses found accuracies of 82% or higher across neighborhood characteristics. In regression analyses controlling for census tract sociodemographics, we find that single-lane roads (an indicator of lower urban development) were linked with chronic conditions and worse mental health. Walkability and urbanicity indicators such as crosswalks, sidewalks, and two or more cars were associated with better health, including reduction in depression, obesity, high blood pressure, and high cholesterol. Street signs and streetlights were also found to be associated with decreased chronic conditions. Chain link fence (physical disorder indicator) was generally associated with poorer mental health. Living in neighborhoods with a built environment that supports social interaction and physical activity can lead to positive health outcomes. Computer vision models using manually annotated Google Street View images as a training dataset were able to accurately identify neighborhood built environment characteristics. These methods increases the feasibility, scale, and efficiency of neighborhood studies on health.