Region fine-grained attention network for accurate bone age assessment.

Bone age assessment plays a vital role in monitoring the growth and development of adolescents. However, it is still challenging to obtain precise bone age from hand radiography due to these problems: 1) Hand bone varies greatly and is always masked by the background; 2) the hand bone radiographs with successive ages offer high similarity. To solve such issues, a region fine-grained attention network (RFGA-Net) was proposed for bone age assessment, where the region aware attention (RAA) module was developed to distinguish the skeletal regions from the background by modeling global spatial dependency; then the fine-grained feature attention (FFA) module was devised to identify similar bone radiographs by recognizing critical fine-grained feature regions. The experimental results demonstrate that the proposed RFGA-Net shows the best performance on the Radiological Society of North America (RSNA) pediatric bone dataset, achieving the mean absolute error (MAE) of 3.34 and the root mean square error (RMSE) of 4.02, respectively.

Assessment of air pollution and air quality perception mismatch using mobility-based real-time exposure.

Air pollution poses a threat to human health. Public perceptions of air pollution are important for individual self-protection and policy-making. Given the uncertainty faced by residence-based exposure (RB) measurements, this study measures individuals' real-time mobility-based (MB) exposures and perceptions of air pollution by considering people's daily movement. It explores how contextual uncertainties may influence the disparities in perceived air quality by taking into account RB and MB environmental factors. In addition, we explore factors that are related to the mismatch between people's perceived air quality and actual air pollution exposure. Using K-means clustering to divide the PM2.5 values into two groups, a mismatch happens when the perceived air quality is poor but the air pollution level is lower than 15.536µg/m3 and when the perceived air quality is good but the air pollution level is higher than 15.608µg/m3. The results show that there is a mismatch between air pollution exposure and perception of air pollution. People with low income are exposed to higher air pollution. Unemployed people and people with more serious mental health symptoms (e.g., depression) have a higher chance of accurately assessing air pollution (e.g., perceiving air quality as poor when air pollution levels are high). Older people and those with a higher MB open space density tend to underestimate air pollution. Students tend to perceive air quality as good. People who are surrounded by higher MB transportation land-use density and green space density tend to perceive air quality as poor. The results can help policymakers to increase public awareness of high air pollution areas, and consider the health effects of landscapes during planning.

Bone age assessment from articular surface and epiphysis using deep neural networks.

Bone age assessment is of great significance to genetic diagnosis and endocrine diseases. Traditional bone age diagnosis mainly relies on experienced radiologists to examine the regions of interest in hand radiography, but it is time-consuming and may even lead to a vast error between the diagnosis result and the reference. The existing computer-aided methods predict bone age based on general regions of interest but do not explore specific regions of interest in hand radiography. This paper aims to solve such problems by performing bone age prediction on the articular surface and epiphysis from hand radiography using deep convolutional neural networks. The articular surface and epiphysis datasets are established from the Radiological Society of North America (RSNA) pediatric bone age challenge, where the specific feature regions of the articular surface and epiphysis are manually segmented from hand radiography. Five convolutional neural networks, i.e., ResNet50, SENet, DenseNet-121, EfficientNet-b4, and CSPNet, are employed to improve the accuracy and efficiency of bone age diagnosis in clinical applications. Experiments show that the best-performing model can yield a mean absolute error (MAE) of 7.34 months on the proposed articular surface and epiphysis datasets, which is more accurate and fast than the radiologists. The project is available at https://github.com/YameiDeng/BAANet/, and the annotated dataset is also published at https://doi.org/10.5281/zenodo.7947923.

Field Evaluation and Calibration of Low-Cost Air Pollution Sensors for Environmental Exposure Research.

This paper seeks to evaluate and calibrate data collected by low-cost particulate matter (PM) sensors in different environments and using different aggregated temporal units (i.e., 5-s, 1-min, 10-min, 30 min intervals). We first collected PM concentrations (i.e., PM1, PM2.5, and PM10) data in five different environments (i.e., indoor and outdoor of an office building, a train platform and lobby of a subway station, and a seaside location) in Hong Kong, using five AirBeam2 sensors as the low-cost sensors and a TSI DustTrak DRX Aerosol Monitor 8533 as the reference sensor. By comparing the collected PM concentrations, we found high linearity and correlation between the data reported by the AirBeam2 sensors in different environments. Furthermore, the results suggest that the accuracy and bias of the PM data reported by the AirBeam2 sensors are affected by rainy weather and environments with high humidity and a high level of hygroscopic salts (i.e., a seaside location). In addition, increasing the aggregation level of the temporal units (i.e., from 5-s to 30 min intervals) increases the correlation between the PM concentrations obtained by the AirBeam2 sensors, while it does not significantly improve the accuracy and bias of the data. Lastly, our results indicate that using a machine learning model (i.e., random forest) for the calibration of PM concentrations collected on sunny days generates better results than those obtained with multiple linear models. These findings have important implications for researchers when designing environmental exposure studies based on low-cost PM sensors.

Evaluating the Inequality of Medical Service Accessibility Using Smart Card Data.

The measurement of medical service accessibility is typically based on driving or Euclidean distance. However, in most non-emergency cases, public transport is the travel mode used by the public to access medical services. Yet, there has been little evaluation of the public transport system-based inequality of medical service accessibility. This work uses massive real smart card data (SCD) and an improved potential model to estimate the public transport-based medical service accessibility in Beijing, China. These real SCD data are used to calculate travel costs in terms of time and distance, and medical service accessibility is estimated using an improved potential model. The spatiotemporal variations and patterns of medical service accessibility are explored, and the results show that it is unevenly spatiotemporally distributed across the study area. For example, medical service accessibility in urban areas is higher than that in suburban areas, accessibility during peak periods is higher than that during off-peak periods, and accessibility on weekends is generally higher than that on weekdays. To explore the association of medical service accessibility with socio-economic factors, the relationship between accessibility and house price is investigated via a spatial econometric analysis. The results show that, at a global level, house price is positively correlated with medical service accessibility. In particular, the medical service accessibility of a higher-priced spatial housing unit is lower than that of its neighboring spatial units, owing to the positive spatial spillover effect of house price. This work sheds new light on the inequality of medical service accessibility from the perspective of public transport, which may benefit urban policymakers and planners.

Biliary stenting for unresectable cholangiocarcinoma: A population-based study of long-term outcomes and hospital costs in Taiwan.

Few studies have compared percutaneous biliary stenting (PBS) and endoscopic biliary stenting (EBS) in terms of long-term effects on cholangiocarcinoma (CC), and few have systematically evaluated outcome associations in Taiwan. This study aimed to compare long-term outcomes between two treatments for unresectable CC: PBS and EBS. After propensity score matching (PSM) to reduce the effect of selection bias, 1002 CC patients were included in this analysis: 501 in the PBS group and 501 in the EBS group. The Kaplan-Meier method was used to construct the survival curve for all CC patients, and the Cox proportional hazards model was used for multivariate assessment of outcome predictors. After PSM, group comparisons revealed a significantly longer length of stay in the PBS group compared to the EBS group (25 days vs. 19 days, respectively; p < 0.001). Hospital costs were also significantly higher in the PBS group than in the EBS group (US$126,575 vs. US$89,326, respectively; p < 0.001). The median survival time was 3.7 months in all CC patients, 3.5 months in the PBS group, and 4.0 months in the EBS group. The 1-year, 3-year, and 5-year survival rates were 17.6%, 6.1%, and 3.2% in all CC patients; 16.6%, 4.8%, and 3.2% in the PBS group; and 18.6%, 7.27%, and 3% in the EBS group, respectively. The most important predictor of survival is extrahepatic CC. Medical professionals and healthcare providers should carefully consider the use of EBS for initial treatment of obstructive jaundice in patients with unresectable CC.