Grant J. Scott

The integration of Generative AI and satellite technology offers groundbreaking potential for addressing chronic health issues like obesity in public health research. This chapter explores the innovative application of deep learning and satellite imagery to predict obesity prevalence in Missouri, emphasizing a multidisciplinary approach that integrates public health, artificial intelligence, and remote sensing. Utilizing medium-resolution Sentinel-2 satellite imagery, deep convolutional neural networks (DCNNs), specifically ResNet-50, are employed to extract deep neural visual features (DNVFs) from each census tract. This method transcends traditional data sources by integrating obesity prevalence estimates from the Centers for Disease Control and Prevention (CDC) with advanced image processing techniques to provide a nuanced analysis of environmental and societal determinants of health. By focusing on specific census tracts, the study illustrates how localized data can inform targeted interventions, thereby enhancing the effectiveness of health policies and practices. This approach not only reflects the practical application of user innovation in the field of public health but also sets a precedent for the future use of Generative AI to tackle complex health disparities. The findings encourage further exploration into the potential of this technology to expand its application to a broader geographical scale, thereby maximizing the impact of AI-driven public health initiatives.

This chapter dives into the transformative role of user innovation and generative AI in enhancing public health responses during the COVID-19 pandemic, with a specific focus on Boone County, Missouri. By integrating advanced AI analytics with proactive community engagement, it provides a nuanced examination of how demographic factors influence COVID-19 positivity rates and outcomes. This discussion underscores the utility of AI-driven tools in providing pivotal insights that guide public health strategies and interventions. Crucially, the chapter illustrates how end-users, by accessing and interpreting these insights, can make informed decisions to protect themselves and their communities. This engagement not only bolsters the responsiveness of public health measures but also empowers individuals with the knowledge to navigate the pandemic more safely and effectively. Moreover, the chapter confronts the technological, logistical, and ethical challenges of weaving AI into public health practices, proposing adaptive solutions and frameworks to cultivate user-driven innovation in managing health crises. The insights offered advocate for a collaborative model where user innovation and AI create a more inclusive and effective public health ecosystem, presenting a roadmap for leveraging technology to enhance community resilience in facing future pandemics.

Wildfires continue to occur every year, resulting in millions of acres of burned areas. This makes the task of mapping the burned areas arduous, and it is exacerbated by the fact that many of these burned areas are in remote and hard-to-reach regions. Satellite imagery with the highest freely available resolution combined with the latest deep learning architectures offer an opportunity to map these burned areas using models trained for semantic segmentation. In this work, we evaluate the state-of-the- art deep learning transformer architecture SegFormer and convolutional attention architecture SegNext specifically trained for the task of burned area semantic segmentation. The trained models are compared against U-Net and DeepLab neural network models trained on the same dataset. SegNext achieved a precision of 88.79%, a 4.44% increase compared to the best precision achieved by DeepLabV3+, and an Intersection-Over-Union score of 76.71%, a 4.71% increase compared to the best score achieved by DeepLabV3.

Background: The global obesity epidemic demands innovative approaches to understand its complex environmental and social determinants. Spatial technologies, such as geographic information systems, remote sensing, and spatial machine learning, offer new insights into this health issue. This study uses deep learning and spatial modeling to predict obesity rates for census tracts in Missouri. Objective: This study aims to develop a scalable method for predicting obesity prevalence using deep convolutional neural networks applied to satellite imagery and geospatial analysis, focusing on 1052 census tracts in Missouri. Methods: Our analysis followed 3 steps. First, Sentinel-2 satellite images were processed using the Residual Network-50 model to extract environmental features from 63,592 image chips (224×224 pixels). Second, these features were merged with obesity rate data from the Centers for Disease Control and Prevention for Missouri census tracts. Third, a spatial lag model was used to predict obesity rates and analyze the association between deep neural visual features and obesity prevalence. Spatial autocorrelation was used to identify clusters of obesity rates. Results: Substantial spatial clustering of obesity rates was found across Missouri, with a Moran I value of 0.68, indicating similar obesity rates among neighboring census tracts. The spatial lag model demonstrated strong predictive performance, with an R2 of 0.93 and a spatial pseudo R2 of 0.92, explaining 93% of the variation in obesity rates. Local indicators from a spatial association analysis revealed regions with distinct high and low clusters of obesity, which were visualized through choropleth maps. Conclusions: This study highlights the effectiveness of integrating deep convolutional neural networks and spatial modeling to predict obesity prevalence based on environmental features from satellite imagery. The model’s high accuracy and ability to capture spatial patterns offer valuable insights for public health interventions. Future work should expand the geographical scope and include socioeconomic data to further refine the model for broader applications in obesity research.

This research study investigates and predicts the obesity prevalence in Missouri, utilizing deep neural visual features extracted from medium-resolution satellite imagery (Sentinel-2). By applying a deep convolutional neural network (DCNN), the study aims to predict the obesity rate of census tracts based on visual features in the satellite imagery that covers each tract. The study utilizes Sentinel-2 satellite images, processed using the ResNet-50 DCNN, to extract deep neural visual features (DNVF). Obesity prevalence data, sourced from the CDC’s 2022 estimates, is analyzed at the census tract level. The datasets were integrated to apply a machine learning model to predict the obesity rates in 1052 different census tracts in Missouri. The analysis reveals significant associations between DNVF and obesity prevalence. The predictive models show moderate success in estimating and predicting obesity rates in various census tracts within Missouri. The study emphasizes the potential of using satellite imagery and advanced machine learning in public health research. It points to environmental factors as significant determinants of obesity, suggesting the need for targeted health interventions. Employing DNVF to explore and predict obesity rates offers valuable insights for public health strategies and calls for expanded research in diverse geographical contexts.

BACKGROUND The global obesity epidemic demands innovative approaches to understand its complex environmental and social determinants. Spatial technologies, such as geographic information systems, remote sensing, and spatial machine learning, offer new insights into this health issue. This study uses deep learning and spatial modeling to predict obesity rates for census tracts in Missouri. OBJECTIVE This study aims to develop a scalable method for predicting obesity prevalence using deep convolutional neural networks applied to satellite imagery and geospatial analysis, focusing on 1052 census tracts in Missouri. METHODS Our analysis followed 3 steps. First, Sentinel-2 satellite images were processed using the Residual Network-50 model to extract environmental features from 63,592 image chips (224×224 pixels). Second, these features were merged with obesity rate data from the Centers for Disease Control and Prevention for Missouri census tracts. Third, a spatial lag model was used to predict obesity rates and analyze the association between deep neural visual features and obesity prevalence. Spatial autocorrelation was used to identify clusters of obesity rates. RESULTS Substantial spatial clustering of obesity rates was found across Missouri, with a Moran I value of 0.68, indicating similar obesity rates among neighboring census tracts. The spatial lag model demonstrated strong predictive performance, with an R2 of 0.93 and a spatial pseudo R2 of 0.92, explaining 93% of the variation in obesity rates. Local indicators from a spatial association analysis revealed regions with distinct high and low clusters of obesity, which were visualized through choropleth maps. CONCLUSIONS This study highlights the effectiveness of integrating deep convolutional neural networks and spatial modeling to predict obesity prevalence based on environmental features from satellite imagery. The model’s high accuracy and ability to capture spatial patterns offer valuable insights for public health interventions. Future work should expand the geographical scope and include socioeconomic data to further refine the model for broader applications in obesity research.

The COVID-19 pandemic has had a deep influence on American life in general and on the American economy in particular. However, the burden of the pandemic has not been distributed equally among members of a population based on their social-determinants-of-health. The purpose of this study was to investigate whether the median income was associated with COVID-19 total number of tests and positivity rate in Boone County, Missouri during the pandemic. We analyzed the geospatial data using three heat maps showing the Census tract-wise COVID-19 positivity rate, Census tract-wise median income, and Census tract-wise total number of COVID-19 tests to highlight our study findings. Our study results support the hypothesis that individuals with lower median income tend to have a lower total number of COVID-19 tests and higher COVID-19 positivity rates in Boone County, Missouri. The Pearson correlation coefficient between the positivity rate and median income is -0.324.

Deforestation poses a critical environmental challenge with far-reaching impacts on climate change, biodiversity, and local communities. As such, detecting and monitoring deforestation are crucial, and recent advancements in deep learning and remote sensing technologies offer a promising solution to this challenge. In this study, we adapt ChangeFormer, a transformer-based framework, to detect deforestation in the Brazilian Amazon, employing the attention mechanism to analyze spatial and temporal patterns in bi-temporal satellite images. To assess the model’s effectiveness, we employed a robust approach to create a deforestation detection dataset, utilizing Sentinel-2 imagery from select conservation areas in the Brazilian Amazon throughout 2020 and 2021. Our dataset comprises 7,734 pairs of bi-temporal image chips with a resolution of 256×256 pixels and 1,406 pairs of image chips with a resolution of 512×512 pixels. The model achieved an overall accuracy of 93% with corresponding F1 score of 90% and IoU score of 82%. These results demonstrate the potential of transformer-based networks for accurate and efficient deforestation detection.