An Exploratory Approach to Deriving Nutrition Information of Restaurant Food from Crowdsourced Food Images: Case of Hartford

Deep learning models can recognize the food item in an image and derive their nutrition information, including calories, macronutrients (carbohydrates, fats, and proteins), and micronutrients (vitamins and minerals). This technology has yet to be implemented for the nutrition assessment of restaurant food. In this paper, we crowdsource 15,908 food images of 470 restaurants in the Greater Hartford region on Tripadvisor and Google Place. These food images are loaded into a proprietary deep learning model (Calorie Mama) for nutrition assessment. We employ manual coding to validate the model accuracy based on the Food and Nutrient Database for Dietary Studies. The derived nutrition information is visualized at both the restaurant level and the census tract level. The deep learning model achieves 75.1% accuracy when compared with manual coding. It has more accurate labels for ethnic foods but cannot identify portion sizes, certain food items (e.g., specialty burgers and salads), and multiple food items in an image. The restaurant nutrition (RN) index is further proposed based on the derived nutrition information. By identifying the nutrition information of restaurant food through crowdsourced food images and a deep learning model, the study provides a pilot approach for large-scale nutrition assessment of the community food environment.

Losses, Gains, and Changes to the Food Environment in a Rural Kentucky County during the COVID-19 Pandemic

The COVID-19 pandemic has caused alterations to be made in the way many people access, prepare, and consume food. Rural communities are particularly impacted due to pre-existing structural vulnerabilities, i.e., poverty, lack of infrastructure, and limited fresh food options. This study aimed to characterize experiences of one rural Appalachian community’s changes to the food environment during the pandemic. In April 2021, six focus groups were conducted with residents of Laurel County, Kentucky. Using grounded theory, we identified losses, gains, and overall changes to the community food environment since the onset of COVID-19. Seventeen Laurel Countians (17 female; ages 30–74) participated in the six focus groups. Three main themes emerged regarding food environment changes—(1) modifications of community food and nutrition resources, (2) expansion and utilization of online food ordering, and (3) implications of the home food environment. Rural communities faced considerable challenges during the COVID-19 pandemic, in part, due to gaps in existing infrastructure and loss of pre-existing resources. This study illustrates the complexity of changes occurring during COVID-19. Using the preliminary data obtained, we can better understand pre-existing issues in Laurel County and suggestions for future programming to address the inequitable access and response during public health emergencies and beyond.

Web Data Mining: Validity of Data from Google Earth for Food Retail Evaluation

To overcome the challenge of obtaining accurate data on community food retail, we developed an innovative tool to automatically capture food retail data from Google Earth (GE). The proposed method is relevant to non-commercial use or scholarly purposes. We aimed to test the validity of web sources data for the assessment of community food retail environment by comparison to ground-truth observations (gold standard). A secondary aim was to test whether validity differs by type of food outlet and socioeconomic status (SES). The study area included a sample of 300 census tracts stratified by SES in two of the largest cities in Brazil, Rio de Janeiro and Belo Horizonte. The GE web service was used to develop a tool for automatic acquisition of food retail data through the generation of a regular grid of points. To test its validity, this data was compared with the ground-truth data. Compared to the 856 outlets identified in 285 census tracts by the ground-truth method, the GE interface identified 731 outlets. In both cities, the GE interface scored moderate to excellent compared to the ground-truth data across all of the validity measures: sensitivity, specificity, positive predictive value, negative predictive value and accuracy (ranging from 66.3 to 100%). The validity did not differ by SES strata. Supermarkets, convenience stores and restaurants yielded better results than other store types. To our knowledge, this research is the first to investigate using GE as a tool to capture community food retail data. Our results suggest that the GE interface could be used to measure the community food environment. Validity was satisfactory for different SES areas and types of outlets.

Associations between Subjective and Objective Measures of the Community Food Environment and Executive Function in Early Childhood

The present study utilized a cross-sectional design to assess whether two indicators of the community food environment, parent perceptions of the community food environment (i.e., as assessed by parent reports of access to, availability, and affordability of foods) and limited food access (via census data), were related to executive function in preschool children. Children were recruited during the 2014–2015 academic year from Head Start and community-based preschools (N = 102) and children’s executive function ability was tested using the Head–Toes–Knees–Shoulders task. Multiple linear regression analysis was used, as well as adjusted standard errors to account for clustering at the classroom level. Parent reports of their food environment were significantly related to children’s executive function, such that children living in higher quality community food environments had better executive function. In contrast, limited food access using census data was not significantly related to executive function. The results suggest that parent reports of the community food environment in early childhood may contribute to young children’s cognitive outcomes more so than being in a limited food access area, as these data may not represent individual behaviors or capture the variability of the accessibility and affordability of healthy foods. Policy makers should consider correlations between the food environment and early executive functioning when developing new community health/wellness legislation.

Children-and-Partner's Food Preferences and Role of Woman as Caregiver: Key Elements of Food Choice Linked to Social Environment in a Consumer Population of Hermosillo City

Objectives The main objective was to identify determinants of food choice linked to the community food environment in a marginalized consumer population in the city of Hermosillo, Mexico. The specific objectives were to develop group interviews with women and to frame the data analysis within the field of research in food and nutrition security. Methods In this qualitative study design, we employed the focus group technique to ask participants “How do you decide which foods to obtain for your family's diet?”. To investigate further we asked, “What are those reasons that explain the selection of those foods?” Women who regularly attended a community center localized in a neighborhood with a very high grade of urban marginalization participated in the focus groups. Interviews were transcribed verbatim and three distinct methods were used to perform analysis: (1) content analysis (2) data organization, using software QSR NVivo, in relation to five dimensions of food and nutrition security: affordability, accessibility, acceptability, food quality, and care; and (3) triangulation between five co-authors (A.C.N., P.B.C., A.D.C.P., G.E.P.A., and M.I.O.V.). Results From May to November of 2019, four focus groups were conducted by a single facilitator in two community centers of Hermosillo (n = 27 participants). Reasons that explained the participant's food choices within the community food environment and that showed the highest number of mentions in the interviews were identified in relation to acceptability factors: children's food preferences, partner's food preferences, all household-members’ food preferences. The following extract reflects the main study findings: “When I can't do a certain thing it is because it's very expensive, but if there is a way, even if it's a little… for example, my middle-aged son really likes peppers and he eats them alone. So, I know that when I go to the store, I have to bring at least one, for him to eat other things.” Conclusions The study of food choice in this group of women denotes that their role as caregiver of food and nutrition in relation to their children-and-partner's food preferences are key elements of food decision-making processes, preceding the socioeconomic factors and constraints, they certainly face. Funding Sources Institutional small grant C.I.A.D., A.C.

Introducing the Facility List Coder: A New Dataset/Method to Evaluate Community Food Environments

Community food environments have been shown to be important determinants to explain dietary patterns. This data descriptor describes a typical dataset obtained after applying the Facility List Coder (FLC), a new tool to asses community food environments that was validated and presented. The FLC was developed in Python 3.7 combining GIS analysis with standard data techniques. It offers a low-cost, scalable, efficient, and user-friendly way to indirectly identify community nutritional environments in any context. The FLC uses the most open access information to identify the facilities (e.g., convenience food store, bar, bakery, etc.) present around a location of interest (e.g., school, hospital, or university). As a result, researchers will have a comprehensive list of facilities around any location of interest allowing the assessment of key research questions on the influence of the community food environment on different health outcomes (e.g., obesity, physical inactivity, or diet quality). The FLC can be used either as a main source of information or to complement traditional methods such as store census and official commercial lists, among others.

Validity and Reliability of the Facility List Coder, a New Tool to Evaluate Community Food Environments

A community food environment plays an essential role in explaining the healthy lifestyle patterns of its community members. However, there is a lack of compelling quantitative approaches to evaluate these environments. This study introduces and validates a new tool named the facility list coder (FLC), whose purpose is to assess food environments based on data sources and classification algorithms. Using the case of Mataró (Spain), we randomly selected 301 grids areas (100 m2), in which we conducted street audits in order to physically identify all the facilities by name, address, and type. Then, audit-identified facilities were matched with those automatically-identified and were classified using the FLC to determine its quality. Our results suggest that automatically-identified and audit-identified food environments have a high level of agreement. The intra-class correlation coefficient (ICC) estimates and their respective 95% confidence intervals for the overall sample yield the result “excellent” (ICC ≥ 0.9) for the level of reliability of the FLC.