Source: The Social & Health Research Center submitted to NRP
DIGITAL FOOD IMAGING ANALYSIS: EVALUATION RESEARCH OF FOOD CONSUMPTION & NUTRIENT ANALYSIS FOR SCHOOL NUTRITION & CHILDHOOD OBESITY PROGRAMS
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
COMPLETE
Funding Source
AFRI COMPETITIVE GRANT
Reporting Frequency
Annual
Accession No.
0224222
Grant No.
2011-67001-30071
Cumulative Award Amt.
$2,000,000.00
Proposal No.
2010-04578
Multistate No.
(N/A)
Project Start Date
Feb 1, 2011
Project End Date
Jan 31, 2016
Grant Year
2011
Program Code
[A2131]- Childhood Obesity Prevention: Methodological Research to Assess the Effectiveness of Obesity Prevention Strategies
Recipient Organization
The Social & Health Research Center
1302 South St. Mary's Street
San Antonio,TX 78210
Performing Department
(N/A)
Non Technical Summary
Obesity is a leading public health concern and is now considered to be the most prevalent nutritional disease of children in the United States. The prevalence of obesity has increased dramatically among US children. Excessive weight during childhood stems from interacting factors including poor diet and physical activity. Consequently, behavioral modifications associated with obesity prevention approaches use the combination of increased physical activity and nutritional assessment. Nutrient analysis is a significant component of obesity research in children because of its predictive ability and function as a weighted cofactor in clinical investigations. However, current dietary tools (24-hr dietary recall, food records, and food frequency questionnaire) are limited when adapted for children. Therefore, the need for a reliable and cost effective dietary tool designed for young children in school settings exists. This study proposal is aimed at developing an innovative instrument in estimating dietary intake. The proposed instrument is the Digital Food Imaging Analysis Technique (DFIA). The DFIA technique was developed using the NIH-Image J analysis software and the NHANES database for nutrient analysis. DFIA provides a qualitative estimation of young children's food consumption and nutrient analysis. DFIA will be able to capture significant micro and macro nutritional changes in children. Consequently, nutritional data obtained from DFIA is useful in determining the effectiveness of obesity prevention and nutritional programs in elementary school systems.
Animal Health Component
100%
Research Effort Categories
Basic
(N/A)
Applied
100%
Developmental
(N/A)
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
7015010101020%
7036010101080%
Knowledge Area
703 - Nutrition Education and Behavior; 701 - Nutrient Composition of Food;

Subject Of Investigation
6010 - Individuals; 5010 - Food;

Field Of Science
1010 - Nutrition and metabolism;
Goals / Objectives
1.To develop a Digital Food Imaging Analysis (DFIA) system: A nutritional quantitative assessment tool in estimating young children's food consumption and nutrient intake. 2.To apply the developed DFIA in evaluating the effectiveness of an evidence-based Bienestar/NEEMA obesity health program as implemented in the schools with children three to eight years of age. 3.To evaluate DFIA's applicability in measuring nutritional change moderated by the following USDA school-based nutritional programs under the scope of obesity prevention: a.National School Lunch Program (NSLP) b.Fresh Fruit & Vegetable Program (FFVP) c.School Breakfast Program (SBP) d.Expanded Food and Nutrition Education Program (EFNEP)
Project Methods
The research activities proposed in this study are twofold. First, to develop an innovative nutritional assessment tool (DFIA) that can be adapted for pediatric research. Second, the developed DFIA tool will then be applied to test the effectiveness of a school-based obesity program (Bienestar/NEEMA) and replicate its capability in evaluating a list of USDA's nutritional programs. Phase 1: Development of the Digital Food Image Analysis (DFIA) tool in school settings. A. DFIA technical development - In this sub-phase several technical procedures will be developed to establish a standard procedure (repeatability) in capturing digital food images of children's meals. Digital photos of children's meals will be taken before and after they are consumed. The photos will be analyzed using NIH-Image J software to provide amount and type of food consumed by children. Information and measured parameters from the image analysis will then be analyzed for macro and micronutrient content using the NHANES reference nutrient database. B. Sensitivity and Specificity Studies - In this sub-phase the work plans are focused on determining the sensitivity and specificity of the DFIA. These procedures are to be designed in estimating micro and macronutrients of various food items commonly consumed in school cafeterias. C. Validity and Reliability Studies - In this sub-phase the work plans are focused on testing the reliability and validity of the DFIA against well-established nutritional assessment tools such as the 24-H Dietary Recall (NDS-R) and the Food Frequency Questionnaire (FFQ). D. Cost Analysis - In this sub-phase the work plans are focused on establishing the cost associated with use of DFIA. DFIA is believed to be a more flexible and less expensive tool compared to the current standard of food nutrient analysis techniques (i.e. 24-H DR, FFQ). Phase 2: Evaluation of DFIA in assessing the Bienestar/NEEMA obesity program and USDA nutritional programs. A. Feasibility Study - In this sub-phase the work plans are focused on determining the applicability of the developed DFIA technique in elementary school settings across 5 grade levels (Pre-K - 3rd). Parents will be asked to take pictures of their child's meals at home or at restaurants before and after the meals are consumed. B. Effectiveness Studies - In this sub-phase the work plans are focused on capturing the capability of the DFIA in measuring behavioral change moderated by a school-based obesity intervention program (Bienestar/NEEMA obesity program). Additionally, in this sub-phase the work plans are also intended to capture the school procedures in implementing USDA's nutritional programs. Therefore, the DFIA will be used to establish the effectiveness of Obesity Behavioral Interventions and USDA programs in reducing nutritional risk factors.

Progress 02/01/11 to 01/31/16

Outputs
Target Audience:Children ages 4 - 9 years Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided? Nothing Reported How have the results been disseminated to communities of interest?See publications What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? Goal 1: DFIA uses digital photography to identify food; new coding system (DFIA number) of food items to elaborate on how food is cooked, prepared, processed, spiced and served; software to enter, code (DFIA number) and approximate new recipes and ingredients not found in the USDA National Nutrient Database for StandardReference (NDB SR) to food items that are found in the NDB SR; before and after meal photographs to measure consumption; software and database to determine the amount of nutrient consumed by linking DFIA number of food in photographs to food in the NDB SR. The DFIA is an expeditious and inexpensive method of quantifying value and amount of food consumed that might improve population-based surveillance of dietary intake and evaluation of health programs implementation. Goal 2: Food weight is the fundamental unit fornutrient analysis. The challenge in assessing dietary intakeis collecting accurate food volumes and converting theseto food weight. The 24-HDR uses memory recalls supportedby visual guides (e.g. deck of cards, baseball, food picturesand models) to estimate food volume and then convert toweight. As a result, the estimation of food weight is susceptible to measurement errors. Because of these limitations,the DFIA technique was developed to improve accuracy inestimating food weight. A total of 120 students (ages 8-9years) from two schools with parent consent andcomplete data were included in this analysis (n¼120 meals;57 breakfast + 63 lunch). A 24-HDR was collected thefollowing day and DFIA was collected the same day of theanalysis. Food weight for 24-HDR was estimated by NutritionData System for Research software (NDSR; version4.04), and for DFIA it was estimated by demarcating surfacearea of food items in photographs before and after consumption,and converted to weights using image analysis properties. For reference, meals were weighed with an electronicscale at the school cafeteria cash register and again atthe disposal window to calculate food weight consumption.Validity was examined by comparing real food weight andthe two dietary intake instruments.A comparison between the estimated foodweight by NDSR and DFIA to real food weight wasdone. The regression model shows that DFIA had a betterfit than NDSR with R square¼ 0.351 > 0.088, respectively.The Pearson's correlation also displayed a strongercorrelation between DFIA and real weight (0.592) thanNDSR and real weight (0.297). Lastly the Intraclass Correlation Coefficient for single measures showed thatDFIA had a reliability of 59% compared with 30% forNDSR; and for average measures DFIA had a reliabilityof 74 % compared to 46% for NDSR. These findings supportthat the DFIA is more accurate than NDSR to estimate food weights. Goal 3: The DFIA was used to collect and analyze over 100,000 school lunches and breakfasts during the school year 2011-2012. The study was conducted in 61 elementary schools nested in 38 school districts across 5 states (New York, Arkansas, Iowa, Texas and Washington). Results of this analysis was presented in the 2013 progress report.

Publications

  • Type: Journal Articles Status: Published Year Published: 2015 Citation: Trevi�o R, Ravelo A, Birkenfeld E, Murad M, Diaz J. Food Weight Estimation: A comparative analysis of digital food imaging analysis and 24-hour dietary recall. Journal of Nutrition Education and Behavior  Volume 47, Number 4S: S105
  • Type: Journal Articles Status: Submitted Year Published: 2016 Citation: Trevi�o R, Ravelo A, Senne-Duff B, Murad M, Diaz J, Birkenfeld E. Poor Validity of Dietary Recall in Low-Income Hispanic Children using Digital Food Imaging Analysis as the Reference. Journal Pediatrics and Child Nutrition


Progress 02/01/13 to 01/31/14

Outputs
Target Audience: Children ages 4-9 years Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided? Nothing Reported How have the results been disseminated to communities of interest? Nothing Reported What do you plan to do during the next reporting period to accomplish the goals? Completion of the project

Impacts
What was accomplished under these goals? Efforts under this period were utilized to accomplish the project’s objectives1, 2 and 3. Objective 1 is the development and testing of the Digital Food Imaging Analysis (DFIA) system. DFIA systems of technologies were completed in year 2 (2012-2013). However, for this period, objective 1 was focused on the completion of data collection for validity and reliability study and the initiation of data analysis for reporting. In this sub-phase the work plans were focused on testing the DFIA against two traditional nutritional assessment tools: (1) 24-hr dietary recall with NDSR and (2) the food frequency questionnaire (FFQ). A total of 405 students and their parents participated in the reliability/validity studies. Of the 405 students, 208 students had complete 3-day dietary data available for this application. The data analyses for this sub-phase are to establish the validity and reliability of DFIA using the following analyses: Pearson Correlation (Test-Rest) Inter-rater Reliability Level of Agreement using Bland-Altman method Objective 2 is to apply the developed DFIA in evaluating the effectiveness of the evidence-based Bienestar/NEEMA obesity health program as implemented in the schools with children four to nine years of age. In this sub-phase the work plans were utilized in recruiting students to participate in the Bienestar/NEEMA obesity health program. To date, 522 students from 6 elementary schools in grade levels Pre-K to 4th were consented / assented for the evaluation phase of the study. For the evaluation phase, the student’s baseline characteristics were collected during the fall of the school year 2013-2014 prior to the school implementation of the Bienestar/NEEMA obesity health program. The collected parameters are the following: Height and weight for BMI Physical Fitness level using 20-meter shuttle run Health Knowledge, and School lunch meals Objective 3 is the application of DFIA technologies in evaluating the USDA’s national school lunch (NSLP) and breakfast programs (SBP). In this sub-phase the work plans were utilized to collect and analyze over 120,000 school lunches and breakfasts during the school year 2012-2013. The study was conducted in 72 elementary schools nested in 49 school districts across 6 states (New York, Arkansas, Iowa, Texas, Washington, and Utah). The school year 2012-2013 will serve as one year post implementation of the HHFKA school nutrition standards.

Publications

  • Type: Conference Papers and Presentations Status: Published Year Published: 2013 Citation: Echon, RM (2013) Evaluation of USDAs school lunch and breakfast programs using Digital Food Image Analysis. Annual Meeting for the Society for Nutrition Education and Behavior, Portland, OR


Progress 02/01/12 to 01/31/13

Outputs
OUTPUTS: Efforts under this period were utilized to accomplish the projects objective 1 and 3. Objective 1 is the development and testing of the Digital Food Imaging Analysis (DFIA) system. For this period, objective 1 is subdivided into three separate goals: a.DFIA - Completion of DFIA systems of technologies. b.Sensitivity and Specificity Studies - In this sub-phase, the work plans are focused on the sensitivity and specificity of the DFIA. Sets of referential photographs for all food items were collected to calculate surface areas and volume for all food items to be analyzed and archived in the DFIA database management system. c.Validity and Reliability Studies - Reliability was examined by test-retest using Pearson correlation coefficient. Three days of DFIA, 24HDR, and FFQ were collected twice, either 2 or 3 weeks apart. The data collection was coordinated with school district food service administrators to assure that the same menu was served on the 2 data collection periods. It was arranged with 1 school district to serve the same menu every 2 weeks and with 2 school districts every 3 weeks. A similar arrangement to that in the school districts was coordinated with parents for home meals. Concurrent validity was examined by comparing DFIA with 24HDR with NDSR and FFQ using Pearson correlation. Three days of DFIA, 24HDR, and FFQ were collected in the same day and time. Objective 3 is the application of DFIA technologies in evaluating the USDAs national school lunch (NSLP) and breakfast programs (SBP). The DFIA was used to collect and analyze over 100,000 school lunches and breakfasts during the school year 2011-2012. The study was conducted in 61 elementary schools nested in 38 school districts across 5 states (New York, Arkansas, Iowa, Texas and Washington). A preliminary data analysis was conducted on NSLP in alignment with the USDAs new school nutrition standards as authorized by the Healthy, Hunger-Free Kids Act of 2010 (HHFKA). The analysis was focused on meal composition of school lunch serving and dietary specification of food items served. Events: Echon, RM (2012) Digital Food Image Analysis: An application of adaptive neural-network protocols to resolve system (bias) and random errors in nutrient analysis. Annual Meeting for the Society for Nutrition Education and Behavior, Washington, DC Wells NM. Myers, EM, Echon, RM, Wilkins JL, & Militello R (2013) Lunch tray photography: Measuring childrens dietary intake at school. Annual Environmental Research Design Symposium, Chicago, IL. Services: Continued collaboration with Cornell University and Washington State Universitys USDAs funded (FNS Fed. Agreement: CN-CGP-11-0047) project entitled: "Healthy Gardens, Healthy Youth". Continued collaboration with the Baylor University DFIA evaluation services for the Texas Huger Initiatives PARTICIPANTS: 1.Roberto P. Trevino, MD - Principal Investigator 2.Ginger Mullaney - Project Manager 3.Eri Birkenfeld - Data Manager and Associate Programmer 4.JudyAnn Gonzales - Project Parent Coordinator 5.Albert Salinas - Technology Assistant 6.Chuya Pierce - Research Associate 7.Jeff Wilwon - Nutritionist 8. Katy Jechow- Registered Dieticians 9. Roger Echon - Co-Investigator, DFIA Technology Developer, Project Engineer, Lead Programmer& Project Statistician. TARGET AUDIENCES: Elementary School Children ages 6-10 years PROJECT MODIFICATIONS: Not relevant to this project.

Impacts
Change in Knowledge - Validity and Reliability Studies - A total of 200 students and their parents participated in the reliability/validity studies. Of the 200 students, 112 students had complete 3-day dietary data available for this progress report. Their mean age was 9.18+/-0.78 years. The mean energy intake by DFIA method at test and retest were 1224+/-380 and 1195+/-323, respectively. Pearson correlation was .70 (p<.001). The mean energy intake by 24HDR was 1023+/-456 and 1123+/-448, respectively (.34; p=.008) and for FFQ it was 1560+/-1322 and 907+/-542, respectively (.24; p=.064). The mean energy intakes by DFIA and 24HDR methods were 1224+/-380 and 1023+/-456, respectively. Pearson correlation was .46 (p<.001). The mean energy intakes by DFIA and FFQ were 1023+/-456 and 1560+/-1322, respectively (.17; p=.10) and by 24HDR and FFQ they were 1023+/-456 and 1560+/-1322, respectively (.06; p=.33). Application study - The meal pattern of serving from the evaluated elementary schools (national average) are as follows: For fruits 105.76+/-85.76 g or 0.42 cup is served daily with 43% of the students meeting the USDAs recommended half cup. For vegetables, 55.24+/-49.46 g or 0.22 cup is served daily with 16% of the students meeting the USDAs recommended one third cup. For grains 36.62+/-45.43 g or 1.08 oz is served daily with 39% of the students meeting the USDAs recommended 1 oz. For meat or meat alternative 56.76+/-59.02 g or 1.7 oz is served daily with 56% of the students meeting the USDAs recommended 1 oz. For milk 207.73+/- 93.51 g or 0.83 cup are served daily with 81% of the students meeting the USDAs recommended 1 cup. Lastly, for calorie tracking 468.72+/-178.54 Kcal are offered daily to students from 4.32+/-.87 food items with 42% of the students meeting the USDAs recommended 550-650 Kcal for this age group. In terms of nutrient compositions, the average daily energy intake is 419.56+/-.186.73 kcal per student (n=8,168). The average daily salt (Na) intake is 658.57+/-.425.68 mg per student (n=8,168). The average daily saturated fat intake is 5.10+/-.5.33 g per student (n=8,168). Change in Actions - DFIA is an innovative nutrient analysis that addresses random & system errors. System errors are mathematical errors resulting in over/underestimation of nutrient values. Random errors are operational errors due selection. Selection errors stem from the enormity of food items listed in sr24 database. Results from the sensitivity/specificity study clearly demonstrated the significance of system & random errors (76.6% of total variation). DFIA technology can eliminate human factor errors in food item selection by up to 99.5 %. Additionally, DFIA also negates system error (nutrient estimation). These two actions are significant contributions to the field of nutrition, if not a step toward system accuracy and specificity in unifying the field of nutritional assessments.

Publications

  • Echon, RM (2013). Random and system errors in nutrient analysis: An application of adaptive neural-network protocols,Journal of Food Engineering, 116(1):218-232.


Progress 02/01/11 to 01/31/12

Outputs
OUTPUTS: Development of technical procedures to capture food images: (a)image capture system with streaming video(2), recording(2,8,17) & web-based application to upload photo via DFIA web site(2,16), (b)engineering & instrumentation of IP cameras(2,8), (c)data acquisition with wireless transceiver & receiver capability(2), (d)video/audio data overlay(2), optical data isolation, signal conditioning, & amplification(2), (e)CAT-5 wiring of cafeterias(8,17,14,15), (f)IT management (LAN/ WAN)(2,8,14,15,17), & (g)database management using LabView(2), Access(2,3), SQL(2,3), & iNET(16). Development of technical procedures to analyze food images using DFIA's adaptive neural network protocols: (a)menu planner(3), (b)menu analysis with linkage to sr24(2,9), (c)pattern recognition algorithms(2), (d)image artifact analysis(2), (e)RGB-color spectrum, classification, extraction, & threshold(2), (f)densitometry(2), (g)edge detection(2), (h)measurement of surface area/volume(2), & (i)data matrix reader(2). Sensitivity and Specificity Studies: In this sub-phase the work plans are focused on determining the sensitivity and specificity of the DFIA. These procedures are composed of (a)generation of unique identifiers for sr24 linkage(3), (b)generating data matrix for menu analysis, subject de-identifier, school menu reference & manufacturer/UPC code(7,8), (c)collection & archiving of 4 sets of reference photographs with top views to represent 100, 75, 50 & 25% of surface areas (length X width)(3,6,8,10), side views to represent volume (surface area X height) (3,6,8,10), (d)4 matching weight sets(6,8,10). Validity and Reliability Study: In this sub-phase the work plans are focused on testing the DFIA against well-established nutritional assessment tools such as the 24-hr dietary recall and the food frequency questionnaire (FFQ). These procedures are composed of (a)school district coordination and communication for study approval and formalization of roles(1), (b)recruitment, data collection, & training(7), (c)parent coordination and communication for data collection at home or outside of schools(6), (d)training of parents to use DFIA website(6), (e)student/school record maintenance(10), (f)data collection of school meal photos(2,3,8,10), home meal photos(6), 24-H Dietary Recall(24-HDR)(4,5,7,12,13,18), and FFQ(4,5,7,12,13,18), (g)Nutrient Data System (NDS-R) data entry and analysis(5,12,13), (h)school food record menu analysis(4,5,9), (i)home food record menu analysis(2), (j)DFIA nutrient analysis(2), and (k)triangulation of DFIA, NDS-R and FFQ(2). Events: LabView training for objectives 1.a-c(2,3,5,7,8,10), & training on 24-HDR for validity and reliability objectives(5,11). Services: Provided consulting services to George Mason University to use DFIA for NIH grant application titled "School lunch waste project(2)". DFIA evaluation services for Cornell and Washington State University's USDA-funded (FNS Fed. Agreement: CN-CGP-11-0047) project titled: "Healthy Gardens, Healthy Youth(2)". Products: DFIA Software, Algorithms, & DFIA Website: www.dfia.sahrc.org Dissemination: Provided interviews with local, state, national and international media to disseminate DFIA(1,2). PARTICIPANTS: Participants: Project roles are referenced (parentheses) to accomplish the above objectives: 1.Roberto P. Trevino, MD - PI 2.Roger M. Echon, PhD - Co-I, DFIA technology developer, Proj. Engineer, Lead Programmer & Statistician. 3.Eri Birkenfeld, BS - Data Manager & Ass. Programmer 4.Julian Trevino, MPH - Proj. Manager 5.Elisa Garcia, BS - Proj. Nutritionist 6.JudyAnn Gonzales, MS - Parent Coordinator 7.Ginger Mullaney, BA - School Coordinator 8.Albert Salinas - Technology Assist. 9.Tricia Garcia, MPH - Research Assist. 10.Chuya Pierce - Research Assist. Collaborators & Contracts: 11.Liset Leal-Vasquez,MS - Staff Nutritionist 12.University of Texas Health Science Center: Dr. Cunningham, Nutrition Dept. 13.University of Incarnate Word: Dr. Elizabeth Senny-Duff, Nutrition Dept 14.San Antonio ISD: Sally Cody Nutrition Dept & Dago Garcia, IT Dept 15.Edgewood ISD: Janice Phelps, Nutrition Dept & Juan Gonzales, IT Dept 16.Denim Group: J Gonzales, B Mather, J Harris & S Chamber 17.Computer Kinetics: Keith Lytle TARGET AUDIENCES: Children 4-8 years of age. PROJECT MODIFICATIONS: Because parents were not downloading pictures correctly it resulted in significant delay & impacted on the rate of grant expenditure. Corrective measures are better training of parents & staff to pick up cameras. Thus, work plans under this phase will overlap (Phase II) for next funding year.

Impacts
Change in Knowledge - DFIA is an innovative nutrient analysis that addresses random & system errors. System errors are mathematical errors resulting in over/underestimation of nutrient values. Random errors are operational errors due to selection. Selection errors stem from the enormity of food items listed in sr24 database. Change in Actions - Results from the sensitivity/specificity study clearly demonstrated the significance of system & random errors (76.6 percent of total variation). DFIA technology can eliminate human factor errors in food item selection by up to 99.5 percent. Additionally, DFIA also negates system error (nutrient estimation). These two actions are significant contributions to the field of nutrition, if not a step toward system accuracy and specificity in unifying the field of nutritional assessments.

Publications

  • Echon, RM. 2012. Digital Food Image Analysis: An application of adaptive neural-network protocols to resolve system (bias) and random errors in nutrient analysis. Annual Meeting for the Society for Nutrition, Education and Behavior, Washington, DC.
  • KSAT 12 News, 2011. ABC News local affiliate in San Antonio, Texas http://www.ksat.com/newsarchive/27859182/detail.html
  • WOAI News 4, 2011. NBC News local affiliate in San Antonio, Texas - http://www.woai.com/content/health/story/School-spending-2-Million-do llars-to-track-what/u5zRCFhVUkiwpX70HuXCCg.cspx
  • MSNBC, 2011. http://www.msnbc.msn.com/id/42994465/ns/health-diet_and_nutrition/t/s mile-kids-youre-calorie-camera/
  • Fox News, 2011. http://www.foxnews.com/us/2011/05/11/texas-schools-pictures-worth-100 0-calories/
  • The Today Show, 2011. http://www.todayonline.com/World/EDC110514-0000248/Smile,-US-students ,-youre-on-calorie-camera
  • CBC News, 2011. http://www.cbc.ca/news/health/story/2011/05/12/school-lunch-trays-pho tograph.html
  • CBN News, 2011, https://www.cbn.com/cbnnews/healthscience/2011/May/Calorie-Camera-New -Childhood-Anti-Obesity-Tool
  • KCRW 89.9 FM, 2011. Good Food with Evan Kleiman,Includes interview with Dr. Roger Echon on details how the San Antonio School District is combating obesity with a video camera. http://www.kcrw.com/etc/programs/gf/gf111022frito_pie_gut_instin
  • Agritalk, 2011. The Voice of Rural America,San Antonio schools photographing school lunches. Includes interview with Dr. Roger Echon http://www.agritalk.com/podcast/p=episode&name=2011-05-17_may_1711.m p3
  • Associated Press, 2011. In Texas schools, a picture worth 1000 calories by Paul Weber includes interview with Dr. Echon http://www.deseretnews.com/article/700134524/In-Texas-schools-a-pictu res-worth-1000-calories.html
  • San Antonio Express-News, 2011. Skipping those veggies The Cafeteria lady knows by Don Finley,Front page article includes interview with Dr. Roberto Trevino www.mysanantonio.com/news/local_news/article/Skipping-those-veggies-T he-cafeteria-lady-knows-1376032.php
  • Washington Times, 2011. Front page article includes interview with Dr. Roger Echon by Tom Reel http://www.washingtontimes.com/news/2011/may/11/in-texas-schools-a-pi ctures-worth-1000-calories/
  • San Angelo News, 2011. http://www.gosanangelo.com/photos/2011/may/11/46746/
  • Huffington Post, 2011. http://www.huffingtonpost.com/2011/05/11/texas-calorie-camera_n_86077 1.html
  • Whittaker, Nicholas, 2011. Fighting obesity has become high tech in San Antonio schools. www.imperfectparent.com/topics/2011/05/12/fighting-obesity-has-become -high-tech-in-san-antonio-schools/
  • www.sanantonio.yalwa.com, 2011. http://sanantonio.yalwa.com/ID_100640869/Associated-Press-News.html
  • www.sglinks.com, 2011. http://sglinks.com/pages/676526-world-smile-us-students-re-calorie-ca mera
  • www.dailymail.co.uk, 2011. Mail Online- http://www.dailymail.co.uk/health/article-1386177/Texas-school-pilots -nationwide-project-photograph-students-eating-lunch.html