1
|
Chen G, Jia W, Zhao Y, Mao ZH, Lo B, Anderson AK, Frost G, Jobarteh ML, McCrory MA, Sazonov E, Steiner-Asiedu M, Ansong RS, Baranowski T, Burke L, Sun M. Food/Non-Food Classification of Real-Life Egocentric Images in Low- and Middle-Income Countries Based on Image Tagging Features. Front Artif Intell 2021; 4:644712. [PMID: 33870184 PMCID: PMC8047062 DOI: 10.3389/frai.2021.644712] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/21/2020] [Accepted: 02/26/2021] [Indexed: 11/25/2022] Open
Abstract
Malnutrition, including both undernutrition and obesity, is a significant problem in low- and middle-income countries (LMICs). In order to study malnutrition and develop effective intervention strategies, it is crucial to evaluate nutritional status in LMICs at the individual, household, and community levels. In a multinational research project supported by the Bill & Melinda Gates Foundation, we have been using a wearable technology to conduct objective dietary assessment in sub-Saharan Africa. Our assessment includes multiple diet-related activities in urban and rural families, including food sources (e.g., shopping, harvesting, and gathering), preservation/storage, preparation, cooking, and consumption (e.g., portion size and nutrition analysis). Our wearable device ("eButton" worn on the chest) acquires real-life images automatically during wake hours at preset time intervals. The recorded images, in amounts of tens of thousands per day, are post-processed to obtain the information of interest. Although we expect future Artificial Intelligence (AI) technology to extract the information automatically, at present we utilize AI to separate the acquired images into two binary classes: images with (Class 1) and without (Class 0) edible items. As a result, researchers need only to study Class-1 images, reducing their workload significantly. In this paper, we present a composite machine learning method to perform this classification, meeting the specific challenges of high complexity and diversity in the real-world LMIC data. Our method consists of a deep neural network (DNN) and a shallow learning network (SLN) connected by a novel probabilistic network interface layer. After presenting the details of our method, an image dataset acquired from Ghana is utilized to train and evaluate the machine learning system. Our comparative experiment indicates that the new composite method performs better than the conventional deep learning method assessed by integrated measures of sensitivity, specificity, and burden index, as indicated by the Receiver Operating Characteristic (ROC) curve.
Collapse
Affiliation(s)
- Guangzong Chen
- Department of Electrical and Computer Engineering, University of Pittsburgh, PA, United States
| | - Wenyan Jia
- Department of Electrical and Computer Engineering, University of Pittsburgh, PA, United States
| | - Yifan Zhao
- Department of Electrical and Computer Engineering, University of Pittsburgh, PA, United States
| | - Zhi-Hong Mao
- Department of Electrical and Computer Engineering, University of Pittsburgh, PA, United States
| | - Benny Lo
- Hamlyn Centre, Imperial College London, London, United Kingdom
| | - Alex K. Anderson
- Department of Foods and Nutrition, University of Georgia, Athens, GA, United States
| | - Gary Frost
- Section for Nutrition Research, Department of Metabolism, Digestion and Reproduction, Imperial College London, London, United Kingdom
| | - Modou L. Jobarteh
- Section for Nutrition Research, Department of Metabolism, Digestion and Reproduction, Imperial College London, London, United Kingdom
| | - Megan A. McCrory
- Department of Health Sciences, Boston University, Boston, MA, United States
| | - Edward Sazonov
- Department of Electrical and Computer Engineering, University of Alabama, Tuscaloosa, AL, United States
| | | | - Richard S. Ansong
- Department of Nutrition and Food Science, University of Ghana, Legon-Accra, Ghana
| | - Thomas Baranowski
- USDA/ARS Children's Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, TX, United States
| | - Lora Burke
- School of Nursing, University of Pittsburgh, Pittsburgh, PA, United States
| | - Mingui Sun
- Department of Electrical and Computer Engineering, University of Pittsburgh, PA, United States
- Department of Neurosurgery, University of Pittsburgh, Pittsburgh, PA, United States
| |
Collapse
|
2
|
Šumonja S, Jevtić M. Accuracy of reported food intake in a sample of 7-10 year-old children in Serbia. Public Health 2016; 138:63-8. [PMID: 27087636 DOI: 10.1016/j.puhe.2016.03.016] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/21/2014] [Revised: 08/04/2015] [Accepted: 03/15/2016] [Indexed: 11/18/2022]
Abstract
OBJECTIVES Children's ability to recall and report dietary intake is affected by age and cognitive skills. Dietary intake reporting accuracy in children is associated with age, weight status, cognitive, behavioural, social factors and dietary assessment techniques. This study analysed accuracy of 7-10 year-old children's reported food intake for one day. STUDY DESIGN Validation study. METHODS Sample included 94 children aged 7-10 years (median = 9 years) from two elementary schools in a local community in Serbia. 'My meals for one day' questionnaire was a combination of 24-h recall and food recognition form. It included recalls for five meals: breakfast at home; snack at home; lunch at home; snack at school and dinner at home. Parental reports were used as reference information about children's food intake for meals obtained at home and observation was used to gain reference information for school meal. Observed and reported amounts were used to calculate omission rate, intrusion rate, corresponding, over-reported and unreported amounts of energy, correspondence rate and inflation ratio. RESULTS Overall omission rate (37.5%) was higher than overall intrusion rate (36.7%). The same food item (bread) has been the most often correctly reported and omitted food item for breakfast, lunch and dinner. Snack at school had the greatest mean correspondence rate (79.6%) and snack at home the highest mean inflation ratio (90.7%). CONCLUSION Most errors in children's recalls were incorrectly reported amounts and not the food items. The questionnaire should be improved to facilitate accurate reports of the amounts.
Collapse
Affiliation(s)
- S Šumonja
- University of Novi Sad, Faculty of Education, Department of Natural Sciences, Serbia.
| | - M Jevtić
- University of Novi Sad, Faculty of Medicine, Department of Hygiene, Serbia
| |
Collapse
|
3
|
Factors Related to the Accuracy of Self-Reported Dietary Intake of Children Aged 6 to 12 Years Elicited with Interviews: A Systematic Review. J Acad Nutr Diet 2016; 116:76-114. [DOI: 10.1016/j.jand.2015.08.024] [Citation(s) in RCA: 34] [Impact Index Per Article: 4.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/02/2014] [Accepted: 08/27/2015] [Indexed: 12/19/2022]
|
4
|
Madowitz J, Liang J, Peterson CB, Rydell S, Zucker NL, Tanofsky-Kraff M, Harnack L, Boutelle KN. Concurrent and convergent validity of the eating in the absence of hunger questionnaire and behavioral paradigm in overweight children. Int J Eat Disord 2014; 47:287-95. [PMID: 24186043 PMCID: PMC4104791 DOI: 10.1002/eat.22213] [Citation(s) in RCA: 20] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Accepted: 09/25/2013] [Indexed: 12/18/2022]
Abstract
OBJECTIVE This study aimed to assess the concurrent and convergent validity of the Eating in the Absence of Hunger (EAH) questionnaire parent report of child (EAH-PC) and child self-report (EAH-C) with the EAH behavioral paradigm (EAH%) and usual dietary intake. METHOD Data were obtained at baseline assessment for 117 treatment-seeking overweight and obese (BMI > 85th percentile) 8- to 12-year old children (53% female, 54% white) and their parents. Children participated in the EAH free access paradigm after a standardized ad libitum meal. Parents and children completed EAH questionnaires, and the children completed three 24 h recalls. EAH External Eating subscale and total scores were assessed. RESULTS EAH% was inversely associated with the EAH-PC total score (p < .04), however, it was not associated with the EAH-PC External Eating scale, EAH-C total score or EAH-C External Eating scale. Daily caloric intake was positively related to both the EAH-C total score (p < .02) and External Eating subscale (p < .007). Daily caloric intake was inversely related to EAH-PC total score (p < .05), but was not related to EAH-PC External Eating subscale or EAH%. DISCUSSION Concurrent validity was not supported for EAH questionnaires, but convergent validity was supported for EAH-C and child daily caloric intake. Further research is warranted to assess whether EAH questionnaires and paradigm are measuring different aspects of EAH in treatment-seeking children.
Collapse
Affiliation(s)
- Jennifer Madowitz
- San Diego State University/University of California, San Diego Joint Doctoral Program in Clinical Psychology, San Diego, California,Correspondence to: Jennifer Madowitz; San Diego State University/ University of California, San Diego Joint Doctoral Program in Clinical Psychology, San Diego, California.
| | - June Liang
- Department of Pediatrics, University of California, San Diego, La Jolla, California
| | - Carol B. Peterson
- Department of Psychiatry, University of Minnesota, Minneapolis, Minnesota
| | - Sarah Rydell
- Department of Public Health, University of Minnesota, Minneapolis, Minnesota
| | - Nancy L. Zucker
- Department of Psychiatry, Duke University, Durham, North Carolina
| | - Marian Tanofsky-Kraff
- Department of Medical and Clinical Psychology Uniformed Services University of the Health Sciences (USUHS), Bethesda, Maryland
| | - Lisa Harnack
- Department of Public Health, University of Minnesota, Minneapolis, Minnesota
| | - Kerri N. Boutelle
- San Diego State University/University of California, San Diego Joint Doctoral Program in Clinical Psychology, San Diego, California,Department of Pediatrics, University of California, San Diego, La Jolla, California,Department of Psychiatry, University of California, La Jolla, California
| |
Collapse
|
5
|
Baxter SD, Paxton-Aiken AE, Tebbs JM, Royer JA, Guinn CH, Finney CJ. Secondary analyses of data from 4 studies with fourth-grade children show that sex, race, amounts eaten of standardized portions, and energy content given in trades explain the positive relationship between body mass index and energy intake at school-provided meals. Nutr Res 2012; 32:659-68. [PMID: 23084638 PMCID: PMC3479430 DOI: 10.1016/j.nutres.2012.07.001] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/08/2012] [Revised: 06/28/2012] [Accepted: 07/05/2012] [Indexed: 12/28/2022]
Abstract
Results from a 2012 article showed a positive relationship between children's body mass index (BMI) and energy intake at school-provided meals. To help explain that positive relationship, secondary analyses investigated (1) whether the relationship differed by sex and race and (2) the relationship between BMI and 6 aspects of school-provided meals--amounts eaten of standardized portions, energy content given in trades, energy intake received in trades, energy intake from flavored milk, energy intake from a la carte ice cream, and breakfast type. Data were from 4 studies conducted 1 per school year (1999-2000 to 2002-2003). Fourth-grade children (n = 328; 50% female; 54% black) from 13 schools total were observed eating school-provided breakfast and lunch on 1 to 3 days per child for 1178 total meals (50% breakfast). Children were weighed and measured. Marginal regression models were fit using BMI as the dependent variable. For purpose 1, independent variables were energy intake at school-provided meals, sex, race, age, and study; additional models included interaction terms involving energy intake and sex/race. For purpose 2, independent variables were the 6 aspects of school-provided meals, sex, race, age, and study. The relationship between BMI and energy intake at school-provided meals differed by sex (P < .0001; stronger for females) and race (P = .0063; stronger for black children). BMI was positively related to amounts eaten of standardized portions (P < .0001) and negatively related to energy content given in trades (P = .0052). Explaining the positive relationship between BMI and energy intake at school-provided meals may contribute to school-based obesity prevention efforts.
Collapse
Affiliation(s)
- Suzanne Domel Baxter
- Institute for Families in Society, University of South Carolina, Columbia, SC 29208, USA.
| | | | | | | | | | | |
Collapse
|
6
|
O’Neil CE, Zanovec M, Nicklas TA, Cho SS. Presweetened and Nonpresweetened Ready-to-Eat Cereals at Breakfast Are Associated With Improved Nutrient Intake but Not With Increased Body Weight of Children and Adolescents. Am J Lifestyle Med 2011. [DOI: 10.1177/1559827610395476] [Citation(s) in RCA: 7] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/16/2022] Open
Abstract
This study compared nutrient intake and body weight measures in children 4 to 8 (n = 1480) and 9 to 13 (n = 1790), and adolescents 14 to 18 years of age (n = 1826) participating in 1999-2002 NHANES and consuming a presweetened ready-to-eat cereal (PSRTEC), a non-PSRTEC (NPSRTEC), or other breakfast (OB). Sample weights were applied, and logistic regression identified differences among breakfast consumption groups controlling for appropriate covariates. Breakfast type varied by the above-mentioned age groups: 34% (n = 507), 33% (n = 593), and 23% (n = 416) consumed PSRTEC, respectively; 12% (n = 178), 9% (n = 165), and 8% (n = 144), respectively, consumed NPSRTEC; and 54% (n = 795), 58% (n = 1032), and 69% (n = 1266), respectively, consumed OBs. RTEC consumers had significantly higher intakes of thiamin, riboflavin, niacin, vitamin B6 and B12, folate, calcium, magnesium, iron, and zinc than OB consumers. In children 4 to 8 years old and 9 to 13 years old, added sugars (103.3 ± 2.6 and 111.3 ± 3.0 g) were highest in PSRTEC consumers. Fiber intake was highest in NPSRTEC (4-8 and 9-13 years of age) consumers (14.3 ± 0.4 and 14.7 ± 0.7 g) and lowest in 14- to 18-year-old OB consumers (11.9 ± 0.3 g). Weight measures were not higher in either group of RTEC consumers compared with OB consumers. Data suggest that RTEC contribute a high level of nutrients to the diet, and PSRTEC is not associated with higher body weight in children.
Collapse
Affiliation(s)
- Carol E. O’Neil
- Louisiana State University Agricultural Center, Baton Rouge, Louisiana (CEO, MZ), NutraSource, Inc, Clarksville, Maryland
- USDA/ARS Children’s Nutrition Research Center at Baylor College of Medicine, Houston, Texas (TAN), NutraSource, Inc, Clarksville, Maryland
| | - Michael Zanovec
- Louisiana State University Agricultural Center, Baton Rouge, Louisiana (CEO, MZ), NutraSource, Inc, Clarksville, Maryland
- USDA/ARS Children’s Nutrition Research Center at Baylor College of Medicine, Houston, Texas (TAN), NutraSource, Inc, Clarksville, Maryland
| | - Theresa A. Nicklas
- Louisiana State University Agricultural Center, Baton Rouge, Louisiana (CEO, MZ), NutraSource, Inc, Clarksville, Maryland
- USDA/ARS Children’s Nutrition Research Center at Baylor College of Medicine, Houston, Texas (TAN), NutraSource, Inc, Clarksville, Maryland
| | - Susan S. Cho
- Louisiana State University Agricultural Center, Baton Rouge, Louisiana (CEO, MZ), NutraSource, Inc, Clarksville, Maryland
- USDA/ARS Children’s Nutrition Research Center at Baylor College of Medicine, Houston, Texas (TAN), NutraSource, Inc, Clarksville, Maryland
| |
Collapse
|
7
|
A systematic review of the effect of breakfast on the cognitive performance of children and adolescents. Nutr Res Rev 2009; 22:220-43. [DOI: 10.1017/s0954422409990175] [Citation(s) in RCA: 298] [Impact Index Per Article: 19.9] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/05/2022]
Abstract
Breakfast is recommended as part of a healthy diet because it is associated with healthier macro- and micronutrient intakes, BMI and lifestyle. Breakfast is also widely promoted to improve cognitive function and academic performance, leading to the provision of breakfast initiatives by public health bodies. Despite this positive and intuitive perception of cognitive benefits, there has been no systematic review of the evidence. Systematic review methodology was employed to evaluate the effects of breakfast on cognitive performance in well-nourished children and nutritionally at-risk or stunted children. Acute experimental studies, school feeding programmes and studies of habitual breakfast intake are reviewed. Comparisons of breakfast v. no breakfast and breakfasts differing in energy and macronutrient composition are discussed. Included are forty-five studies described in forty-one papers published between 1950 and 2008. The evidence indicates that breakfast consumption is more beneficial than skipping breakfast, but this effect is more apparent in children whose nutritional status is compromised. There is a lack of research comparing breakfast type, precluding recommendations for the size and composition of an optimal breakfast for children's cognitive function. Few studies examined adolescents. Studies of school breakfast programmes suggest that such interventions can have positive effects on academic performance, but this may be in part explained by the increased school attendance that programmes encourage. The present systematic review considers methodological issues in this field and makes recommendations for future research design and policy priorities.
Collapse
|
8
|
Beydoun MA, Wang Y. Parent-child dietary intake resemblance in the United States: evidence from a large representative survey. Soc Sci Med 2009; 68:2137-44. [PMID: 19375837 PMCID: PMC2730650 DOI: 10.1016/j.socscimed.2009.03.029] [Citation(s) in RCA: 95] [Impact Index Per Article: 6.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/11/2008] [Indexed: 10/20/2022]
Abstract
We studied the association in dietary intakes and patterns between parents (aged 20-65 years) and their children (aged 2-18 years), using nationally representative data collected by the US Department of Agriculture (USDA) in the Continuing Survey of Food Intake by Individuals 1994-96. We analyzed two 24-h recall dietary data for 1061 fathers, 1230 mothers, 1370 sons and 1322 daughters. All analyses adjusted for sampling design complexity. We assessed multivariate-adjusted parent-child correlations in selected nutrients, food groups and overall dietary quality assessed using the new USDA 2005 Healthy Eating Index score (HEI(n)). The parent-child correlations were weak or moderate (0.20-0.33) for most intake measures. There were clear patterns of interaction with gender dyads in the intakes of calcium and dairy products (P<0.05 for dyadxparental intake), whereby multivariate-adjusted correlations in mother-daughter or mother-child dyads were significantly stronger compared to their father-child counterparts. The reverse was true for multivariate-adjusted correlations in HEI(n). Hispanics and other ethnic groups had significantly stronger resemblance than Non-Hispanic whites and blacks in soft drinks and HEI(n). Resemblance in general was stronger among older children, though the reverse was true when considering agreement in HEI(n)'s upper quintile. The influence of family income and parental education on the resemblance was small. In conclusion, parent-child dietary resemblance in the US is relatively weak, and varies by nutrients and food groups, and by the types of parent-child dyad and population groups. Factors other than parental eating behaviors seem to play an important role in affecting American young people's dietary intake.
Collapse
Affiliation(s)
| | - Youfa Wang
- Johns Hopkins School of Public Health, Center for Human Nutrition, Department of International Health, 615 N Wolfe St., Room E2546, Baltimore, MD 21205, United States
| |
Collapse
|