Deep learning accurately predicts food categories and nutrients based on ingredient statements

Peihua Ma; Zhikun Zhang; Ying Li; Ning Yu; Jiping Sheng; Hande Küçük McGinty; Qin Wang; Jaspreet K C Ahuja

doi:10.1016/j.foodchem.2022.133243

Deep learning accurately predicts food categories and nutrients based on ingredient statements

Food Chem. 2022 Oct 15:391:133243. doi: 10.1016/j.foodchem.2022.133243. Epub 2022 May 19.

Authors

Peihua Ma¹, Zhikun Zhang², Ying Li³, Ning Yu⁴, Jiping Sheng⁵, Hande Küçük McGinty⁶, Qin Wang⁷, Jaspreet K C Ahuja³

Affiliations

¹ Department of Nutrition and Food Science, College of Agriculture and Natural Resources, University of Maryland, College Park, MD 20740, USA.
² CISPA Helmholtz Center for Information Security, Stuhlsatzenhaus 5, 66123 Saarbrucken, Germany.
³ U.S. Department of Agriculture, Agricultural Research Service, Beltsville Human Nutrition Research Center, 10300 Baltimore Ave, Bldg. 005, BARC-WEST, Beltsville, MD 20705, USA.
⁴ Department of Computer Science, College of Mathematical and Natural Sciences, University of Maryland, College Park, MD 20742, USA.
⁵ School of Agricultural Economics and Rural Development, Renmin University of China, Beijing 100872, China.
⁶ U.S. Department of Agriculture, Agricultural Research Service, Beltsville Human Nutrition Research Center, 10300 Baltimore Ave, Bldg. 005, BARC-WEST, Beltsville, MD 20705, USA; Department of Chemistry and BioChemistry, Ohio University, Athens, OH, USA.
⁷ Department of Nutrition and Food Science, College of Agriculture and Natural Resources, University of Maryland, College Park, MD 20740, USA. Electronic address: wangqin@umd.edu.

PMID: 35623276
DOI: 10.1016/j.foodchem.2022.133243

Abstract

Determining attributes such as classification, creating taxonomies and nutrients for foods can be a challenging and resource-intensive task, albeit important for a better understanding of foods. In this study, a novel dataset, 134 k BFPD, was collected from USDA Branded Food Products Database with modification and labeled with three food taxonomy and nutrient values and became an artificial intelligence (AI) dataset that covered the largest food types to date. Overall, the Multi-Layer Perceptron (MLP)-TF-SE method obtained the highest learning efficiency for food natural language processing tasks using AI, which achieved up to 99% accuracy for food classification and 0.98 R² for calcium estimation (0.93 ∼ 0.97 for calories, protein, sodium, total carbohydrate, total lipids, etc.). The deep learning approach has great potential to be embedded in other food classification and regression tasks and as an extension to other applications in the food and nutrient scope.

Keywords: Deep learning; Ingredients; Nutrients; USDA branded food products database.

MeSH terms

Artificial Intelligence*
Deep Learning*
Food
Neural Networks, Computer
Nutrients