Background: The complexity of food structure is such as to hinder its inclusion in mathematical models predicting food properties and transformations, although a considerable impulse is being determined by using artificial intelligence. As a matter of fact, food definition currently neglects the structural description, even in those fields for which structure is demonstrated to have a decisive role, such as nutrition. Scope and approach: This review aims to analyse the current knowledge about the structure of foods and its potential use to numerically define the sensory and nutritional quality, as well as the stability properties. Starting from this information, a possible methodology is explored to build, even in an automated way, mathematical models for simulating and predicting the properties of food. A model pipeline has been proposed and applied to pasta, in particular exploiting the description of the structural changes occurring upon cooking. Key findings and conclusions: Foods may be designed in silico, based on automated pipelines for direct extraction of information on rheological and sensory properties as derived from structure images and from data on the dynamic state of the water. The ultimate goal of these approaches is to make more limited use of expensive and time-consuming experiments on physically prepared foods to get to use digital twins of foods designed in the laboratory.

Food structure, function and artificial intelligence

Mengucci, Carlo;Picone, Gianfranco;Capozzi, Francesco
2022

Abstract

Background: The complexity of food structure is such as to hinder its inclusion in mathematical models predicting food properties and transformations, although a considerable impulse is being determined by using artificial intelligence. As a matter of fact, food definition currently neglects the structural description, even in those fields for which structure is demonstrated to have a decisive role, such as nutrition. Scope and approach: This review aims to analyse the current knowledge about the structure of foods and its potential use to numerically define the sensory and nutritional quality, as well as the stability properties. Starting from this information, a possible methodology is explored to build, even in an automated way, mathematical models for simulating and predicting the properties of food. A model pipeline has been proposed and applied to pasta, in particular exploiting the description of the structural changes occurring upon cooking. Key findings and conclusions: Foods may be designed in silico, based on automated pipelines for direct extraction of information on rheological and sensory properties as derived from structure images and from data on the dynamic state of the water. The ultimate goal of these approaches is to make more limited use of expensive and time-consuming experiments on physically prepared foods to get to use digital twins of foods designed in the laboratory.
Mengucci, Carlo; Ferranti, Pasquale; Romano, Annalisa; Masi, Paolo; Picone, Gianfranco; Capozzi, Francesco
File in questo prodotto:
Eventuali allegati, non sono esposti

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11585/878869
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? ND
social impact