{"title":"Effect of novel phage cocktail on Salmonella recovered from broiler sources and its anti-biofilm effect on food contact surface model","authors":"Wattana Pelyuntha , David Yembilla Yamik , Nichapatr Vetboocha , Kitiya Vongkamjan","doi":"10.1016/j.foodcont.2024.111000","DOIUrl":"10.1016/j.foodcont.2024.111000","url":null,"abstract":"<div><div>The contamination of <em>Salmonella</em> in poultry products remains an important food safety and economic issue as the products have been recalled or rejected. Traditional antibiotic-based therapy in the animal production chain can cause antibiotic resistance (ABR) in bacteria, leading to severe illness and failure to treat diseases. Some strains of <em>Salmonella</em> produce genotoxin and extracellular polymeric substances called “biofilm” that support the survival of this bacterium. From this study, phages were applied to control cytolethal distending toxin B (<em>CdtB</em>)-producing, antibiotic-resistant, and biofilm-producing <em>Salmonella</em> isolated from the broiler sources. Out of 116 isolates, only 11 isolates (9.5%) representing six serovars were characterized. Of these, seven isolates were characterized as multidrug-resistant (MDR), three were resistant to one antibiotic, and only one isolate was susceptible to all antibiotics tested. For biofilm formation ability, 18.2% of the isolates were identified as strong biofilm producers, while 63.6% and 18.2% were moderate and weak, respectively. Up to 12 <em>Salmonella</em> phages from our collection were chosen to test their ability on given <em>Salmonella</em>. All the phages showed a strong lytic ability up to 90.9%. Three phages with the highest lysis potential (100% lytic ability) were further selected for a phage cocktail preparation, including WPX5, WPX8, and WPX9. Overall, the phage cocktail completely reduced representative <em>Salmonella</em> counts by three log units <em>in vitro</em> at a multiplicity of infection (MOI) of 10<sup>4</sup> and 10<sup>5</sup> at 6 h of treatment. The phage cocktail was tested to reduce <em>Salmonella</em> on different food contact surfaces. At MOI at 10<sup>3</sup>, the highest reduction of <em>Salmonella</em> attachment was observed in stainless steel surfaces, indicated by a 67% reduction, followed by polyvinyl chloride (PVC) and ceramic by 64% and 58%, respectively. These results suggest that a developed phage cocktail could be a potential biocontrol to control toxin-producing, biofilm-producing, and MDR <em>Salmonella</em> in the poultry industry.</div></div>","PeriodicalId":319,"journal":{"name":"Food Control","volume":"169 ","pages":"Article 111000"},"PeriodicalIF":5.6,"publicationDate":"2024-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142662467","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Food ControlPub Date : 2024-11-04DOI: 10.1016/j.foodcont.2024.111003
Asad Nawaz , Atka Afzal , Ali Khatibi , Amar Shankar , Heena Madan , Hafiza Sana Faisal , Amna Shahbaz , Ifrah Usman , Nimra Zulfiqar , Farhan Saeed , Aftab Ahmed , Ali Imran , Muhammad Afzaal , Awais karni , Faiyaz Ahmed , Noor Akram , Musarrat Rasheed , Fakhar Islam
{"title":"Role of artificial intelligence in halal authentication and traceability: A concurrent review","authors":"Asad Nawaz , Atka Afzal , Ali Khatibi , Amar Shankar , Heena Madan , Hafiza Sana Faisal , Amna Shahbaz , Ifrah Usman , Nimra Zulfiqar , Farhan Saeed , Aftab Ahmed , Ali Imran , Muhammad Afzaal , Awais karni , Faiyaz Ahmed , Noor Akram , Musarrat Rasheed , Fakhar Islam","doi":"10.1016/j.foodcont.2024.111003","DOIUrl":"10.1016/j.foodcont.2024.111003","url":null,"abstract":"<div><div>The convergence of artificial intelligence (AI) and the halal food industry represents a significant and transformative intersection, driven by the need to ensure the authenticity, safety, and compliance of halal food products in an increasingly complex global market. This review paper examines the applications, implications, and impact of AI technologies within the halal food sector. AI technologies, including machine learning, computer vision, natural language processing, and blockchain, are explored in the context of their contributions to addressing key challenges and opportunities. These technologies play a vital role in enhancing quality control, supply chain transparency, and consumer trust, and they offer solutions for the certification and monitoring of halal compliance. It further delves into examining how AI contributes to ensuring halal compliance, enhancing quality control, improving supply chain transparency, and fortifying consumer trust.</div></div>","PeriodicalId":319,"journal":{"name":"Food Control","volume":"169 ","pages":"Article 111003"},"PeriodicalIF":5.6,"publicationDate":"2024-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142662473","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Food ControlPub Date : 2024-11-03DOI: 10.1016/j.foodcont.2024.110992
Youte Wang , Yu Wang , Xianlong Yu , Kunhua Wang , Yiwu Lu , Huaide Xu , Lili Wang , Jun Wang
{"title":"Effect of buckwheat inclusion fractions on the drying characteristics and quality attributes of instant noodle sheets","authors":"Youte Wang , Yu Wang , Xianlong Yu , Kunhua Wang , Yiwu Lu , Huaide Xu , Lili Wang , Jun Wang","doi":"10.1016/j.foodcont.2024.110992","DOIUrl":"10.1016/j.foodcont.2024.110992","url":null,"abstract":"<div><div>This study investigated the effects of varying proportions of buckwheat flour mixed with wheat flour on the drying characteristics, physical properties, and nutritional quality of instant buckwheat noodle sheets (IBNS). The results indicated that incorporating buckwheat into pure wheat sheets reduced the drying time of IBNS by 10–16.7 min. With the increase in drying time, the amount of free water decreased significantly. The transverse relaxation times (T<sub>2</sub>) range gradually decreased, and the overall peak shifted towards the left. During rehydration, sheets with 10% buckwheat content exhibited a rehydration ratio of 17%, which was higher than that of pure wheat sheets. Initially, pure wheat sheets displayed the highest levels of hardness and chewiness during rehydration. The inclusion of buckwheat flour greatly improved the nutritional content and antioxidant capacity of the IBNS. The addition of buckwheat modified the secondary structure of proteins and the ordered structure of starch. Raw sheets with a 10% buckwheat addition demonstrated the highest value of short-range ordered structure (DO) at 2.16. In contrast, noodles with a 30% buckwheat addition showed the highest lacunarity at 38.44 × 10<sup>−2</sup>. The findings of this study will serve as a theoretical foundation and provide guidance for the production of mixed-grain composite products in the instant food industry.</div></div>","PeriodicalId":319,"journal":{"name":"Food Control","volume":"169 ","pages":"Article 110992"},"PeriodicalIF":5.6,"publicationDate":"2024-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142593053","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Food ControlPub Date : 2024-11-02DOI: 10.1016/j.foodcont.2024.110990
Wenbo Li , Peng Zhang , Lianye Li , Fei Xia , Lingling Wu , Wenxiu Sun
{"title":"A dynamic word conversion indicator based on water resistance poly (vinyl alcohol)/basic lead acetate nanofibers for monitoring shrimp freshness","authors":"Wenbo Li , Peng Zhang , Lianye Li , Fei Xia , Lingling Wu , Wenxiu Sun","doi":"10.1016/j.foodcont.2024.110990","DOIUrl":"10.1016/j.foodcont.2024.110990","url":null,"abstract":"<div><div>Visual monitoring of seafood is particularly important for dietary health. Hydrophilic materials provide more direct visual results, yet improving their water resistance while maintaining high sensitivity is challenging. In response, we introduce a visual monitoring fibrous film characterized by high water resistance and sensitivity. A poly (vinyl alcohol)/basic lead acetate (PVA/Pb) nanofiber film responding to hydrogen sulfide (H<sub>2</sub>S) released during seafood spoilage, employing poly (vinyl alcohol) (PVA) as the support and basic lead acetate (Pb(Ac)<sub>2</sub>) as the indicator, has been fabricated through electrospinning. Through studies on different crosslinking durations, 4 h was found optimal, giving the film satisfactory water resistance (with a mass loss rate of 35.93%) and sensitivity (2.3 ppm<sup>−1</sup>), alongside investigations into its operational mechanism. To enhance the widespread application of the PVA/Pb nanofiber film, we employed its color transition from white to brown as a dynamic word conversion indicator (DWCI), accurately indicating shrimp freshness from “edible\" to “inedible\". This innovative design provides a clear and convenient solution for special populations (including visually impaired individuals, older people, and children), enabling barrier-free real-time detection of seafood freshness.</div></div>","PeriodicalId":319,"journal":{"name":"Food Control","volume":"169 ","pages":"Article 110990"},"PeriodicalIF":5.6,"publicationDate":"2024-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142662409","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Food ControlPub Date : 2024-11-02DOI: 10.1016/j.foodcont.2024.110989
Menglong Liu, Yiyao Liu, Jingyi Xiao, Yi Ren, Xue Gao
{"title":"Colorimetric-ratiometric fluorescence sensors dual-mode detection of ofloxacin based on Fe-Ni bimetallic metal-organic frameworks and carbon dots","authors":"Menglong Liu, Yiyao Liu, Jingyi Xiao, Yi Ren, Xue Gao","doi":"10.1016/j.foodcont.2024.110989","DOIUrl":"10.1016/j.foodcont.2024.110989","url":null,"abstract":"<div><div>Ofloxacin (OFL) is an effective antibiotic against bacteria. However, the overuse of ofloxacin can lead to residues in animal-derived foods, posing a potential threat to human health. Due to the issue of false positives in single-mode detection, it is crucial to establish a multi-mode method for accurate and sensitive detection of ofloxacin. Peroxidase-like active iron-nickel bimetallic organic framework (FeNi-MOF) and orange fluorescent carbon dots (OCDs) synthesized by a simple one-pot method were used to establish a dual-mode colorimetric/ratiometric fluorescence biosensor for the detection of ofloxacin residues. In colorimetric mode, in the presence of hydrogen peroxide (H<sub>2</sub>O<sub>2</sub>), FeNi-MOF can catalyze the oxidation of 3,3′,5,5′-tetramethylbenzidine (TMB) to produce blue oxTMB. However, the fluorescence of OCDs at 588 nm will be quenched by oxTMB, thus forming a ratiometric fluorescence signal with the fluorescence of FeNi-MOF at 420 nm. The linear range of the dual-mode detection method was 10<sup>−7</sup>-10<sup>−2</sup> mg/mL, and the detection limits were 1.1 ng/kg (colorimetric) and 0.34 ng/kg (radiometric fluorescence). In addition, the constructed smartphone sensing platform is also well applied for on-site portable detection. Therefore, we have established a dual-mode sensor with high specificity, accuracy, and excellent anti-interference ability for OFL detection, and satisfactory results are also obtained in the actual spiked samples.</div></div>","PeriodicalId":319,"journal":{"name":"Food Control","volume":"169 ","pages":"Article 110989"},"PeriodicalIF":5.6,"publicationDate":"2024-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142586995","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Food ControlPub Date : 2024-11-02DOI: 10.1016/j.foodcont.2024.110972
Yingying Sun, Yana Liu, Han Wang, Bo Zou, Yijie Zhao, Xingmin Li, Ruitong Dai
{"title":"Revealing the mechanism of ohmic heating against Bacillus cereus spores through intracellular homeostasis and lipid analysis","authors":"Yingying Sun, Yana Liu, Han Wang, Bo Zou, Yijie Zhao, Xingmin Li, Ruitong Dai","doi":"10.1016/j.foodcont.2024.110972","DOIUrl":"10.1016/j.foodcont.2024.110972","url":null,"abstract":"<div><div>B. cereus spores can cause food spoilage, leading to food-borne illness and significant economic losses. Ohmic heating (OH) is an emerging and alternative thermal processing technique widely used for pasteurizing or sterilizing food, with the potential to inactivate spores. The results of this study indicated that OH can effectively inactivate <em>B. cereus</em> spores. To elucidate the inactivation mechanism of <em>B. cereus</em> spores by OH, several physicochemical indices were investigated, including spore characteristics, spore membrane integrity, intracellular substance concentration, and spore lipid composition. OH treatment neither produced H<sub>2</sub>O<sub>2</sub> during processing nor caused lipid oxidation in spores. Both oil bath heating (OB) and OH treatments affected catalase, glutathione peroxidase, ATPase, Na<sup>+</sup>/K<sup>+</sup>-ATPase, and Ca<sup>2+</sup>/Mg<sup>2+</sup>-ATPase activity, suggesting a disruption in the intracellular homeostasis of the spores. Additionally, K<sup>+</sup>, Ca<sup>2+</sup>, and Mg<sup>2+</sup> leakage from spores, and a decrease in inner membrane fluidity was observed following OB and OH treatments. However, OH treatment at 10 V/cm and 50 Hz resulted in minimal cation leakage and more pronounced catalase activity damage. OH treatments had a more significant impact on the lipid composition and metabolism pathways of spores than OB treatment, particularly concerning major membrane-forming phospholipids and glycerophospholipid metabolism. This study provides a comprehensive understanding of the OH inactivation mechanism on spores from various perspectives, promoting theoretical advancement of OH application in food sterilization.</div></div>","PeriodicalId":319,"journal":{"name":"Food Control","volume":"169 ","pages":"Article 110972"},"PeriodicalIF":5.6,"publicationDate":"2024-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142586990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Food ControlPub Date : 2024-10-30DOI: 10.1016/j.foodcont.2024.110987
Shruthy Seshadrinathan , V.M. Balasubramaniam , Abigail B. Snyder , Robert Dupont , Xiaoguang Wang
{"title":"Utilizing superheated steam to inactivate Enterococcus faecium NRRL B-2354 on various material surfaces used in food and produce industries","authors":"Shruthy Seshadrinathan , V.M. Balasubramaniam , Abigail B. Snyder , Robert Dupont , Xiaoguang Wang","doi":"10.1016/j.foodcont.2024.110987","DOIUrl":"10.1016/j.foodcont.2024.110987","url":null,"abstract":"<div><div>Traditional sanitation in dry food and produce manufacturing plants is challenging as residual moisture can harbor microbial contamination. Superheated steam (SHS) is produced when water is heated to temperatures (125–400 °C) beyond the boiling point. SHS does not lead to condensation on surfaces. There are limited studies available on the sanitation efficacy of SHS on different surfaces. This study was conducted to understand the sanitation efficacy of SHS to inactivate <em>Enterococcus faecium</em> NRRL B-2354 on different surfaces (stainless steel, concrete, plywood, leather, polytetrafluoroethylene (PTFE), silicon rubber, cotton, and cardboard). The surfaces were either spot inoculated or coated with a representative food matrix (baby formula) and treated with SHS at 150 °C for a total process time (come-up time + treatment time) of 3 min. Spot-inoculated surfaces showed a higher microbial inactivation (range: 9.2 ± 0.9 to 10.7 ± 0.7 log reduction/coupon) whereas surfaces with inoculum coated with baby formula had lower microbial inactivation (range: 7.8 ± 0.4 to 9.6 ± 0.4 log reduction/coupon). The effects of surface characteristics like thermal inertia, surface roughness, and hydrophobicity of microbial inactivation were studied. Prolonged usage of SHS (150 °C for 5 min once a week over 8 weeks) did not alter surface characteristics, particularly roughness of stainless steel, silicone rubber and plywood.</div></div>","PeriodicalId":319,"journal":{"name":"Food Control","volume":"169 ","pages":"Article 110987"},"PeriodicalIF":5.6,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142578012","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Food ControlPub Date : 2024-10-30DOI: 10.1016/j.foodcont.2024.110978
Cláudia Esteves da Silva , Fernanda Vitória Leimann , Yuri Yohan Ribeiro Sanches , Vanessa de Carvalho Rodrigues , Ali Tfayli , Douglas N. Rutledge , Paulo Henrique Março
{"title":"Food freshness and composition evaluated by Colorimetry, TPA, and spectroscopy through ICA-based ComDim: A case study of a peanut-based protein-enriched food","authors":"Cláudia Esteves da Silva , Fernanda Vitória Leimann , Yuri Yohan Ribeiro Sanches , Vanessa de Carvalho Rodrigues , Ali Tfayli , Douglas N. Rutledge , Paulo Henrique Março","doi":"10.1016/j.foodcont.2024.110978","DOIUrl":"10.1016/j.foodcont.2024.110978","url":null,"abstract":"<div><div>This study aimed to apply Common Components and Specific Weights Analysis (CCSWA or ComDim) to explore the relations between colorimetry (color), Texture Profile Analysis (TPA), and Near-Infrared Spectroscopy (NIR) besides determining which of these techniques is most effective in differentiating the freshness and composition of the studied samples. ComDim-ICA, a recent modification of ComDim based on Independent Components Analysis (ICA) decomposition, was used. This approach was chosen to provide more straightforward and interpretable scores and loadings compared to the classical ComDim, which is based on Principal Components Analysis (PCA). The experiment was performed on a peanut-based food enriched with powdered proteins derived from pumpkin seed, rice, pea, sunflower seed, water lentil (duckweed), flaxseed, soybean, and whey. Measurements (Color, NIR, and TPA) were taken on the day of the food preparation, after seven days, and after fourteen days. The global scores indicated that CC2 carried information regarding freshness, while CC3, CC4, and CC5 were associated with food composition. According to the saliences, NIR and color were the most important techniques for determining food freshness (the most important in CC2). Additionally, in CC3, NIR was responsible for distinguishing lentil and rice proteins from the other protein sources. The best differentiation regarding food composition was found in CC4 and CC5, where colorimetry and TPA were most significant. These findings may encourage new applications of multiblock analysis to elucidate differences in food quality based on diverse evaluation techniques.</div></div>","PeriodicalId":319,"journal":{"name":"Food Control","volume":"169 ","pages":"Article 110978"},"PeriodicalIF":5.6,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142578011","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Food ControlPub Date : 2024-10-30DOI: 10.1016/j.foodcont.2024.110983
Yongqing Li , Yikai Fan , Jingyi Gao , Li Liu , Lijun Cao , Bo Hu , Zunongjiang Abula , Yeerlan Xieermaola , Haitong Wang , Chu Chu , Zhuo Yang , Guochang Yang , Peipei Wen , Dongwei Wang , Wenxin Zheng , Shujun Zhang
{"title":"Rapid detection and spectroscopic feature analysis of mineral content in camel milk using fourier-transform mid-infrared spectroscopy and traditional machine learning algorithms","authors":"Yongqing Li , Yikai Fan , Jingyi Gao , Li Liu , Lijun Cao , Bo Hu , Zunongjiang Abula , Yeerlan Xieermaola , Haitong Wang , Chu Chu , Zhuo Yang , Guochang Yang , Peipei Wen , Dongwei Wang , Wenxin Zheng , Shujun Zhang","doi":"10.1016/j.foodcont.2024.110983","DOIUrl":"10.1016/j.foodcont.2024.110983","url":null,"abstract":"<div><div>Camel milk is rich in nutrients and bioactive factors, with mineral content generally higher than that of cow milk, but there is currently no internationally established, rapid, batch-testing method for the mineral content. This study collected samples of camel milk from 113 locations in Xinjiang, China. For the first time internationally, based on the true mineral values determined by ICP-OES (Inductively Coupled Plasma Optical Emission Spectroscopy) and the extracted mid-infrared spectra data, a quantitative prediction model for 10 key minerals (Ca, Fe, K, Mg, Mn, Na, P, Sr, Zn, and Se) was established using Fourier-Transform Mid-Infrared Spectroscopy (FT-MIRS) and the traditional machine learning algorithm Partial Least Squares Regression. The R<sub>t</sub><sup>2</sup> of the test set ranged from 0.61 to 0.91, RMSE<sub>t</sub> ranged from 2.21ug/kg(Se) to 197.08 mg/kg(K) and the RPD<sub>t</sub> from 1.59 to 3.28. In addition, the distribution, patterns, and correlations of mineral-related characteristic wavenumbers in camel milk were summarized. This study opens a new avenue for the rapid detection of minerals in camel milk and fills the research gap in in using FT-MIRS to detect mineral content in camel milk.</div></div>","PeriodicalId":319,"journal":{"name":"Food Control","volume":"169 ","pages":"Article 110983"},"PeriodicalIF":5.6,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142586872","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Food ControlPub Date : 2024-10-29DOI: 10.1016/j.foodcont.2024.110979
Arman Arefi , Barbara Sturm , Thomas Hoffmann
{"title":"Explainability of deep convolutional neural networks when it comes to NIR spectral data: A case study of starch content estimation in potato tubers","authors":"Arman Arefi , Barbara Sturm , Thomas Hoffmann","doi":"10.1016/j.foodcont.2024.110979","DOIUrl":"10.1016/j.foodcont.2024.110979","url":null,"abstract":"<div><div>Explainable AI is gaining popularity as a way to understand the decision-making processes of neural networks and gain insight into their predictions. In this paper, Integrated Gradients (IG) was applied to assess the relevance of spectral features used by deep convolutional neural networks in predicting the starch content of potatoes. For this purpose, spectral information of 7651 tubers of 12 potato varieties was acquired using a NIR spectrometer in the spectral range of 940–1650 nm. This was followed by a reference measurement of starch content. Three one-dimensional deep convolutional neural networks i.e. VGG-19, InceptionV3, and SpectraNet-32 were developed using the Keras API. The deep networks outperformed traditional models in the starch content prediction, with SpectraNet-32 achieving the highest prediction accuracy (R<sup>2</sup> = 0.84, RMSE = 1.41%, RPD = 2.46, and rRMSE = 9.88%). Further analysis of the neural networks by IG indicated that the predictions were generated based on starch relevant spectral bands. The results of this study demonstrated that the deep convolutional neural networks not only could accurately predict starch content in potatoes, but also provided certainty in the predictions.</div></div>","PeriodicalId":319,"journal":{"name":"Food Control","volume":"169 ","pages":"Article 110979"},"PeriodicalIF":5.6,"publicationDate":"2024-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142662410","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}