Rapid recognition of processed milk type using electrical impedance spectroscopy and machine learning

Huang, Ziyu; Xiao, Yanghao; Xiao, Yexiang; Cai, Honghao; Ni, Hui

doi:10.1111/ijfs.16440

Cited by 3 publications

(2 citation statements)

References 65 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared to the accuracy of 0.997 achieved using the Swin Transformer model, there was a decrease of 0.205. Although EIS combined with machine learning has shown promising results in binary or ternary classification problems with accuracies above 0.9 (Huang et al., 2023; Tiitta et al., 2020), a multi‐classification method that can differentiate subtle differences is needed for quantitative analysis. Moreover, EIS data is sensitive to temperature, therefore, in order to obtain reliable EIS data, experiments need to be conducted under constant temperature conditions, and typically require a waiting time of around 30 min for data stabilization.…”

Section: Resultsmentioning

confidence: 99%

Real‐time quantitative detection of hydrocolloid adulteration in meat based on Swin Transformer and smartphone

Gao,

Chen,

Huang

et al. 2024

Journal of Food Science

Self Cite

View full text Add to dashboard Cite

Hydrocolloids are widely used in meat products as common food additives. However, research has indicated that excessive consumption of these hydrocolloids may have potential health implications. Currently, consumers mainly rely on sensory evaluation to identify hydrocolloid adulteration in meat products. Although many studies on quantitative detection of hydrocolloids have been conducted by biochemical methods in laboratory environments, there is currently a lack of effective tools for consumers and regulators to obtain real‐time and reliable information on hydrocolloid adulteration. To address this challenge, a smartphone‐based computer vision method was developed to quantitatively detect carrageenan adulteration in beef in this work. Specifically, Swin Transformer models, along with pre‐training and fine‐tuning techniques, were used to successfully automate the classification of beef into nine different levels of carrageenan adulteration, ranging from 0% to 20%. Among the tested models, Swin‐Tiny (Swin‐T) achieved the highest trade‐off performance, with a Top‐1 accuracy of 0.997, a detection speed of 3.2 ms, and a model size of 103.45 Mb. Compared to computer vision, the electrochemical impedance spectroscopy achieved a lower accuracy of 0.792 and required a constant temperature environment and a waiting time of around 30 min for data stabilization. In addition, Swin‐T model was also capable of distinguishing between different types of hydrocolloids with a Top‐1 accuracy of 0.975. This study provides consumers and regulators with a valuable tool to obtain real‐time quantitative information about meat adulteration anytime, anywhere.Practical ApplicationThis research provides a practical solution for regulators and consumers to non‐destructively and quantitatively detect the content and type of hydrocolloids in beef in real‐time using smartphones. This innovation has the potential to significantly reduce the costs associated with meat quality testing, such as the use of chemical reagents and expensive instruments.

show abstract

Section: Resultsmentioning

confidence: 99%

Real‐time quantitative detection of hydrocolloid adulteration in meat based on Swin Transformer and smartphone

Gao,

Chen,

Huang

et al. 2024

Journal of Food Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…Electrical impedance spectroscopy (EIS) has been widely used for analysing biological objects in various fields, such as medical (Gamal et al ., 2018), agronomic (Liu et al ., 2021b), and food (Pliquett, 2010; Huang et al ., 2023). Unlike traditional biochemical methods that require expensive instruments, complex sample preparation, and chemical reagents, EIS is a cost‐effective alternative.…”

Section: Introductionmentioning

confidence: 99%

Rapid beef quality detection using spectra pre‐processing methods in electrical impedance spectroscopy and machine learning

Qiu,

Lin,

et al. 2024

Int J of Food Sci Tech

Self Cite

View full text Add to dashboard Cite

SummaryThe fraudulent practice of beef adulteration is a growing concern, as it violates consumer rights. Electrical impedance spectroscopy (EIS) combined with machine learning has emerged as a widely used approach to identify low‐quality meat. Unlike traditional biochemical methods that require expensive instruments, complex sample preparation, and chemical reagents, EIS is a cost‐effective alternative. However, EIS data are susceptible to temperature fluctuations, requiring a waiting period under consistent temperature conditions for data stabilisation before measurements. This process becomes impractical when dealing with a large number of samples. To overcome this limitation, standardisation, normalisation, and smoothing methods were introduced in the meat quality detection based on EIS data. A recognition model for detecting carrageenan adulteration in beef was established. Under an inconsistent temperature condition, by applying the spectra pre‐processing methods to the prediction dataset, the model accuracy reached 84%, whereas the accuracy of the unprocessed prediction dataset dropped to 54%. This study demonstrates that acquiring EIS data under consistent temperature conditions is unnecessary if proper spectra pre‐processing methods are applied. By eliminating the waiting time for data stabilisation, this practical approach enhances the efficiency and accuracy of meat quality detection.

show abstract

Assessing nutritional probing and storage stability of functional Aloe vera (Aloe barbadensis) based guava jam: a machine learning approach for predictive modelling

Ibraheem,

Tariq,

Ali

et al. 2024

Int J of Food Sci Tech

View full text Add to dashboard Cite

SummaryThis research aimed to explore the influence of nutritional adjustments and storage stability on a functional, reduced‐calorie guava jam incorporating Aloe vera. Over two months, comprehensive analysis assessed physicochemical properties, sensory traits, microbial stability, and shelf‐life. The addition of Aloe vera gel resulted in significant improvements in pH (3.00 to 3.65), total soluble solids (40.10 to 42.20°Brix), antioxidant activity (36.85% to 81.09%), moisture content (29.48% to 38.82%), water activity (0.78 to 0.84), ash content (0.29% to 0.45%), fat content (0.14% to 0.19%), fibre content (1.05% to 1.86%), and the colour values. Moreover, b* scores for colour indication improved from 15.09 to 18.86. Texture attributes of cohesiveness and firmness improved significantly. Sensory evaluation favoured the T2 variant (20% Aloe vera gel), suggesting it as the optimal formulation. Furthermore, artificial neural networks (ANNs), a technique of machine learning, were utilised to predict guava jam behaviour, with 99% accuracy. The study discovered substantial changes in pH, total soluble solids, antioxidant activity, moisture content, and textural qualities, indicating that Aloe vera supplementation could improve guava jam quality and shelf‐life. The results of ANN predictions about antioxidants and cohesiveness provide information about the product's performance during storage.

show abstract

Rapid recognition of processed milk type using electrical impedance spectroscopy and machine learning

Cited by 3 publications

References 65 publications

Real‐time quantitative detection of hydrocolloid adulteration in meat based on Swin Transformer and smartphone

Real‐time quantitative detection of hydrocolloid adulteration in meat based on Swin Transformer and smartphone

Rapid beef quality detection using spectra pre‐processing methods in electrical impedance spectroscopy and machine learning

Assessing nutritional probing and storage stability of functional Aloe vera (Aloe barbadensis) based guava jam: a machine learning approach for predictive modelling

Contact Info

Product

Resources

About