The Analysis of Nutrition Toxicology Detection Based on Big Data and Deep Learning
Public health and safety are increasingly concerned as public awareness of health-related issues grows. To find a rapid, convenient, and non-destructive testing method for detecting human nutritional toxicology detection, this study selects sildenafil, phenolphthalein, and metformin hydrochloride&am...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10328874/ |
| _version_ | 1797400982996385792 |
|---|---|
| author | Jing Shi Renjuan Chen Yonghong Ma Yancheng Feng Ke Men |
| author_facet | Jing Shi Renjuan Chen Yonghong Ma Yancheng Feng Ke Men |
| author_sort | Jing Shi |
| collection | DOAJ |
| description | Public health and safety are increasingly concerned as public awareness of health-related issues grows. To find a rapid, convenient, and non-destructive testing method for detecting human nutritional toxicology detection, this study selects sildenafil, phenolphthalein, and metformin hydrochloride—commonly found additives in health products—as the focal point. The research endeavors to tackle the paramount issue of public health and safety. The study begins by elucidating the public health and safety concept and then outlines the computational process for determining the terahertz (THz) optical properties. Subsequently, it provides a brief overview of deep learning (DL) methods, including the Back Propagation Neural Network (BPNN), Convolutional Neural Network (CNN), Residual Network (ResNet), and MobileNet model. Finally, the study compares and tests the THz absorption spectrum data of 22 pure samples containing sildenafil, phenolphthalein, and metformin hydrochloride by DL technique to evaluate the model’s classification performance. The findings demonstrate that, with increased training iterations, the model’s accuracy consistently improves and stabilizes. For instance, after 12 training iterations, CNN’s accuracy under the verification set stabilizes, frequently reaching nearly 100%. After 83 iterations, the accuracy remains steady at 98.96%. Similarly, the MobileNet model reaches stability after 17 iterations, achieving 100% accuracy. The BPNN demonstrates the fastest prediction time among the four DL algorithm models, at 310–5 seconds. Meanwhile, the MobileNet model exhibits the highest accuracy and stability. This study using THz waves to identify contaminants in medical items can significantly enhance public health and safety. |
| first_indexed | 2024-03-09T02:03:21Z |
| format | Article |
| id | doaj.art-0125c6df65d54e79abfd16f3d4acba48 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-09T02:03:21Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-0125c6df65d54e79abfd16f3d4acba482023-12-08T00:05:21ZengIEEEIEEE Access2169-35362023-01-011113510613511910.1109/ACCESS.2023.333694610328874The Analysis of Nutrition Toxicology Detection Based on Big Data and Deep LearningJing Shi0Renjuan Chen1Yonghong Ma2Yancheng Feng3Ke Men4https://orcid.org/0009-0009-3056-7626Institute for Research on Health Information and Technology, Xi’an Medical University, Xi’an, ChinaDepartment of Nutrition Hygiene and Toxicology, Xi’an Medical University, Xi’an, ChinaInstitute for Research on Health Information and Technology, Xi’an Medical University, Xi’an, ChinaInstitute for Research on Health Information and Technology, Xi’an Medical University, Xi’an, ChinaInstitute for Research on Health Information and Technology, Xi’an Medical University, Xi’an, ChinaPublic health and safety are increasingly concerned as public awareness of health-related issues grows. To find a rapid, convenient, and non-destructive testing method for detecting human nutritional toxicology detection, this study selects sildenafil, phenolphthalein, and metformin hydrochloride—commonly found additives in health products—as the focal point. The research endeavors to tackle the paramount issue of public health and safety. The study begins by elucidating the public health and safety concept and then outlines the computational process for determining the terahertz (THz) optical properties. Subsequently, it provides a brief overview of deep learning (DL) methods, including the Back Propagation Neural Network (BPNN), Convolutional Neural Network (CNN), Residual Network (ResNet), and MobileNet model. Finally, the study compares and tests the THz absorption spectrum data of 22 pure samples containing sildenafil, phenolphthalein, and metformin hydrochloride by DL technique to evaluate the model’s classification performance. The findings demonstrate that, with increased training iterations, the model’s accuracy consistently improves and stabilizes. For instance, after 12 training iterations, CNN’s accuracy under the verification set stabilizes, frequently reaching nearly 100%. After 83 iterations, the accuracy remains steady at 98.96%. Similarly, the MobileNet model reaches stability after 17 iterations, achieving 100% accuracy. The BPNN demonstrates the fastest prediction time among the four DL algorithm models, at 310–5 seconds. Meanwhile, the MobileNet model exhibits the highest accuracy and stability. This study using THz waves to identify contaminants in medical items can significantly enhance public health and safety.https://ieeexplore.ieee.org/document/10328874/Public health and safetyadditives in health productsdeep learningBP neural networkterahertz wave |
| spellingShingle | Jing Shi Renjuan Chen Yonghong Ma Yancheng Feng Ke Men The Analysis of Nutrition Toxicology Detection Based on Big Data and Deep Learning IEEE Access Public health and safety additives in health products deep learning BP neural network terahertz wave |
| title | The Analysis of Nutrition Toxicology Detection Based on Big Data and Deep Learning |
| title_full | The Analysis of Nutrition Toxicology Detection Based on Big Data and Deep Learning |
| title_fullStr | The Analysis of Nutrition Toxicology Detection Based on Big Data and Deep Learning |
| title_full_unstemmed | The Analysis of Nutrition Toxicology Detection Based on Big Data and Deep Learning |
| title_short | The Analysis of Nutrition Toxicology Detection Based on Big Data and Deep Learning |
| title_sort | analysis of nutrition toxicology detection based on big data and deep learning |
| topic | Public health and safety additives in health products deep learning BP neural network terahertz wave |
| url | https://ieeexplore.ieee.org/document/10328874/ |
| work_keys_str_mv | AT jingshi theanalysisofnutritiontoxicologydetectionbasedonbigdataanddeeplearning AT renjuanchen theanalysisofnutritiontoxicologydetectionbasedonbigdataanddeeplearning AT yonghongma theanalysisofnutritiontoxicologydetectionbasedonbigdataanddeeplearning AT yanchengfeng theanalysisofnutritiontoxicologydetectionbasedonbigdataanddeeplearning AT kemen theanalysisofnutritiontoxicologydetectionbasedonbigdataanddeeplearning AT jingshi analysisofnutritiontoxicologydetectionbasedonbigdataanddeeplearning AT renjuanchen analysisofnutritiontoxicologydetectionbasedonbigdataanddeeplearning AT yonghongma analysisofnutritiontoxicologydetectionbasedonbigdataanddeeplearning AT yanchengfeng analysisofnutritiontoxicologydetectionbasedonbigdataanddeeplearning AT kemen analysisofnutritiontoxicologydetectionbasedonbigdataanddeeplearning |