Methodology to classify hazardous compounds via deep learning based on convolutional neural networks

Miri Seo; Sang Wook Lee

doi:10.1016/j.cap.2022.06.003

Methodology to classify hazardous compounds via deep learning based on convolutional neural networks

Miri Seo, Sang Wook Lee

Department of Physics

Research output: Contribution to journal › Article › peer-review

4 Scopus citations

Abstract

Compounds information such as Chemical Abstracts Service (CAS) registry number, hazards, and properties have been provided through Globally Harmonized System (GHS) based Material Safety Data Sheet (MSDS). This information can help users avoid hazardous compounds and handle chemicals in proper way. GHS specifies that hazards of compounds are categorized through animal testing (or in vivo testing), in vitro testing, epidemiological surveillance, and clinical trials. In this study, artificial intelligence (AI) is used to replace traditional approaches in predicting the toxicity of chemicals. A database of hazardous compounds is generated by data provided by the Ministry of Environment (ME), training and learning based on convolutional neural network (CNN) are carried out following data featurization. As a result, 90% of accuracy for CNN-based model is obtained using the image dataset. In contrast to the previous methods, the classification method based on CNN-based model in this study allows for the efficient discrimination of hazard chemicals without any additional tests.

Original language	English
Pages (from-to)	59-65
Number of pages	7
Journal	Current Applied Physics
Volume	41
DOIs	https://doi.org/10.1016/j.cap.2022.06.003
State	Published - Sep 2022

Bibliographical note

Funding Information:
The authors would like to thank Dr. Kookjin Lee at Intel Corporation for providing valuable comments on the machnie learning algorithm and data analysis on this work. This research was supported by the Basic Research Program ( NRF-2022R1A2B5B01001640 , NRF-2021R1A6A1A10039823 ) and Global Research and Development Center Program ( NRF-2018K1A4A3A01064272 ) through the National Research Foundation of Korea (NRF).

Publisher Copyright:
© 2022 Korean Physical Society

Keywords

Artificial intelligence
Classification
Deep learning
GHS
Hazardous compounds
MSDS

Access to Document

10.1016/j.cap.2022.06.003

Cite this

@article{e619a9f93fb4486da919c9f029db385b,

title = "Methodology to classify hazardous compounds via deep learning based on convolutional neural networks",

abstract = "Compounds information such as Chemical Abstracts Service (CAS) registry number, hazards, and properties have been provided through Globally Harmonized System (GHS) based Material Safety Data Sheet (MSDS). This information can help users avoid hazardous compounds and handle chemicals in proper way. GHS specifies that hazards of compounds are categorized through animal testing (or in vivo testing), in vitro testing, epidemiological surveillance, and clinical trials. In this study, artificial intelligence (AI) is used to replace traditional approaches in predicting the toxicity of chemicals. A database of hazardous compounds is generated by data provided by the Ministry of Environment (ME), training and learning based on convolutional neural network (CNN) are carried out following data featurization. As a result, 90% of accuracy for CNN-based model is obtained using the image dataset. In contrast to the previous methods, the classification method based on CNN-based model in this study allows for the efficient discrimination of hazard chemicals without any additional tests.",

keywords = "Artificial intelligence, Classification, Deep learning, GHS, Hazardous compounds, MSDS",

author = "Miri Seo and Lee, {Sang Wook}",

note = "Funding Information: The authors would like to thank Dr. Kookjin Lee at Intel Corporation for providing valuable comments on the machnie learning algorithm and data analysis on this work. This research was supported by the Basic Research Program ( NRF-2022R1A2B5B01001640 , NRF-2021R1A6A1A10039823 ) and Global Research and Development Center Program ( NRF-2018K1A4A3A01064272 ) through the National Research Foundation of Korea (NRF). Publisher Copyright: {\textcopyright} 2022 Korean Physical Society",

year = "2022",

month = sep,

doi = "10.1016/j.cap.2022.06.003",

language = "English",

volume = "41",

pages = "59--65",

journal = "Current Applied Physics",

issn = "1567-1739",

publisher = "Elsevier",

}

TY - JOUR

T1 - Methodology to classify hazardous compounds via deep learning based on convolutional neural networks

AU - Seo, Miri

AU - Lee, Sang Wook

N1 - Funding Information: The authors would like to thank Dr. Kookjin Lee at Intel Corporation for providing valuable comments on the machnie learning algorithm and data analysis on this work. This research was supported by the Basic Research Program ( NRF-2022R1A2B5B01001640 , NRF-2021R1A6A1A10039823 ) and Global Research and Development Center Program ( NRF-2018K1A4A3A01064272 ) through the National Research Foundation of Korea (NRF). Publisher Copyright: © 2022 Korean Physical Society

PY - 2022/9

Y1 - 2022/9

N2 - Compounds information such as Chemical Abstracts Service (CAS) registry number, hazards, and properties have been provided through Globally Harmonized System (GHS) based Material Safety Data Sheet (MSDS). This information can help users avoid hazardous compounds and handle chemicals in proper way. GHS specifies that hazards of compounds are categorized through animal testing (or in vivo testing), in vitro testing, epidemiological surveillance, and clinical trials. In this study, artificial intelligence (AI) is used to replace traditional approaches in predicting the toxicity of chemicals. A database of hazardous compounds is generated by data provided by the Ministry of Environment (ME), training and learning based on convolutional neural network (CNN) are carried out following data featurization. As a result, 90% of accuracy for CNN-based model is obtained using the image dataset. In contrast to the previous methods, the classification method based on CNN-based model in this study allows for the efficient discrimination of hazard chemicals without any additional tests.

AB - Compounds information such as Chemical Abstracts Service (CAS) registry number, hazards, and properties have been provided through Globally Harmonized System (GHS) based Material Safety Data Sheet (MSDS). This information can help users avoid hazardous compounds and handle chemicals in proper way. GHS specifies that hazards of compounds are categorized through animal testing (or in vivo testing), in vitro testing, epidemiological surveillance, and clinical trials. In this study, artificial intelligence (AI) is used to replace traditional approaches in predicting the toxicity of chemicals. A database of hazardous compounds is generated by data provided by the Ministry of Environment (ME), training and learning based on convolutional neural network (CNN) are carried out following data featurization. As a result, 90% of accuracy for CNN-based model is obtained using the image dataset. In contrast to the previous methods, the classification method based on CNN-based model in this study allows for the efficient discrimination of hazard chemicals without any additional tests.

KW - Artificial intelligence

KW - Classification

KW - Deep learning

KW - GHS

KW - Hazardous compounds

KW - MSDS

UR - http://www.scopus.com/inward/record.url?scp=85133850213&partnerID=8YFLogxK

U2 - 10.1016/j.cap.2022.06.003

DO - 10.1016/j.cap.2022.06.003

M3 - Article

AN - SCOPUS:85133850213

SN - 1567-1739

VL - 41

SP - 59

EP - 65

JO - Current Applied Physics

JF - Current Applied Physics

ER -

Methodology to classify hazardous compounds via deep learning based on convolutional neural networks

Abstract

Bibliographical note

Keywords

Access to Document

Fingerprint

Cite this