Deep belief network-based drug identification using near infrared spectroscopy

Huihua Yang,Yanchun Feng,Changqin Hu,Xipeng Pan,Lihui Yin,Shengke Yan,Xuebo Zhang,Baichao Hu

doi:10.1142/s1793545816300111

Abstract

Near infrared spectroscopy (NIRS) analysis technology, combined with chemometrics, can be effectively used in quick and nondestructive analysis of quality and category. In this paper, an effective drug identification method by using deep belief network (DBN) with dropout mechanism (dropout-DBN) to model NIRS is introduced, in which dropout is employed to overcome the overfitting problem coming from the small sample. This paper tests proposed method under datasets of different sizes with the example of near infrared diffuse reflectance spectroscopy of erythromycin ethylsuccinate drugs and other drugs, aluminum and nonaluminum packaged. Meanwhile, it gives experiments to compare the proposed method’s performance with back propagation (BP) neural network, support vector machines (SVMs) and sparse denoising auto-encoder (SDAE). The results show that for both binary classification and multi-classification, dropout mechanism can improve the classification accuracy, and dropout-DBN can achieve best classification accuracy in almost all cases. SDAE is similar to dropout-DBN in the aspects of classification accuracy and algorithm stability, which are higher than that of BP neural network and SVM methods. In terms of training time, dropout-DBN model is superior to SDAE model, but inferior to BP neural network and SVM methods. Therefore, dropout-DBN can be used as a modeling tool with effective binary and multi-class classification performance on a spectrum sample set of small size.

Highlights

Counterfeit and substandard drugs, cause great harms to people and require identication under pharmaceutical supervision
In terms of constructing classication model, Deconinck et al.[4] used decision tree to identify the spectra of Viagra and Cialis drugs, which obtained good classication accuracy; Storme et al.[5] adopted support vector machines (SVM) to do the drug identication of Near infrared spectroscopy (NIRS), the result of which was proved more accurate than that of decision tree
There are few researches on the spectroscopy multi-classication, for the multiclassication is more complex than binary classication, and the shallow structure of SVM and the two-layer back propagation (BP) neural network make the ability of its function mapping weak, the ability of which still remains to be investigated in the identication of multi-classication drugs

Summary

Introduction

Counterfeit and substandard drugs, cause great harms to people and require identication under pharmaceutical supervision. It is of great value to identify these differences in pharmaceutical supervision. Near infrared spectroscopy (NIRS) analysis technology can be used for quick and nondestructive detection of the drugs and has been widely used in the pharmaceutical industry,[1,2,3] for it has abundant frequency doubling and frequency vibration information in molecular groups. In terms of constructing classication model, Deconinck et al.[4] used decision tree to identify the spectra of Viagra and Cialis drugs, which obtained good classication accuracy; Storme et al.[5] adopted support vector machines (SVM) to do the drug identication of NIRS, the result of which was proved more accurate than that of decision tree. There are few researches on the spectroscopy multi-classication, for the multiclassication is more complex than binary classication, and the shallow structure of SVM and the two-layer back propagation (BP) neural network make the ability of its function mapping weak, the ability of which still remains to be investigated in the identication of multi-classication drugs

Methods

Results

Conclusion