Identification of Multi-Class Drugs Based on Near Infrared Spectroscopy and Bidirectional Generative Adversarial Networks

Abstract

Drug detection and identification technology are of great significance in drug supervision and management. To determine the exact source of drugs, it is often necessary to directly identify multiple varieties of drugs produced by multiple manufacturers. Near-infrared spectroscopy (NIR) combined with chemometrics is generally used in these cases. However, existing NIR classification modeling methods have great limitations in dealing with a large number of categories and spectra, especially under the premise of insufficient samples, unbalanced samples, and sensitive identification error cost. Therefore, this paper proposes a NIR multi-classification modeling method based on a modified Bidirectional Generative Adversarial Networks (Bi-GAN). It makes full utilization of the powerful feature extraction ability and good sample generation quality of Bi-GAN and uses the generated samples with obvious features, an equal number between classes, and a sufficient number within classes to replace the unbalanced and insufficient real samples in the courses of spectral classification. 1721 samples of four kinds of drugs produced by 29 manufacturers were used as experimental materials, and the results demonstrate that this method is superior to other comparative methods in drug NIR classification scenarios, and the optimal accuracy rate is even more than 99% under ideal conditions.

A Concise Review on Recent Developments of Machine Learning for the Prediction of Vibrational Spectra

“…Considering the structure–spectrum bidirectional relationship, it will in principle also be possible to make use of well-developed ML strategies of language translation (e.g., seq2seq, attention, transformers). Studies − applying generative adversarial networks to extract descriptors for the Raman spectrum predictions are also known, and same NN structure can potentially be used in the aspect of molecular representations. Besides, ΔML (an algorithm that learns from the corrections from low-level theory to high-level theory), which has been applied in other fields of computational chemistry (e.g., force field development), can also be utilized for achieving high accuracy. − …”

Exploring the Steps of Infrared (IR) Spectral Analysis: Pre-Processing, (Classical) Data Modelling, and Deep Learning

“…Zheng and his team suggested a tweaked bidirectional generative adversarial network (BI-GAN) approach for classifying near-infrared spectra of drugs. The findings show that models based on CNN performed better than the stacked autoencoder, which had the highest error rates in estimating soil properties [106].…”

A Review of Machine Learning for Near-Infrared Spectroscopy

“…Zheng et al proposed a modified bidirectional generative adversarial network (BI-GAN) method for classifying the near-infrared spectra of drugs, given the dilemma of insufficient samples within a class and sample imbalance between classes. The results demonstrate that CNN-based modeling outperformed the stacked autoencoder, which had the highest error rates when estimating soil characteristics [155]. Yang et al proposed machine learning algorithms combined with NIR spectroscopy to accurately distinguish cumin and fennel from three different regions, Turpan, Yumen, and Dezhou, with all model parameters remaining unchanged [156].…”