A novel quality evaluation method for magnolia bark using electronic nose and colorimeter data with multiple statistical algorithms

Abstract

The bark of the magnolia tree, which includes both the aromatic and medicinal varieties known as magnolia
officinalis and magnolia officinalis, is often used in Chinese patent medicines and clinical prescriptions. The chemical
composition of Hou Po is a key indicator of its quality, which is in turn connected to its therapeutic efficacy; the site of origin
plays a pivotal role in determining its quality. An innovative approach to quickly, correctly, and fully determining where Hou Po
came from and what its key chemical components are is the goal of this research.
Methods: The magnolol and honokiol components were analysed using high performance liquid chromatography, while the
magnocurarine and magno- fluorine components were analysed using ultra-performance liquid chromatography. We found out
what was in the water-soluble extracts by using the cold soak technique. The colorimeter and E-nose were used to identify the Hou
Po samples' smell and colour, respectively.
In order to determine where Hou-Po came from and what chemicals were in its water-soluble extracts, we used a number of
statistical techniques to build discriminant models that relied on E-nose and colorimeter data. With a classification accuracy of
99.53%, the Random Forest classifier in conjunction with the ten-fold cross-validation approach outperformed the other models
tested. For each of the five chemical components, the correlation coefficient between experimental and predicted values was more
than 0.96. This research concludes that the electronic nose and colorimeter show promise as quantitative and qualitative tools for
assessing the quality of Chinese herbal remedies.