Leaf-Face Dendrobium Classifier Based on an Integrated Electrochemical Tongue and Machine Learning

Botanical sourcing seriously impacts the safety and potency of herbal medicines, restricting the development of the traditional Chinese medicinal industry. Rapid and convenient identification of plant resources is important to address this problem. Herein, we innovated a portable, intelligent, and integrated platform, termed the Smart Electronic Tongue (SET), for right recognizing Dendrobium bonsai of different subspecies and origins. The device is miniaturized with a hollow microneedle array for leaf-face extraction, a press-pumping pipeline for on-demand sample sap suction, plus a five-electrode chip for electrochemical readout. Differential pulse voltammograms on three simplexes of self-assembled monolayers as well as their multiplexed configurations are specialized to generate high-dimensional fingerprinting of Dendrobium subtypes. Taking advantage of machine learning algorithms, the platform achieves automatic authentication over eight varieties with a discriminatory accuracy of 95%. Given this, we anticipate that such a bionic sensing setup would offer a versatile solution toward the classification of diverse officinals at the point of need.