Deep learning-enabled mobile application for efficient and robust herb image recognition.

With the increasing popularity of herbal medicine, high standards of the high quality control of herbs becomes a necessity, with the herb recognition as one of the great challenges. Due to the complicated processing procedure of the herbs, methods of manual recognition that require chemical materials and expert knowledge, such as fingerprint and experience, have been used. Automatic methods can partially alleviate the problem by deep learning based herb image recognition, but most studies require powerful and expensive computation hardware, which is not friendly to resource-limited settings. In this paper, we introduce a deep learning-enabled mobile application which can run entirely on common low-cost smartphones for efficient and robust herb image recognition with a quite competitive recognition accuracy in resource-limited situations. We hope this application can make contributions to the increasing accessibility of herbal medicine worldwide.

"Possible mechanisms underlying treatment of Alzheimer's disease with Traditional Chinese Medicine: active components, potential targets and synthetic pathways of Bulao Elixir".

OBJECTIVE: To elucidate the mechanisms underlying the treatment of Alzheimer's disease (AD) with Traditional Chinese Medicine (TCM), by examining the active components, potential targets and synthetic pathways of Bulao Elixir (BLE). METHODS: The Absorption, Distribution, Metabolism, Excretion (ADME) / Toxicology (T) calculation was used to screen the active components of Bulao Elixir. Based on the TCM Systems Pharmacology Analysis Platform (TCMSP database) and a text mining tool (GoPubMed database), we predicted and screened the active components of Bulao Elixir and its therapeutic targets for AD. Using the Database for Annotation, Visualization and Integrated Discovery (DAVID), we obtained the targets for AD. Cytoscape software was used to establish a network map of the active component-target and target-pathway of Bulao Elixir. Gene function, related biological processes and signaling pathways were analyzed using the DAVID database. RESULTS: Twelve active components were selected from 196 components of Bulao Elixir. Among 2209 targets, 102 effective targets were selected, and 30 important targets were identified via matching with the disease targets. After further analysis, 14 core targets were identified. Enrichment analysis revealed that most of these important targets were involved in multiple biological processes, including apoptosis, inflammatory reactions, and cell regulation cycles. The synthetic pathways for AD treatment were identified after analyzing and confirming the relevant pathways, providing potentially useful information for diagnosis and treatment methods for AD. CONCLUSION: The current study elucidated the potential treatment mechanisms of Bulao Elixir in AD using network pharmacology, providing a foundation for further clarification of its treatment targets.

Integrated Analysis of the Mechanisms of Da-Chai-Hu Decoction in Type 2 Diabetes Mellitus by a Network Pharmacology Approach.

BACKGROUND: The incidence of type 2 diabetes mellitus (T2DM) has increased year by year, which not only seriously affects people's quality of life, but also imposes a heavy economic burden on the family, society, and country. Currently, the pathogenesis, diagnosis, and treatment of T2DM are still unclear. Therefore, exploration of a precise multitarget treatment strategy is urgent. Here, we attempt to screen out the active components, effective targets, and functional pathways of therapeutic drugs through network pharmacology with taking advantages of traditional Chinese medicine (TCM) formulas for multitarget holistic treatment of diseases to clarify the potential therapeutic mechanism of TCM formulas and provide a systematic and clear thought for T2DM treatment. METHODS: First, we screened the active components of Da-Chai-Hu Decoction (DCHD) by absorption, distribution, metabolism, excretion, and toxicity (ADME/T) calculation. Second, we predicted and screened the active components of DCHD and its therapeutic targets for T2DM relying on the Traditional Chinese Medicine Systems Pharmacology Analysis Platform (TCMSP database) and Text Mining Tool (GoPubMed database), while using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) to obtain T2DM targets. Third, we constructed a network of the active component-target, target-pathway of DCHD using Cytoscape software (http://cytoscape.org/,ver.3.5.1) and then analyzed gene function, related biological processes, and signal pathways through the DAVID database. RESULTS: We screened 77 active components from 1278 DCHD components and 116 effective targets from 253 ones. After matching the targets of T2DM, we obtained 38 important targets and 7 core targets were selected through further analysis. Through enrichment analysis, we found that these important targets were mainly involved in many biological processes such as oxidative stress, inflammatory reaction, and apoptosis. After analyzing the relevant pathways, the synthetic pathway for the treatment of T2DM was obtained, which provided a diagnosis-treatment idea for DCHD in the treatment of T2DM. CONCLUSIONS: This article reveals the mechanism of DCHD in the treatment of T2DM related to inflammatory response and apoptosis through network pharmacology, which lays a foundation for further elucidation of drugs effective targets.