Chinese herbal pieces are an essential component of traditional Chinese medicine. Accurate identification and classification of these materials are crucial in clinical practice.

This study aims to enhance the recognition efficiency of Chinese herbal pieces using deep learning technology, while addressing the limitations of traditional manual classification methods in terms of both quality and efficiency.

A comprehensive dataset containing 201 types of Chinese herbal pieces was established. Based on Real-time Detection Transformer (RT-DETR), we designed and integrated a Feature-focused Diffusion Network (FDN), resulting in an improved model termed RT-DETR-FDN. The proposed FDN includes a Feature-focus Module and a feature diffusion mechanism, enabling the model to capture more extensive feature information from Chinese herbal pieces and diffuse it across multiple detection scales.

Experimental results show that RT-DETR-FDN achieved a precision of 0.925, a recall of 0.943, and an mAP50-95 of 0.851. In addition, the model was compared with representative You Only Look Once series models commonly used in object detection. Compared with these models, RT-DETR-FDN achieved higher recognition accuracy while maintaining a lightweight architecture.

This study integrates deep learning with traditional Chinese medicine, providing a more effective solution for the recognition of Chinese herbal pieces.

Murraya, a genus of shrubs and trees in the Rutaceae family, consists of approximately 9 species in China with significant medicinal and horticultural value. However, the phylogeny and taxonomy of Murraya species remain controversial, particularly with respect to Murraya exotica and M. paniculata.

This study aimed to provide insights into the taxonomy, phylogeny, and identification of Murraya.

In this study, the chloroplast (CP) genomes of 7 Murraya species were sequenced, assembled, and subjected to comparative and phylogenetic analyses.

The CP genomes of Murraya ranged from 158,573 to 160,817 bp in length and encoded 112 unique genes, including 78 protein-coding genes, 30 tRNA genes, and 4 rRNA genes. Similar to other angiosperms, the inverted repeat regions of the CP genomes exhibited lower sequence divergence than the single-copy regions, and coding regions were more conserved than noncoding regions. Comparative analysis identified several highly variable regions (eg, matK, ycf1, ndhI-ndhA, trnH-GUG-psbA, rpl32-trnL) that could serve as molecular markers for species identification in Murraya. Among these, the ycf1 gene was validated as a useful marker for distinguishing M. exotica from M. paniculata. Positive selection was detected in 10 genes, including rbcL, psaJ, ndhD, ndhF, rpl2, rpl20, ycf1, accD, ccsA, and rpl32. Phylogenetic analysis based on CP genomes supported the recognition of M. exotica and M. paniculata as independent species. Moreover, the phylogenetic trees indicated that Murraya is not monophyletic, with sect. Bergera showing a closer relationship to Clausena. Molecular dating results suggested that the diversification of M. paniculata, M. alata, and M. exotica occurred approximately 9.11 Mya (95% highest posterior density: 4.90-13.87 Mya).

These findings provide valuable CP genome data for clarifying the phylogenetic relationships between M. exotica and M. paniculata, and for advancing the study of DNA markers and the evolutionary history of Murraya.