Digital Morphology Meets Chemometrics: Multi-Sensor Combination for Rapid Quality Grading and Geographical Origin Discrimination of Atractylodes lancea Rhizome

The dried rhizome of Atractylodes lancea (RAL) is a widely used traditional Chinese medicine (TCM). Its quality evaluation and origin authentication have long relied on time-consuming chromatographic methods, which are poorly suited for rapid, on-site decisions in commercial supply chains, and existing studies generally focus on isolated morphological indicators without systematic digital characterization and practical on-site grading tools. Guided by the traditional empirical knowledge of “Bianzhuang Lunzhi”, which holds that external morphological traits can reflect the internal quality of TCM, this study presents the first systematic multi-dimensional digital characterization of RAL morphological traits using an integrated multi-sensor approach and quantitatively explores the underlying correlations between digital traits and key bioactive constituent contents. Nighty samples from three major producing regions were analyzed. Significant correlations were observed between odor indices, color parameters, density, oil cavity area ratio and bioactive component contents in the authentic Maoshan-sourced RAL (p < 0.01 or p < 0.05). Such associations were absent in the emerging regions (Dabie and Qin−Ba Mountains). A three-grade quality classification system based on density thresholds (Grade A: ≥0.73 g/cm3; B: 0.58–0.73 g/cm3; C: <0.58 g/cm3) was established specifically for Maoshan RAL. Additionally, an electronic nose-based classification model was constructed for geographical origin discrimination, which delivered reliable and robust classification performance in external validation with independent blind test samples. This work provides practical, low-cost tools for rapid quality grading and origin identification of RAL. The proposed trait-driven analytical strategy offers a generalizable framework for the quality control of other complex herbal medicines.