To develop a traditional Chinese medicine (TCM) knowledge graph (KG) for diabetic retinopathy (DR) diagnosis and treatment by integrating literature and medical records, thereby enhancing TCM knowledge accessibility and providing innovative approaches for TCM inheritance and DR management.

First, a KG framework was established with a schema-layer design. Second, high-quality literature and electronic medical records served as data sources. Named entity recognition was performed using the ALBERT-BiLSTM-CRF model, and semantic relationships were curated by domain experts. Third, knowledge fusion was mainly achieved through an alias library. Subsequently, the data layer was mapped to the schema layer to refine the KG, and knowledge was stored in Neo4j. Finally, exploratory work on intelligent question answering was conducted based on the constructed KG.

In Neo4j, a KG for TCM diagnosis and treatment was constructed, incorporating 6 types of labels, 5 types of relationships, 5 types of attributes, 822 nodes, and 1,318 relationship instances. This systematic KG supports logical reasoning and intelligent question answering. The question answering model achieved a precision of 95%, a recall of 95%, and a weighted F1-score of 95%.

This study proposes a semi-automatic knowledge-mapping scheme to balance integration efficiency and accuracy. Clinical data-driven entity and relationship construction enables digital dialectical reasoning. Exploratory applications show the KG’s potential in intelligent question answering, providing new insights for TCM health management.

To develop a classifier for traditional Chinese medicine (TCM) syndrome differentiation of diabetic retinopathy (DR), using optimized machine learning algorithms, which can provide the basis for TCM objective and intelligent syndrome differentiation.

Collated data on real-world DR cases were collected. A variety of machine learning methods were used to construct TCM syndrome classification model, and the best performance was selected as the basic model. Genetic Algorithm (GA) was used for feature selection to obtain the optimal feature combination. Harris Hawk Optimization (HHO) was used for parameter optimization, and a classification model based on feature selection and parameter optimization was constructed. The performance of the model was compared with other optimization algorithms. The models were evaluated with accuracy, precision, recall, and F1 score as indicators.

Data on 970 cases that met screening requirements were collected. Support Vector Machine (SVM) was the best basic classification model. The accuracy rate of the model was 82.05%, the precision rate was 82.34%, the recall rate was 81.81%, and the F1 value was 81.76%. After GA screening, the optimal feature combination contained 37 feature values, which was consistent with TCM clinical practice. The model based on optimal combination and SVM (GA_SVM) had an accuracy improvement of 1.92% compared to the basic classifier. SVM model based on HHO and GA optimization (HHO_GA_SVM) had the best performance and convergence speed compared with other optimization algorithms. Compared with the basic classification model, the accuracy was improved by 3.51%.

HHO and GA optimization can improve the model performance of SVM in TCM syndrome differentiation of DR. It provides a new method and research idea for TCM intelligent assisted syndrome differentiation.