TCMPR: TCM Prescription Recommendation Based on Subnetwork Term Mapping and Deep Learning

Abstract

Traditional Chinese medicine (TCM) has played an indispensable role in clinical diagnosis and treatment. Based on a patient’s symptom phenotypes, computation-based prescription recommendation methods can recommend personalized TCM prescription using machine learning and artificial intelligence technologies. However, owing to the complexity and individuation of a patient’s clinical phenotypes, current prescription recommendation methods cannot obtain good performance. Meanwhile, it is very difficult to conduct effective representation for unrecorded symptom terms in an existing knowledge base. In this study, we proposed a subnetwork-based symptom term mapping method (SSTM) and constructed a SSTM-based TCM prescription recommendation method (termed TCMPR). Our SSTM can extract the subnetwork structure between symptoms from a knowledge network to effectively represent the embedding features of clinical symptom terms (especially the unrecorded terms). The experimental results showed that our method performs better than state-of-the-art methods. In addition, the comprehensive experiments of TCMPR with different hyperparameters (i.e., feature embedding, feature dimension, subnetwork filter threshold, and feature fusion) demonstrate that our method has high performance on TCM prescription recommendation and potentially promote clinical diagnosis and treatment of TCM precision medicine.

Traditional Chinese medicine diagnostic prediction model for holistic syndrome differentiation based on deep learning

“…Currently, deep learning is undergoing rapid development in TCM. 34 , 35 , 36 The most common application involves predicting various TCM syndromes under a single disease, which significantly limits the model's extrapolation and fails to encompass the TCM holistic syndromes. 37 The main challenge for TCM text-based deep learning is that current models are unlikely to be able to accommodate TCM holistic syndromes diagnostic methods, and are limited by standards for data processing and artificial intelligence technology of TCM syndrome characteristics.…”

Machine learning in TCM with natural products and molecules: current status and future perspectives

“…In their approach, transformer was used for TCM prescription generation, while the GAN model aims to augment the training set to further enhance the overall system performance by reducing overfitting effect. Dong [ 86 ] proposed a TCM prescription recommendation based on subnetwork term mapping and deep learning. They used TCM clinical case data to construct a natural product-symptom-related knowledge graph, constructed a symptom network by combining a meta path method and knowledge graph, proposed a subnetwork-based symptom term mapping method, utilized CNN as the train model, and finally output the prediction probability of each natural product to obtain the recommended prescription [ 86 ].…”

“…Dong [ 86 ] proposed a TCM prescription recommendation based on subnetwork term mapping and deep learning. They used TCM clinical case data to construct a natural product-symptom-related knowledge graph, constructed a symptom network by combining a meta path method and knowledge graph, proposed a subnetwork-based symptom term mapping method, utilized CNN as the train model, and finally output the prediction probability of each natural product to obtain the recommended prescription [ 86 ]. Dengzhan Shengmai capsule is a patented TCM preparation for the secondary prevention of stroke.…”

Analysis of the Real-World Application Laws for Zhengan Xifeng Decoction in the Treatment of Hypertension Based on Machine Learning

“…At present, studies on the combination of the micromechanisms of herbs and clinical data are lacking. Dong et al [13] constructed a mapping method for TCM prescription recommendations based on symptom term network. Shu et al [14] used network diversity to analyse protein-protein interactions in TCM symptom clusters to elucidate the potential molecular network mechanism of symptom phenotypes.…”

“…A convolutional neural network (CNN) can respond to a certain range of neurons and generate arti cial neural network. CNN is more suitable for TCM prescription feature processing [13] . CNN exhibits excellent performance for the two-dimensional structure processing of input data and includes fewer parameters than other feedforward neural networks.…”