Good Formulas: Empirical Evidence in Mid-Imperial Chinese Medical Texts, written by Ruth Yun-ju Chen

BackgroundTraditional Chinese medicine (TCM) in China and Kampo medicine in Japan are representative forms of traditional medicine in East Asia, in which mineral medicines constitute an indispensable source of therapeutic agents. Due to concerns about toxicity and safety, the frequency of mineral medicine use in modern times has declined compared with historical practice. Existing research on mineral medicines in traditional medicine has predominantly focused on toxicity and safety issues in an international context, overviews of historical mineral medicines in Japan, or records of specific mineral medicines in single ancient medical texts in China. However, there is a lack of comparative studies spanning multiple countries, historical periods, and a wide range of historical medical literature.MethodsThis study utilized modern computational techniques and a self-constructed database, the Chinese–Japanese Traditional Medical Literature Corpus. Based on the medical history of China and Japan, 56 representative historical medical texts from the period 219–1863 were selected, from which data on mineral medicines were extracted. The methods of prescription metrology and data mining were applied to analyze the co-occurring medicines in prescriptions containing mineral medicines, while word frequency statistics were used to examine the conditions treated by these medicines. Data cleaning, statistical analysis, and visualization were performed using Microsoft Excel and Python scripts.ResultsThe “Prescription” category of historical medical texts is the primary source of mineral medicine data for both countries. A total of 106 mineral medicines were recorded in Chinese historical texts, compared with 100 in Japanese historical texts, with 97 mineral medicines shared between the two. Based on a cation-based classification system, the mineral medicines documented in the historical texts of China and Japan were divided into 16 categories; all 16 were found in Chinese texts, while Japanese texts contained 14 categories. The top three categories of mineral medicines by number of occurrences were the same in both countries, though their ranking order differed slightly. For pharmacological analysis, mercury- and mercury-compound-based mineral medicines (hereafter referred to as mercury-based mineral medicines) were selected due to their high toxicity and high number of occurrences. Historical Chinese texts recorded 189 medical conditions treated with mercury-based mineral medicines or compound prescriptions containing them, while Japanese texts recorded 98 such conditions, with two conditions unique to Japan. Six conditions were identified as core conditions strongly associated with mercury-based mineral medicines in both countries. Historical Chinese texts documented 257 co-occurring medicines with mercury-based mineral medicines, while Japanese texts recorded 240, with 17 species unique to Japan. Twelve co-occurring medicines were identified as core drugs strongly paired with mercury-based mineral medicines in both countries. Gypsum was selected for further pharmacological analysis, as it is included in both modern authoritative pharmacopoeias and ranks just below salt in number of occurrences in historical texts of both countries. Historical Chinese texts documented 429 co-occurring medicines with gypsum, while Japanese texts recorded 168. The core medicines strongly paired with gypsum showed minimal differences between the two countries. The top five conditions most strongly associated with gypsum, in terms of number of occurrences, were the same in both Chinese and Japanese historical texts, although their ranking varied slightly. Compared with the indications recorded in modern pharmacopoeias and medical literature of both countries, the descriptions of gypsum-related core conditions in historical texts were more diverse and detailed.ConclusionThe classification of historical medical texts in this study is based on the characteristics of their content. The types of historical texts serving as data sources for mineral medicines are similar in China and Japan, and the recorded mineral medicine species, compound types, and frequently recorded varieties also show a high degree of similarity, indicating that Kampo medicine in Japan extensively absorbed the theoretical foundations of mineral medicines from TCM. However, the higher number of occurrences of sodium compound–based mineral medicines in Japan, as well as the differences in the occurrence probabilities of commonly recorded mineral medicines between the two countries, to some extent reflect the localization tendencies of mineral medicine use in Kampo medicine. Mercury-based mineral medicines and gypsum documented in historical Chinese and Japanese medical texts showed minimal differences in associated conditions and co-occurring medicines. Many mercury-based mineral medicines shared generalizable features, highlighting the research significance and value of distinguishing mineral medicines by compound type to reveal overarching pharmacological trends. The comparison of gypsum’s principal therapeutic indications between historical and modern records revealed a clear trend toward a narrower application scope for mineral medicines in the modern era. From the perspective of preserving and inheriting traditional mineral medicine knowledge, a large amount of mineral medicine knowledge in historical Chinese and Japanese medical texts remains to be explored. Furthermore, research supported by objective data—such as analyses of the pharmacological effects of co-occurring medicines related to mineral medicines and studies on the associations between these co-occurring medicines and their related conditions—remains urgently needed.

This book is a wide-ranging study of vernacularization in East Asia, and for this purpose East Asia includes not only China, Japan, Korea, and Vietnam but also other societies that no longer exist, such as the Tangut and Khitan empires. It takes the reader from the early centuries of the Common Era, when the Chinese script was the only form of writing and Chinese Buddhist, Confucian, and medical texts spread throughout East Asia, through the centuries when vernacular scripts evolved, right up to the end of the nineteenth century when nationalism created new roles for vernacular languages and vernacular scripts. Through an examination of oral approaches to Chinese texts, it shows how highly valued Chinese texts came to be read through the prism of the vernaculars and ultimately to be translated. This long process has some parallels with vernacularization in Europe, but a crucial difference is that literary Chinese was, unlike Latin, not a spoken language. As a consequence, people who spoke different East Asian vernaculars had no means of communicating in speech, but they could communicate silently by means of written conversation in literary Chinese; a further consequence is that within each society Chinese texts assumed vernacular garb: in classes and lectures, Chinese texts were read and declaimed in the vernaculars. What happened in the nineteenth century and why are there still so many different scripts in East Asia? How and why were Chinese texts dethroned and what replaced them? These are some of the questions addressed in this book.

The Chinese text classification task is challenging compared to tasks based on other languages such as English due to the characteristics of the Chinese text itself. In recent years, some popular methods based on deep learning have been used for text classification, such as the convolutional neural network (CNN) and the long short-term memory (LSTM) network. However, some problems are still encountered when classifying Chinese text. For example, important but obscure context information in Chinese text is not easily extracted. To improve the effect of Chinese text classification, we propose a novel classification model in this paper named the hierarchical comprehensive context modeling network (HCCMN) that can extract more comprehensive context. Our approach aims to extract contextual information and integrate it with the original input and then extract hierarchically more context, spatial information and high-weight local features from the integrated results. In addition, our method can remember long-term historical obscure information. Since Chinese radiology texts are complicated and difficult to obtain, we collected a Chinese radiology medical text dataset (CIRTEXT) containing more than 56,000 real-world data samples to verify the effect of this work. We conducted experiments on four datasets and showed that our HCCMN performs at state-of-the-art levels on three selected evaluation metrics compared to baselines. We present promising results showing that our hierarchical context modeling network extracts useful context from Chinese text more effectively and comprehensively.

Deep neural network-based recognition of entities in Chinese online medical inquiry texts

Named entity annotation means an entity that needs to be labeled in a prediction sequence on a given text sequence. Labeling high-quality medical entities from Chinese medical texts plays an important role in named entity recognition, and construction of medical knowledge graph. Named entity annotation in medical texts is the premise of the full-supervised and semi-supervised named entities recognition. The current mainstream named entity annotation require a lot of manpower on the corpus labeling, which is laborious and time consuming. For medical entities widely distributed in Chinese medical texts, in this paper, we propose a small number of manually labeled medical entities to autonomously learn medical text features, and iteratively generating new labeled entities. The model automatically iterates the annotations from the original medical text collection to be processed and generates a valid medical entity. The autonomously medical entity labeling work makes it easy to label Chinese medical texts. This framework is tested on real Chinese medical records, and the experimental results show that the method can effectively identify the entities, and has certain practical value.

Recently, a lot of Chinese patients consult treatment plans through social networking platforms, but the Chinese medical text contains rich information, including a large number of medical nomenclatures and symptom descriptions. How to build an intelligence model to automatically classify the text information consulted by patients and recommend the correct department for patients is very important. In order to address the problem of insufficient feature extraction from Chinese medical text and low accuracy, this paper proposes a dual channel Chinese medical text classification model. The model extracts feature of Chinese medical text at different granularity, comprehensively and accurately obtains effective feature information, and finally recommends departments for patients according to text classification. One channel of the model focuses on medical nomenclatures, symptoms and other words related to hospital departments, gives different weights, calculates corresponding feature vectors with convolution kernels of different sizes, and then obtains local text representation. The other channel uses the BiGRU network and attention mechanism to obtain text representation, highlighting the important information of the whole sentence, that is, global text representation. Finally, the model uses full connection layer to combine the representation vectors of the two channels, and uses Softmax classifier for classification. The experimental results show that the accuracy, recall and F1-score of the model are improved by 10.65%, 8.94% and 11.62% respectively compared with the baseline models in average, which proves that our model has better performance and robustness.

The question of national-cultural conditioning of manipulative influence in Russian and Chinese medical commercial advertising texts are considered. A model of speech manipulative strategy in the advertising text is considered as the interaction of the main strategies: cognitive, prescriptive, emotive and accessory aesthetic. The novelty of the research consists in the following. (1) The author notes cultural differences between the Russian and Chinese texts when implementing manipulative influence, due to ethno-speech bans of Chinese linguistic culture, and at the same time states a significant coincidence of the communicative speech techniques in the Russian and Chinese advertising texts. (2) The notion of “manipulative effect” is clarified. The author sets out his own understanding of the interaction between the author of manipulating text and its recipient. Methods of multidimensional comparative linguistic and cultural studies are improved. The relevance of the work is based primarily on the relevance of the results of cross-cultural communicative studies, including advertising, closely related to the development of international business communication. The results of the study are significant for the teaching of stylistics of the Russian language and practical Russian language in foreign (Chinese) audience, as well as in the training of specialists in advertising and intercultural communication.

In this paper, a survey is done to introduce the named entity recognition task in Chinese medical text and its practical significance. First, the existing datasets for the named entity recognition task of Chinese medical text are presented, then the survey is given on the algorithms for this task, mainly from the perspectives on matching and sequence labeling. Finally, the future development of named entity recognition in Chinese medical text is discussed.

The integration of lexicon information into character-based models is a hot topic in Chinese Named Entity Recognition(NER) research. Most methods only utilize information from a single lexicon which is usually a general lexicon. However, In the Chinese medical text scenario, due to the large amount of medical terminology, a single lexicon, especially a general lexicon, offers little performance improvement to the Chinese NER. In this paper, we propose a Multi-source Lexicon Information Fusion method for Named Entity Recognition in Chinese Medical Text(MLNER) which can utilize information from both general and medical lexicons. Considering the small medical annotated corpus, we combine the model with the pre-trained model to improve the performance of the model on small datasets by exploiting the rich representation capability of the pre-trained model. Experiments show that our method can effectively improve the performance of NER in Chinese medical text. Our model is also applicable to Chinese NER tasks in other domain specific fields, with good scalability and application value.

Benjamin A. Elman (Editor). <i>Antiquarianism, Language, and Medical Philology: From Early Modern to Modern Sino-Japanese Medical Discourses</i>. (Sir Henry Wellcome Asian Series, 12.) viii + 232 pp., figs., index. Leiden/Boston: Brill, 2015. $135 (cloth).

Addressing the privacy protection and data sharing issues in Chinese medical texts, this paper introduces a federated learning approach named FLCMC for Chinese medical text classification. The paper first discusses the data heterogeneity issue in federated language modeling. Then, it proposes two perturbed federated learning algorithms, FedPA and FedPAP, based on the self-attention mechanism. In these algorithms, the self-attention mechanism is incorporated within the model aggregation module, while a perturbation term, which measures the differences between the client and the server, is added to the local update module along with a customized PAdam optimizer. Secondly, to enable a fair comparison of algorithms' performance, existing federated algorithms are improved by integrating a customized Adam optimizer. Through experiments, this paper first conducts experimental analyses on hyperparameters, data heterogeneity, and validity on synthetic datasets, which proves that the proposed federated learning algorithm has significant advantages in classification performance and convergence stability when dealing with heterogeneous data. Then, the algorithm is applied to Chinese medical text datasets to verify its effectiveness on real datasets. The comparative analysis of algorithm performance and communication efficiency shows that the algorithm exhibits strong generalization ability on deep learning models for Chinese medical texts. As for the synthetic dataset, upon comparing with comparison algorithms FedAvg, FedProx, FedAtt, and their improved versions, the experimental results show that for data with general heterogeneity, both FedPA and FedPAP show significantly more accurate and stable convergence behavior. On the real Chinese medical dataset of doctor-patient conversations, IMCS-V2, with logistic regression and long short-term memory network as training models, the experiment results show that in comparison to the above three comparison algorithms and their improved versions, FedPA and FedPAP both possess the best accuracy performance and display significantly more stable and accurate convergence behavior, proving that the method in this paper has better classification effects for Chinese medical texts.

The Chinese medical sector has been somewhat lacking in knowledge graphs, a deficiency this study aims to address. By leveraging the prowess of the BERT pre-training model, a two-tier approach has been innovated that utilizes separate pre-trained encoders for both entity and relational models. These models are intricately linked: the output from the entity model seamlessly flows into the relational one, making it possible to adeptly extract entity relationships from Chinese medical texts. This research is anchored in the CMeIE dataset, sourced from the esteemed CHIP (China Health Information Processing) conference. This dataset stands as a recognized benchmark in evaluating Chinese medical texts. By harnessing this data, the methods have been rigorously tested and validated. The promising experimental results underscore the effectiveness of the approach in distilling relationships from Chinese medical literature. The implications of this research are profound. Beyond just enriching the Chinese medical domain, the boundaries of NLP technology are also being pushed. Potential applications are manifold: from constructing comprehensive Chinese medical knowledge graphs to assisting in early-stage medical diagnoses. This innovative approach not only addresses an existing gap but also sets the stage for future advancements in medical NLP.

In order to solve the problem of Chinese medical image text being missed and misdetected under the CTPN model, a new convolutional neural network DVNet based on the fusion of VGG convolutional neural network and DenseNet dense network was proposed. DVNet takes the first two layers of VGG network for deep feature extraction, and then connects DenseNet dense modules. Using the idea of feature reuse, the features of the front convolutional layer and the features of the back convolutional layer are output together. During post-processing, NMS is used to filter out redundant text boxes. In the Chinese medical text data set provided, three different networks, VGG, DenseNet and DVNet, were used to detect the text. The experimental results showed that the precision rate of DVNet were improved by 2%-3% compared with VGG and DenseNet.

Medical short text classification is of great significance to medical information extraction and medical auxiliary diagnosis. However, medical short texts face challenges such as sparse features, semantic ambiguity, and the specialized nature of the medical field, resulting in relatively low accuracy in short text classification. Taking into consideration the characteristics of medical short texts, this paper proposes a Chinese medical short text classification model based on DPECNN. First, ERNIE is utilized to learn text knowledge and information in order to enhance the model’s semantic representation capabilities. Then, the DPECNN model is employed to extract rich feature information, and the classification results are generated through a fully connected layer. In the case of DPCNN, it only considers deep-level contextual semantic information, overlooking the correlation of adjacent semantic information between channels. To address this, ECA channel attention is introduced to account for adjacent semantic information. The use of a self-normalizing activation function helps avoid the problem of vanishing gradients. To enhance the model’s robustness and generalization ability, the FGM adversarial training algorithm is employed to perturb the data. The F1 values achieved on the THUCNews, KUAKE-QIC, and CHIP-CTC datasets are 95.00%, 79.45%, and 82.81%, respectively.

To solve the problem of poor feature extraction ability of traditional text recognition methods in Chinese medical record text, this paper proposes a Res-RNN network for feature extraction based on residual error. Combined with residual characteristics, this network not only improves the depth of the network, but also ensures that there will be no degradation of the network, and strengthens the network's ability to extract Chinese character features. In the residual module, 1 x 1 convolution kernel is used to replace 3 x 3 convolution kernel, effectively reducing the parameters. Combined with feature maps of different scales, the feature information of Chinese characters at different levels is effectively utilized. According to the characteristics of Chinese characters, the vertical sensing field of the feature map is adjusted to retain more vertical fine-grained feature information, thus effectively improving the representational ability of the network. Experiments on actual Chinese medical record text image data set show that the accuracy of the proposed model is 4% higher than that of CRNN.