A Bibliometric Analysis of Research on Myofascial Pain Trigger Points
Description
The data for bibliometric analysis was sourced from the Web of Science Core Collection (WoSSC). The search was conducted on February 12, 2025, using the following search string: TS= (“myofascia” OR “fascia” OR “muscle fascia” OR “myofascial pain”) AND (“trigger points” OR “pain points” OR “sensitive points” OR “myofascial trigger points”). A total of 978 articles were retrieved. Article type was limited to "article," and other types were excluded. Only English-language articles were included. Data were manually screened for authenticity and reliability, with two researchers independently conducting the screening. In case of discrepancies, a third researcher was consulted for consensus. After manual screening, 699 articles were retained for analysis. Data retrieval and downloading were performed on the same day to minimize biases caused by daily database updates. The complete records were downloaded in file formats including plain text, BibTex, Tab delimited file, and Excel for further analysis.The retrieved data were imported into Citespace (version 6.2.R7 Advanced), VOSviewer (version 1.6.20), R (version 4.4.1) with the bibliometrix package, KH Coder (3b07d), Word Cloud Maker (https://wordart.com/), and Edraw Soft Online (https://mm.edrawsoft.cn/app/create) for analysis and visualization. VOSviewer was used to visualize co-authorship, co-occurrence, and co-citation networks, which are useful for understanding the collaboration and relationships among authors, institutions, and countries. Different nodes represent authors, countries, institutions, journals, and keywords, with node size corresponding to citation or co-citation frequency. Edges between nodes indicate collaboration or co-occurrence. The color of nodes and edges reflects clusters or the associated year. CiteSpace, a tool designed for the visualization and analysis of scientific literature networks, was employed to visualize keyword co-occurrence networks and detect emerging research trends through co-citation burst detection. We used CiteSpace to perform keyword and reference clustering analysis, as well as temporal distribution, to identify the evolution of research dynamics in this field. Topic modeling, a technique from natural language processing (NLP), was used to discover potential themes in the literature. Latent Dirichlet Allocation (LDA) is a popular method for topic modeling that can effectively handle large amounts of unstructured text data. By analyzing word frequency co-occurrences within documents, LDA calculates the relevance of each article to a particular theme. The number of topics was determined using various criteria, such as cosine distance, Kullback-Leibler divergence, and model coherence.