Building and Utilizing a Knowledge Graph for Randomized Controlled Trials Investigating the Effects of Acupuncture

Abstract

Background: Acupuncture clinicians and researchers are required to read numerous RCT papers for their scientific work. However, time is precious, particularly for clinicians who must balance both clinical practice and research. Having well-structured knowledge at their disposal would significantly save time and improve efficiency. Therefore, discovering knowledge from existing achievements, accurately understanding the effects of acupuncture, and drawing experiences from current literature necessitate the organization and mining of acupuncture scientific papers.

Methods: We created an Excel spreadsheet to organize the acupuncture RCT literature from the AcuEBase V1.0 database. The spreadsheet included fields such as acupuncture points, needling techniques. We applied relevant standards to ensure the consistency and standardization of the fields.To measure the similarity between entities, we employed the DICE coefficient. The DICE coefficient calculates the similarity between two entities based on the shared features and the total number of features in both entities. This similarity measure allowed us to assess the relationships and connections between different entityin the dataset.

Results: We present our efforts on constructing the first publicly available Acupuncture EffectKnowledge Graph, denoted as AEKG. AEKG includes nodes representing medical entities in clinical trials (e.g., Literature, Disease and Trail), and edges representing the relations among these entities (e.g., Literaturehas aTrial). Our analysis demonstrates the potential utilities of AEKG in various applications such as similarity search.

Conclusions: AEKG has shown promising application scenarios. However, it is important to exercise caution when interpreting the results of the analysis due to the quality of RCT literature. These results have not been clinically validated and should be approached with caution.

Note:
Funding declaration: This study was supported by the Science and Technology Innovation Project administered by the China Academy of Chinese Medical Science（CI2021A05307).

Conflict of Interests: The author stated that they have no conflicts of interest related to the research, authorship, and/or publication of this article.

Keywords: Acupuncture effect, Knowledge graph, Knowledge retrieval