In this cross-sectional bibliometric analysis, PubMed results for the search query “Disorders of Consciousness” were reviewed. Abstracts and full texts were separately analysed using a text mining algorithm which searched for anatomical brain terminology. 

We evaluated impact on the results if the analyses were based on abstracts or full texts or topic models (non-negative matrix factorisation was used to create subgroups of each collection based on their key topics). The number of appearances of brain region in each group was recorded. Terms were ranked by their overall frequency across all documents and concordance was measured. Graphical analysis was performed to explore the relationships between the brain regions mentioned. The PageRank algorithm (used by Google to rank websites) was used to rank global importance of the regions.

The PubMed search yielded 14,945 abstracts and 2178 (14.6%) freely available full texts. The topic-modelled subgroups contained 2440 abstracts and 367 full texts.  Bibliometric analysis across all documents yielded a Kendall’s concordance value of 0.8380 (p = 0.000002). 

Network analysis revealed 4 clusters: cluster 1 has 20 members with the most important member being the insular cortex [PageRank =0.167]; cluster 2 has 29 members with most important member being amygdala [PageRank =0.0199]; cluster 3 has 10 members with the most important member being thalamus [PageRank = 0. 0205]; cluster 4 has 19 members with the cingulate cortex being most important member [PageRank = 0. 0.020]).

Visualisation is available at: (https://mango117.github.io/BMedScDOC/network.html)