Automated Image Label Extraction from Radiology Reports

This data set contains the raw data from Figures 3 and 4 of the manuscript "Automated Image Label Extraction from Radiology Reports - A Review"

The data used to create Figure 3 (co.autorships) provides Co-authorship links among authors of the studies included in the synthesis. The "nodes" tab contains information on each author, such as their modularity class and their degree (the number of colaborations with other authors). The "edges" tab contains information on the links between authors, such as the weight (the number of colaborations) of each link between two authors. A modularity algorithm (https://dx.doi.org/10.1088/1742-5468/2008/10/P10008) was used for community detection. This data was used to create Figure 3 of the manuscript.

The data used to create Figure 4 of the manuscript (wordcloud_counts) provides the number of occurences of named entities across the abstracts of the included studies. The entities were extracted using the spaCy model en_core_sci_lg, which is appropriate for processing biomedical, scientific and clinical text (https://spacy.io/). This data was used to create Figure 4 of the manuscript.