Search for a command to run...
AI in Anatomy and Medical Education This repository contains the data, search criteria, and analytical code for the perspective article: "Artificial Intelligence in Anatomy and Medical Education: Growth, Thematic Structure, Collaboration, and Global Inequality" Overview Artificial intelligence (AI) is increasingly shaping the landscape of anatomy-related medical education, from image-based learning and simulation to adaptive educational technologies and generative tools. However, the structure of this literature and the factors associated with its visibility remain insufficiently examined. Drawing on a Scopus-based dataset of 598 articles and reviews published between 2000 and 2026, this study uses descriptive, network-based, and exploratory inferential analyses to examine how the field has developed. We focus on three primary issues: Temporal growth Thematic and geographic structure Bibliometric correlates of visibility Repository Contents 1. Supplementary Material/ This folder contains the raw data and the exact parameters used for literature retrieval: Supplementary Material 1.docx (Search Strategy & Criteria) Contains the full, reproducible search criteria. Includes the specific Scopus search string used. Outlines the specific domains, sub-domains, and Boolean operators utilized to build the corpus at the intersection of Artificial Intelligence, Anatomy, and Medical Education. Supplementary Material 2.csv (Scopus Dataset) The raw data export from Scopus (exported March 21, 2026). Contains the 598 bibliographic records (articles and reviews) analyzed in the study. Includes metadata such as Authors, Author Affiliations, Year, Source Title, Document Type, Open Access status, and Citation Counts (Cited by). 2. src/ This folder contains the analytical code used to process the dataset and generate the results: code.R (Main Analysis Script) A comprehensive R script used for all data manipulation, descriptive statistics, and inferential modeling. It follows a structured pipeline: Basic Data Cleaning: Standardizing strings, cleaning missing values, and filtering by year criteria (2000–2026). Variable Derivation: Creating variables for article_age, citations_per_year, log-transformed citations, number of authors, and a simplified doc_type binary. Geographic Classification: Using regex and the countrycode package to extract the first author's country from the complex Scopus affiliation string. Collaboration Structure: Parsing multiple affiliations to determine single-country vs. international collaboration. Income-Group Mapping: Manually mapping the extracted countries to the World Bank 2025 income classifications (High income, Upper middle income, Lower middle income, Low income) and generating summary tables (Table 1). Descriptive Figures (ggplot2): Plotting annual publication output by document type (Figure 1a), top 10 contributing countries (Figure 1b), and top 10 source journals (Figure 1c). Exploratory Inferential Analysis: Fitting a multivariable linear regression model (lm(log_citations_per_year ~ ...)) to examine the association between citation visibility and factors like international collaboration, document type, open access, author count, and publication year (Table 2). Reproducibility and Requirements To run the src/code.R script, you will need an R environment with the following packages installed: install.packages(c("readr", "dplyr", "stringr", "tidyr", "ggplot2", "forcats", "broom", "scales", "tibble", "countrycode")) Note: The network visualizations for country co-authorship (Figure 2) and keyword co-occurrence (Figure 3) presented in the manuscript were constructed separately using VOSviewer, based on the bibliometric data derived from this corpus.