Publication Trends in Artificial Intelligence and Explainable Artificial Intelligence for Medical Image Analysis: An OpenAlex Bibliometric Dataset (2020–2025)
Description
This dataset contains bibliometric records retrieved from the OpenAlex scholarly database for publications related to Artificial Intelligence (AI) and Explainable Artificial Intelligence (XAI) in medical image analysis published between 2020 and 2025. Publications were identified using predefined keyword-based search strategies covering AI methodologies. The AI query combined the terms: (artificial intelligence, machine learning, deep learning, convolutional neural network (CNN), transformer, vision transformer, foundation model, vision-language model, and large language model). The XAI query combined: (explainable artificial intelligence, explainable AI, explainability, interpretability, interpretable, XAI, Grad-CAM, SHAP, LIME, saliency, attention map, and class activation map). Each query was combined with medical imaging terms (medical imaging, radiology, medical image analysis, MRI, CT, PET, ultrasound, mammography, chest X-ray, pathology, and histopathology) using the Boolean expression (AI/XAI terms) AND (medical imaging terms). Duplicate records were removed using unique OpenAlex identifiers. After duplicate records were removed, a total of 67,116 and 7,059 publications were retained and analyzed for AI and XAI publications, respectively. The dataset includes annual publication counts and processed outputs used to evaluate publication trends and the relative growth of explainable AI research within medical image analysis. This repository also includes the Python source code used to query the OpenAlex API, retrieve publication records, perform deduplication, generate annual publication statistics, and create all figures presented in the accompanying analysis. The scripts are provided to ensure full reproducibility and reuse of the bibliometric workflow. An OpenAlex API Key is required.
Files
Steps to reproduce
1. Download the Python analysis script. 2. Install the required Python packages l (e.g., pyalex, pandas, matplotlib, numpy, openpyxl, and other dependencies). 3. Run the Python script. 4. The script automatically queries the OpenAlex database using the predefined search terms for: Artificial Intelligence (AI) in medical image analysis and Explainable Artificial Intelligence (XAI) in medical image analysis. 5. Ensure to obtain OpenAlex API Key. The script requires this key to successfully run. 6. Publication records are retrieved for each year from 2020 through 2025. 7. Duplicate publications are removed using their unique OpenAlex identifiers to ensure that each publication is counted only once within each research category. 8. The script aggregates the annual publication counts and calculates the Explainability Attention Rate, defined as: Explainability Attention Rate = (Number of AI publications/Number of XAI publications) ×100 9. The processed datasets are exported as CSV/Excel files, and all figures are automatically generated and saved.
Institutions
- University of Illinois Urbana-ChampaignIllinois, Urbana