Explainable AI (XAI) and Interpretable Machine Learning (IML) in Healthcare Dataset
Description
Explainable AI (XAI) and Interpretable Machine Learning (IML) in Healthcare Dataset was compiled from the Web of Science database, encompassing a comprehensive collection of 5,083 research articles focused on AI explainability in healthcare. The dataset includes various document types, such as Articles, Review Articles, Early Access Papers, and Proceeding Papers, all in English. Spanning from 1986 to April 2023, the dataset features key attributes, including Authors, Author Full Names, Article Title, Document Type, Author Keywords, Keywords Plus, Abstract, Publication Year, DOI Link, WoS Categories, and Research Areas. This dataset has been created to provide essential insights into the utilization of XAI and IML in healthcare contexts. In our work [1], we utilized this dataset to perform a systematic analysis, resulting in the identification and categorization of 13 key parameters across three macro-parameters: Research Methods, Health Disorders, and Disease Prevention. Informed by a focused review of over 200 articles, this analysis illuminates specific applications and highlights challenges in XAI, showcasing its impact on enhancing diagnostic accuracy, treatment efficacy, and preventive strategies. We then developed the FIXAIH framework to transform these insights into actionable guidelines, enhancing the interpretability, explainability, and accountability of AI systems in healthcare. Designed to ensure that AI technologies are ethically sound and comprehensible to healthcare professionals, the FIXAIH framework bridges the gap between technical proficiency and clinical utility, promoting the practical application of AI for a more reliable and patient-centric approach. This dataset and its analysis are integral to our broader research and development strategy, focusing on multiperspective parameter discovery and the advancement of autonomous systems [2]. Our approach leverages big data, deep learning, and digital media to explore and analyze cross-sectional, multi-perspective insights, supporting improved decision-making and more effective governance frameworks. These perspectives span academic, public, industrial, and governmental domains. We have applied this approach across various fields and sectors, including AI governance [3], energy [4], education [5], healthcare [6–8], transportation [9,10], labor markets [11,12], tourism [13], service industries [14], and others. References 1. doi:10.2139/SSRN.5086713. 2. doi: 10.54377/95e5-08b3 3. doi:10.3389/FNINF.2024.1472653/BIBTEX. 4. doi:10.3389/FENRG.2023.1071291. 5. doi:10.3389/FRSC.2022.871171/BIBTEX. 6. doi:10.3390/SU14063313. 7. doi:10.3390/TOXICS11030287. 8. doi:10.3390/app10041398. 9. doi:10.3390/SU14095711. 10. doi:10.3390/s21092993. 11. doi:10.3390/JOURNALMEDIA4010010. 12. doi:10.1177/00368504231213788. 13. doi:10.3390/SU15054166. 14. doi:10.3390/SU152216003.