Extracted and coded characteristics of 55 studies on explainable AI for ECG-based arrhythmia detection
Description
This dataset contains the extracted and coded characteristics of the 55 studies included in a systematic review and bibliometric analysis of explainable artificial intelligence (XAI) methods applied to electrocardiogram (ECG) data for arrhythmia detection (2021-2025), identified through a PRISMA 2020-compliant search of the Web of Science database. For each study, the dataset records the citation key, title, publication year, journal, data source, train/test structure, external validation status, XAI method and validation approach, clinician validation/assessment status, and arrhythmia type addressed. A supplementary sheet documents the coding methodology and criteria used. This dataset supports the taxonomy and PROBAST-adapted methodological quality appraisal presented in the associated manuscript.