Which software packages did researchers use to meta-analyze fMRI data? A literature

Andy Wai Kan Yeung^*

• University of Hong Kong, Hong Kong, China

There are various methods available for conducting meta-analyses of fMRI data, with coordinate-based meta-analysis (CBMA) being a frequently used approach due to the limited availability of effect size and statistical maps. Since the literature has accumulated many fMRI meta-analyses, several reports have been published to reveal the prevalence of numerous meta-analytic software packages without investigating into the recency of their versions used. To address this gap, a literature survey was conducted to identify the software packages and version numbers used for fMRI meta-analyses published between 2019 and 2024. The online databases of Web of Science Core Collection (WOSCC) and Scopus were queried to identify relevant papers. After screening, the analysis consisted of data manually extracted from 820 papers. The most frequently used software was GingerALE (407 out of 820 papers, 49.6%), followed by SDM-PSI (27.4%) and Neurosynth (11.0%). Overall, 540 papers (65.9%) fully disclosed the names and version numbers of the software packages used in their analyses, whereas 19 papers (2.3%) reported neither the names nor the version numbers. For GingerALE, the most frequently used versions were 2.3.6 and 3.0.2, but it should be noted that versions prior to 2.3.6 have an issue of inflated false positive rates. For SDM-PSI, the most frequently used versions were 5. 141, 5.15, 6.21, and 6.22, but the metaanalytic method adopted for version 6 differs from those used in prior versions. To ensure transparency and reproducibility in research, researchers should clearly report the name and version number of software package used.