Reporting practices and sourcing of food images in fMRI research: implications for reproducibility

The fMRI literature has accumulated many studies that investigated the neural correlates of viewing food images. However, there has been no literature survey to provide an overall understanding of which food image datasets were used by these studies. To address this gap, a meta-research was conducted to identify if the fMRI studies used food images from established datasets and exactly from which datasets. The online database of Web of Science Core Collection (WOSCC) was queried to identify relevant papers. After screening, the analysis consisted of data manually extracted from 490 papers published during 2001-2024. Approximately 53.7% of the 490 analyzed papers did not specify their source of food images. About 29.8% of papers referred to published datasets for their food images, whereas some papers utilized a combination of online stock photo and in-house library. The most frequently cited food image datasets were IAPS and Blechert et al. (2014). The popularity of IAPS seemed to be declining since the mid-2010s in favor of newer datasets such as Blechert et al. (2014), Charbonnier et al. (2016), and Blechert et al. (2019). Some datasets are no longer accessible or have changed their websites. The findings highlight substantial heterogeneity and underreporting of visual stimulus materials, which may contribute to variability in observed neural responses and limit cross-study comparability in fMRI research.