Diversity, inclusivity and traceability of mammography datasets used in development of Artificial Intelligence technologies: a systematic review

• Published in: Clinical imaging 2025-02, Vol.118, p.110369, Article 110369
• Main Authors: Laws, Elinor, Palmer, Joanne, Alderman, Joseph, Sharma, Ojasvi, Ngai, Victoria, Salisbury, Thomas, Hussain, Gulmeena, Ahmed, Sumiya, Sachdeva, Gagandeep, Vadera, Sonam, Mateen, Bilal, Matin, Rubeta, Kuku, Stephanie, Calvert, Melanie, Gath, Jacqui, Treanor, Darren, McCradden, Melissa, Mackintosh, Maxine, Gichoya, Judy, Trivedi, Hari, Denniston, Alastair K., Liu, Xiaoxuan
• Format: Article
• Language: English
• Subjects: Artificial intelligence; Mammography
• ISSN: 0899-7071; 1873-4499; 1873-4499
• DOI: 10.1016/j.clinimag.2024.110369

There are many radiological datasets for breast cancer, some which have supported the development of AI medical devices for breast cancer screening and image classification. This review aims to identify mammography datasets (including digitised screen film mammography, 2D digital mammography and digital breast tomosynthesis) used in the development of AI technologies and present their characteristics, including their transparency of documentation, content, populations included and accessibility. MEDLINE and Google Dataset searches identified studies describing AI technology development and referencing breast imaging datasets up to June 2024. The characteristics of each dataset are summarised. In particular, the accompanying documentation was reviewed with a focus on diversity and inclusion of populations represented within each dataset. 254 datasets were referenced in the literature search, 190 were privately held, 36 had barriers which prevented access, and 28 were accessible. Most datasets originated from Europe, East Asia and North America. There was poor reporting of individuals' attributes: 32 (12 %) datasets reported race or ethnicity; 76 (30 %) reported female/male categories with only one dataset explicitly defining whether these categories represented sex or gender attributes. Through this review, we demonstrate gaps in the data landscape for mammography, highlighting poor representation globally. To ensure datasets in breast imaging have maximum utility for researchers, their characteristics should be documented and limitations of datasets, such as their representativeness of populations and settings, should inform scientific efforts to translate data-driven insights into technologies and discoveries. •This review identified 254 mammography datasets used in the development of Artificial Intelligence technologies, but only 28 are freely accessible.•The source of mammography datasets span across 34 different countries but no African country except Egypt is represented.•Only 13% of datasets reported race or ethnicity; 2% of datasets reported socioeconomic status; 30% of datasets reported female/male categories.