TR-BI-RADS: a novel dataset for BI-RADS based mammography classification

Ülgü M. M., ZALLUHOĞLU C., Birinci S., Yarbay Y., SEZER E.

Neural Computing and Applications, vol.36, no.7, pp.3699-3709, 2024 (SCI-Expanded) identifier

  • Publication Type: Article / Article
  • Volume: 36 Issue: 7
  • Publication Date: 2024
  • Doi Number: 10.1007/s00521-023-09251-z
  • Journal Name: Neural Computing and Applications
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, PASCAL, Applied Science & Technology Source, Biotechnology Research Abstracts, Compendex, Computer & Applied Sciences, Index Islamicus, INSPEC, zbMATH
  • Page Numbers: pp.3699-3709
  • Keywords: BI-RADS, Deep learning, Mammography classification, Mammography dataset
  • Hacettepe University Affiliated: Yes

Abstract

Breast cancer is still a crucial public health problem worldwide, especially among women. Early diagnosis and treatment can be provided to patients with regular mammography. The BI-RADS system, which is a standard approach used when interpreting mammography results, is widely used worldwide. The number of datasets classified according to the BI-RADS system is mostly limited. Based on this shortcoming, in this study, we introduce a new benchmark dataset, "TR-BI-RADS", for mammogram classification based on BI-RADS standardization. A convolution neural network (CNN) is evaluated on this dataset. In addition to the newly defined (TR-BI-RADS) dataset, experiments are also carried out on the other dataset (INbreast Dataset), available in the literature and consists of BI-RADS categories. We believe that the TR-BI-RADS dataset will be beneficial for detecting breast cancer in the future studies.