A Short Breast Imaging Reporting and Data System-Based Description for Classification of Breast Mass Grade
2024

Breast Mass Classification Using BI-RADS Standards

Sample size: 107 publication 10 minutes Evidence: high

Author Information

Author(s): Grande-Barreto Jonas, Lopez-Armas Gabriela C., Sanchez-Tiro Jose Antonio, Peregrina-Barreto Hayde

Primary Institution: Tecnologías de la Información, Universidad Politécnica de Puebla

Hypothesis

Quantifying breast mass features in mammograms could provide a suitable description that agrees with the BI-RADS standard and would be helpful for the radiologist in assessing benign/malignancy.

Conclusion

The study demonstrates that a description based on BI-RADS allows for effective identification of breast masses, achieving high accuracy in classification.

Supporting Evidence

  • The methodology achieved an accuracy of 0.90 in benign/malignancy classification.
  • The study utilized the INbreast dataset, which includes 107 mammograms.
  • Results showed a general accuracy and sensitivity of 0.88±0.07.
  • Automatic classification was linked to BI-RADS standards for better clinical relevance.

Takeaway

This study helps doctors identify breast lumps better by using special computer methods that follow a set of rules called BI-RADS.

Methodology

The study used a dataset of mammograms to test various descriptors for breast mass classification based on BI-RADS standards, employing neural networks and image processing techniques.

Potential Biases

Potential bias may arise from the reliance on a single dataset for training and validation.

Limitations

The study is limited to a single dataset, which may not represent the full variability of breast masses in clinical practice.

Participant Demographics

The dataset includes diverse lesions with corresponding BI-RADS grades assessed by experts.

Statistical Information

P-Value

0.90

Confidence Interval

0.88±0.07

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.3390/life14121634

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication