Breast Cancer Wisconsin Prognostic Data: Cell Nuclei Features and Recurrence Outcomes

Follow-up data for breast cancer cases seen by Dr. Wolberg since 1984, featuring only invasive cases without distant metastases at diagnosis. The dataset includes 30 features computed from digitized images of fine needle aspirates, describing characteristics of cell nuclei. It was created by researchers at the University of Wisconsin, including Dr. William H. Wolberg, W. Nick Street, and Olvi L. Mangasarian.

Use Cases

• Predicting recurrence outcome (R/N) based on cell nuclei features like radius and texture.
• Modeling time to recurrence or disease-free time using the provided Time feature.
• Feature selection analysis on the 30 computed cell nuclei characteristics.
• Benchmarking classification algorithms like decision trees on real-world medical data.

Strengths

• Includes 30 specific features per record computed from medical images, such as radius, texture, and concavity.
• Data represents consecutive patients from a specific clinical source since 1984, providing a focused cohort.
• Features were selected using an exhaustive search method described in the provided references.

Limitations

• Row count is unknown, which may limit suitability assessment for large-scale modeling.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

UCI Machine Learning Repository, originally from University of Wisconsin researchers.

Collection Method

Features computed from digitized images of fine needle aspirates (FNA) of breast masses.

Time Range

Cases from patients seen since 1984.

Geography

Likely Madison, Wisconsin, USA, based on the institution.