Function AUC from PDFs: Breast DWI Sample Size Error Analysis

24 published breast diffusion-weighted imaging studies were reviewed and compared to an in-house cohort of 171 patients with suspicious breast lesions. The dataset, authored by Jessica V. Eberle and last updated in June 2026, analyzes dominant error types in apparent diffusion coefficient-based lesion characterization. The median sample size at which finite N errors and precision errors were equal was determined to be n=932.

Use Cases

• Estimating required sample sizes for breast DWI studies based on the analysis of finite N versus precision errors.
• Comparing error profiles between a local patient cohort and published literature as described in the methods.
• Informing study design optimization for ADC-based lesion characterization using the derived median sample size threshold.

Strengths

• Analysis is based on a review of 24 published studies and an in-house cohort of 171 patients.
• The dataset provides a specific, calculated median sample size threshold (n=932) for error balance.
• The work received Institutional Review Board approval, suggesting adherence to ethical research standards.

Limitations

• The dataset is very small (462 bytes), indicating it likely contains summary statistics or code rather than raw patient data.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for certain analyses.

Provenance

Source

figshare, author Jessica V. Eberle.

Collection Method

Literature review and analysis of a university hospital patient cohort.

Freshness

Last updated 2026-06-04 17:32:06; freshness should be verified.