# Title Page

## Title

Diagnostic accuracy of three machine-learning classifiers for distinguishing malignant from benign breast masses on fine-needle aspirate cytomorphometry

## Running title

ML classifier diagnostic accuracy on breast FNA

## Authors

Demonstration Author¹

¹ medsci-skills v3.7.0 pipeline demonstration (DEMO 1, clean-room regeneration)

## Corresponding author

Demonstration Author (placeholder)

## Article type

Original article — diagnostic accuracy study (STARD 2015)

## Word count

- Abstract: structured (Background / Objective / Methods / Results / Conclusion)
- Main text (Introduction through Discussion): approximately 1,150 words

## Keywords

breast cancer; fine-needle aspiration; diagnostic accuracy; machine learning; logistic regression; STARD

## Funding

None (methods demonstration; no funding).

## Conflicts of interest

None declared.

## Reporting guideline

STARD 2015 (diagnostic accuracy).

## Artificial-intelligence-use disclosure

During preparation of this demonstration manuscript, the authors used Claude Opus 4.8 (Anthropic), accessed through the Claude Code command-line interface (API channel), during 2026-06 to run the medsci-skills v3.7.0 reproducible-reporting pipeline (figure generation, drafting, reporting-checklist scoring, and self-review). All quantitative claims were verified against committed analysis tables, and the authors take full responsibility for the content. Generative AI was not used to create, modify, or fabricate any data, statistical result, or figure; every number traces to a committed analysis artifact.

> Note: this disclosure paragraph is placed on the title page, not in the manuscript body, per the classical-style convention (in-body AI-disclosure paragraphs are flagged). It carries version (Opus 4.8), access channel (Claude Code CLI / API), date range (2026-06), and responsible party (the authors).