Perspective on AI-Driven High-Risk Pregnancy Prediction and Safeguards

Figshare hosts a perspective article by Qingqing Ji, updated in March 2026, examining AI-based perinatal risk prediction. The 17.9 KB DOCX file synthesizes evidence on obstetric AI, prenatal screening impacts, and principles for trustworthy digital health. It argues risk labeling can increase anxiety and outlines discrimination pathways from biased data.

Use Cases

• Analyze the discussion of biased data and labels to understand over-flagging risks for socially disadvantaged groups.
• Extract proposed safeguards like transparent model documentation and continuous fairness monitoring for ethical AI framework design.
• Study the outlined discrimination pathways, including defensive clinical escalation, for research on over-medicalization.
• Review the synthesis of evidence on AI-based obstetric risk prediction for literature reviews on digital health tools.

Strengths

• Article is recent, last updated on March 26, 2026.
• Released under a permissive CC BY 4.0 license for reuse.
• Synthesizes evidence from three distinct domains: AI prediction, prenatal screening, and digital health governance.

Limitations

• Dataset is a 17.9 KB text document, not a structured data table for quantitative analysis.
• Content is a non-systematic narrative synthesis, not primary research data or model outputs.
• No sample data or defined columns are provided for machine learning applications.

Provenance

Source

Author Qingqing Ji via figshare.

Collection Method

Targeted, non-systematic narrative synthesis of published evidence and principles.

Time Range

Synthesis covers evidence from the past five years prior to publication.

Freshness

Last updated 2026-03-26.

Geography

null