Tracing the Heart: An Evidence-Linked Pipeline for Heart-Failure Feature Engineering

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Soorya Ram Shimgekar, Michelle Hu, Dorisa Shehi, Daniel Kang, Roy Ka-Wei Lee, Koustuv Saha, Christian Poellabauer, Christopher Lee, Sajeev Singh, Piyum Zonooz, Navin Kumar, Zeeshan Ahmed, Priyadarshini Kachroo
• Venue: arXiv: cs.AI
• Published: 2026-08-06
• Collected: 2026-08-07
• URL: http://arxiv.org/abs/2608.06366v1
• PDF: https://arxiv.org/pdf/2608.06366v1
• DOI: UNKNOWN

Plain-Language Brief

Researchers developed an AI system called nMAS to automate the process of extracting structured and aggregated features from Electronic Health Records (EHR) for heart failure research. The system generates these features using a rubric-grounded pipeline, ensures their integrity, compliance with guidelines, and audits them by a restricted large language model (LLM). The system improved the accuracy of heart failure phenotyping and was found to be auditable. However, the evaluation was limited to a single institution's data, and external validation is needed.

What the Document Says

• The research focuses on electronic health record (EHR) feature engineering, which is a significant challenge in clinical research and AI, accounting for 39-45% of data scientists' workload.
• Heart failure, affecting 6.7 million U.S. adults, requires integrating fragmented EHR data with disease-specific, guideline-based clinical reasoning.
• Existing rule-based and large language model (LLM)-based approaches offer partial automation with limited maintainability and evidence traceability.
• The researchers developed nMAS, an evidence-linked, rubric-grounded pipeline for automated heart-failure feature engineering.
• nMAS was tested on 500 dummy patient records from nine EHR source tables.
• nMAS generated 132 structured and 70 rubric-scored aggregated features.
• The structural integrity, rubric compliance, and provenance of the features were verified.
• An auditing process by a restricted LLM was conducted on the features.
• Adding the aggregated features improved the held-out AUROC (Area Under the Receiver Operating Characteristic Curve) for HFrEF (Heart Failure with Reduced Ejection Fraction) and HFpEF (Heart Failure with Preserved Ejection Fraction) phenotyping.
• The features were assessed by an independent LLM-based rubric, scoring 81.5% of maximum points for evidence support and methodological soundness.
• The evaluation was limited to a single-institution cohort, and external validation is needed.

Key Concepts Mentioned

• Electronic health record (EHR)
• Heart failure
• Feature engineering
• AI
• nMAS
• Rubric
• Large Language Model (LLM)
• AUROC

Claims Made by the Document

• nMAS improves the accuracy of heart failure phenotyping.
• The features generated by nMAS are auditable.
• Evaluation was limited to a single-institution cohort and external validation is needed.

Evidence Used

• Improved held-out AUROC for HFrEF and HFpEF phenotyping after adding the aggregated features generated by nMAS.
• An independent LLM-based rubric assessment of evidence support and methodological soundness scored the features at 81.5% of maximum points.
• The evaluation was limited to a single-institution cohort, as stated in the text.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Electronic health record (EHR) feature engineering is a major bottleneck in clinical research and AI, accounting for 39-45% of data scientists' workload. This is especially pronounced in heart failure, which affects an estimated 6.7 million U.S. adults and requires integrating fragmented EHR data with disease-specific, guideline-based clinical reasoning. Existing rule-based and large language model (LLM)-based approaches offer only partial automation with limited maintainability and evidence traceability. We developed the Nimblemind Multi-Agent System (nMAS), an evidence-linked, rubric-grounded pipeline for automated heart-failure feature engineering, and evaluated it on 500 dummy patient records from nine EHR source tables. nMAS generated 132 structured and 70 rubric-scored aggregated features, verified for structural integrity, rubric compliance, and provenance, and audited by a restricted LLM. Adding the aggregated features improved held-out AUROC from 0.895 to 0.963 for HFrEF and 0.870 to 0.910 for HFpEF phenotyping, and an independent LLM-based rubric assessment of evidence support and methodological soundness scored the features at 81.5% of maximum points. These results demonstrate the feasibility of automated, auditable feature engineering for complex cardiovascular EHR data, though evaluation was limited to a single-institution cohort and external validation is needed.