A Synthetic-First Quality Assurance Pipeline for Pre-Deployment Clinical AI Validation

Testing clinical AI workflows presents a paradox: rigorous validation needs realistic data, but using real patient data for testing violates privacy regulations like HIPAA. Generic MLOps tools do not close this gap because they lack clinical context. This paper describes Qualisera AI, a pre-deployment quality assurance framework that validates healthcare AI workflows using 100% synthetic and public test data. The architecture uses three layers: synthetic data ingestion (Synthea, HL7 FHIR, NIH metadata), declarative clinical safety gates, and a mock inference API. The framework executes an automated 113-test suite covering positive, negative, and API latency scenarios, achieving a 100% pass rate in the recorded run. The system is designed to catch physiologically impossible inputs and ambiguous data before they reach a model. This work is an independent engineering prototype and does not constitute clinical validation or regulatory certification.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-17
DOI
https://doi.org/10.5281/zenodo.24312872
Primary Topic
Artificial Intelligence in Healthcare and Education
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

A Synthetic-First Quality Assurance Pipeline for Pre-Deployment Clinical AI Validation

Moumita Ghosh
1 links
Zenodo (CERN European Organization for Nuclear Research)
Artificial Intelligence in Healthcare and Education
preprint

A Synthetic-First Quality Assurance Pipeline for Pre-Deployment Clinical AI Validation

Moumita Ghosh
preprint en

Abstract

Testing clinical AI workflows presents a paradox: rigorous validation needs realistic data, but using real patient data for testing violates privacy regulations like HIPAA. Generic MLOps tools do not close this gap because they lack clinical context. This paper describes Qualisera AI, a pre-deployment quality assurance framework that validates healthcare AI workflows using 100% synthetic and public test data. The architecture uses three layers: synthetic data ingestion (Synthea, HL7 FHIR, NIH metadata), declarative clinical safety gates, and a mock inference API. The framework executes an automated 113-test suite covering positive, negative, and API latency scenarios, achieving a 100% pass rate in the recorded run. The system is designed to catch physiologically impossible inputs and ambiguous data before they reach a model. This work is an independent engineering prototype and does not constitute clinical validation or regulatory certification.

Zenodo (CERN European Organization for Nuclear Research)
Good health and well-being
Artificial Intelligence in Healthcare and Education
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon