Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
393P Privacy-preserving error analysis loop for ML-based extraction of oncology EHR data
0
Zitationen
12
Autoren
2025
Jahr
Abstract
Accurate extraction of clinical variables from electronic health records (EHR) using machine learning (ML) is vital for real-world oncology research, yet extraction errors can occur. Existing ML methods for EHRs often lack systematic error analysis, limiting reliability and regulatory acceptance. We developed a privacy-compliant workflow enabling clinical experts and data scientists to collaboratively pinpoint, categorise, and resolve ML extraction errors against a manually extracted gold standard, ensuring robust and trustworthy data.
Ähnliche Arbeiten
"Why Should I Trust You?"
2016 · 14.179 Zit.
A Comprehensive Survey on Graph Neural Networks
2020 · 8.561 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.071 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.429 Zit.
Analysis of Survival Data.
1985 · 4.379 Zit.