OpenAlex · Aktualisierung stündlich · Letzte Aktualisierung: 15.03.2026, 08:53

Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.

272P A multicenter study on human-machine interaction: Can CNN-support boost clinical decision-making in rare skin tumors?

2025·0 Zitationen·ESMO Real World Data and Digital OncologyOpen Access
Volltext beim Verlag öffnen

0

Zitationen

11

Autoren

2025

Jahr

Abstract

evaluating FP and FN rates in 4,111 records enriched for 22 key terms to derive Recall and F1-score. Results:The mNSCLC DP included 268 terms (mean[range] of alternative expressions 8.4[1-41]): 125 (46.6%) characteristics, 115 (42.9%) treatments, and 28 (10.5%)outcomes.The average (SD) Precision for the 268 terms was 0.97 (0.04), characteristics 0.98 (0.04), treatments 0.97 (0.04), and outcomes 0.96 (0.06).For key terms, the average (SD) Precision, Recall, and F1-score were 0.94 (0.05), 0.91 (0.08), and 0.92 (0.05), respectively. Conclusions:The strong detection performance of EHRead (Precision 0.97, F1-score 0.92) highlights the potential of cNLP combined with validated DP as an efficient alternative to manual chart review or ICD/claims-based approaches.This methodology offers scalable, clinically nuanced RWE generation aligned with regulatory standards, reinforcing its value as a foundation for high-quality real-world research.

Ähnliche Arbeiten