EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Expert consensus on general requirements for high-quality dataset development in healthcare settings (2026 edition)

Zhenchang Wang, Jiahong Dong, Junbo Ge, Junmin Wei

Intelligent Medicine · 2026-11

DOI: 10.1016/j.imed.2026.11.003 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Intelligent Medicine, 2026-11. Verfasst von Zhenchang Wang und 3 weiteren Autor:innen. Die Arbeit nennt 144 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This expert consensus addresses key scenarios including healthcare delivery, clinical research, hospital management, public health, and the development, evaluation, and deployment of medical artificial intelligence (AI). It sets out principles, processes, and governance requirements systematically for the development and evaluation of high-quality datasets in healthcare settings. The consensus emphasizes that high-quality datasets are not merely resources for model training but also foundational infrastructure linking clinical value claims, graded evidence aligned with those claims, real-world validation, and continuous governance. The Consensus Development Working Group formulated recommendations around 10 consensus questions covering definitions and principles, the overall framework, needs assessment and scenario matching, data collection, preprocessing, annotation, transparency, validation and evaluation, compliance and security, and sustainable maintenance. The consensus recommends expanding the evaluation of healthcare dataset quality beyond static data quality to include task fitness, representativeness and fairness, external validation, workflow usability, evidence of clinical value, post-deployment drift monitoring, and the sustainable operation of data assets. For AI-related applications, retrospective model performance or dataset scale should not be used as substitutes for evidence of actual clinical benefit. This consensus is intended to provide a practical foundational framework for the development, trusted data circulation, and regulated application of high-quality healthcare datasets and medical AI in China.

Zitieren

APA

Wang, Z., Dong, J., Ge, J., & Wei, J. (2026). Expert consensus on general requirements for high-quality dataset development in healthcare settings (2026 edition). Intelligent Medicine. https://doi.org/10.1016/j.imed.2026.11.003

Vancouver

Wang Z, Dong J, Ge J, Wei J. Expert consensus on general requirements for high-quality dataset development in healthcare settings (2026 edition). Intelligent Medicine. 2026. doi:10.1016/j.imed.2026.11.003

BibTeX
@article{Wang2026,
  author = {Wang, Zhenchang and Dong, Jiahong and Ge, Junbo and Wei, Junmin},
  title = {Expert consensus on general requirements for high-quality dataset development in healthcare settings (2026 edition)},
  journal = {Intelligent Medicine},
  year = {2026},
  doi = {10.1016/j.imed.2026.11.003}
}
Kostenlos starten →