Internal benchmark evaluation and descriptive RNA quality audit of ChatBiobank (IMAGE)

International Society for Health Data Science

Internal benchmark evaluation and descriptive RNA quality audit of ChatBiobank

Caption

(A) Benchmark strata, internal evidence-support and routing tiers, and blinded expert-review workflow. (B) Response-tier distribution by benchmark stratum. (C) Expert-rated outcomes cross-tabulated by internal response tier after restoration of the concealed labels; numbers within bars are counts. All 77 outputs assigned to Tier B met the professional-acceptability criterion (75 directly acceptable and 2 broadly acceptable). (D) Single-center descriptive RNA quality audit of 153 fresh-tissue biospecimens processed under the implemented local workflow, of which 150 met the prespecified local qualification criteria. The RNA audit was not designed to estimate the causal effect of ChatBiobank. KB, knowledge base; LLM, large language model; QC, quality control.

Credit

Zhao Sun, Xuejiao Xia, Rui Deng, Chang Sun, Yunfei Wang, Zhuohang Li, Jiahui Li, Guanghui Li, Mingliang Yuan, Sanfei Peng, Xiangyang Yin, Zhiwei Zhao, Xinhua Dong, Qi Liu, Jing-Hua Yang, Zhen Yang, Yang Fu

Usage Restrictions

none

License

Original content

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.