Fig. 7: Fine-grained evaluations on the coding results from the pilot study.
From: Optimising the paradigms of human AI collaborative clinical coding
a Distribution of the experimental setups for the nine participants in the pilot study. Box plots for b the recall rates and c the precision rates of the nine participants. d Averaged coding time under three coding modes. e Averaged correction rates for three error types under the HITL ACC mode and the manual mode. f Averaged ROC curve of CliniCoCo and coding distribution of the participants under HITL ACC mode and manual mode.
