Skip to main content
. 2023 Oct 12;19(5):791–799. doi: 10.1007/s11548-023-03022-w

Table 2.

Precision, recall and F1-scores for clinical role prediction, comparing a heuristic-based method and a neural network-based Graphormer for track scoring

Role Heuristic-based Graphormer
Prec Rec F1 Prec Rec F1
Patient 0.99 0.98 0.99 0.99 0.92 0.96
Head surgeon 0.93 1.00 0.96 0.96 1.00 0.98
Assistant surgeon 0.74 0.74 0.74 0.87 0.96 0.91
Circulating nurse 0.65 0.60 0.62 0.91 0.86 0.88
Anesthetist 0.61 0.45 0.52 0.72 0.52 0.60
Macroaverage 0.78 0.75 0.77 0.89 0.85 0.87

Bold indicates the best result between heuristic based and graphormer based