Table 7.

Performance of the multiturn question answering model for feature extraction.

Feature	Accuracy	Precision	Recall	F1 score
Tumor density	0.940	0.875	0.915	0.893
MLNLAa	0.965	0.927	0.927	0.927
TLAb	0.974	0.974	0.974	0.974
Lobulation	0.923	0.993	0.716	0.832
TSAc	0.972	0.972	0.972	0.972
MLNSAd	0.986	0.918	0.931	0.924
Pleural indentation	0.917	0.903	0.938	0.920
Tumor location	0.994	0.990	0.990	0.990
Spiculation	0.979	0.988	0.945	0.966
Vessel invasion	0.982	0.932	0.788	0.854
HLNLAe	0.965	1.000	0.811	0.896
HLNSAf	0.986	0.982	0.848	0.911

^aMLNLA: mediastinal lymph node long axis.

^bTLA: tumor long axis.

^cTSA: tumor short axis.

^dMLNSA: mediastinal lymph node short axis.

^eHLNLA: hilar lymph node long axis.

^fHLNSA: hilar lymph node short axis.