Table 5.

Top 10 important features for pN2 lymph node metastasis prediction.

Rank	LRa			L2-LRb			RFc			LGBMd			All
	Feature	Weight	Feature		Weight	Feature		Weight	Feature		Weight
1	pGGOe,f	–10.383	CEAg		3.530	CEA		0.229	CEA		46.0	CEA
2	CEA	6.010	CA125h		3.067	CA125		0.094	Age		23.3	Solid nodulef
3	CA125	4.728	pGGOf		–1.799	Solid nodulef		0.094	Solid nodulef		18.8	CA125
4	Solid nodulef	3.683	Solid nodulef		1.773	MLNSAf,i		0.073	TLAf,j		17.6	Age
5	TLAf	–2.701	Age		–1.315	MLNLAf,k		0.072	TSAf,l		15.1	MLNLAf
6	Age	–1.908	SCCAgm		0.944	TLAf		0.054	CA125		13.3	TLAf
7	SCCAg	1.763	MLNLAf		0.896	TSAf		0.048	Cyfra211n		12.9	pGGOf
8	mGGOf,o	1.759	Pleural indentationf		0.836	Cyfra211		0.038	NSEp		12.7	SCCAg
9	RMLf,q	–1.729	Cardiovascular disease		0.807	SCCAg		0.037	MLNLAf		11.6	Lobulationf
10	TSAf	1.601	Lobulationf		0.725	Lobulationf		0.036	SCCAg		9.0	TSAf

^aLR: logistic regression.

^bL2-LR: L2-logistic regression.

^cRF: random forest.

^dLGBM: LightGBM.

^epGGO: pure ground glass opacity.

^fFeatures recorded in computed tomography reports.

^gCEA: carcinoembryonic antigen.

^hCA125: carbohydrate antigen 12-5.

ⁱMLNSA: mediastinal lymph node short axis.

^jTLA: tumor long axis.

^kMLNLA: mediastinal lymph node long axis.

^lTSA: tumor short axis.

^mSCCAg: squamous cell carcinoma antigen.

ⁿCyfra211: cytokeratin 19-fragments.

^omGGO: mixed ground glass opacity.

^pNSE: neuron-specific enolase.

^qRML: right middle lobe.