Thyroid nodule classification in ultrasound imaging using deep transfer learning

Xu, Yan; Xu, Mingmin; Geng, Zhe; Liu, Jie; Meng, Bin

doi:10.1186/s12885-025-13917-3

BMC Cancer

Table 3 Diagnostic performance of machine learning model in predicting benign and malignant thyroid nodules

From: Thyroid nodule classification in ultrasound imaging using deep transfer learning

Model_name	Accuracy	AUC	95% CI	Sensitivity	Specificity	PPV	NPV	Precision	Recall	F1	Overall Score	Task
LR	0.653	0.711	0.6779—0.7440	0.78	0.538	0.604	0.729	0.604	0.78	0.681	0.709	label-train
LR	0.67	0.726	0.6608—0.7909	0.719	0.619	0.656	0.686	0.656	0.719	0.686	0.709	label-test
NaiveBayes	0.642	0.678	0.6440—0.7129	0.831	0.471	0.587	0.754	0.587	0.831	0.688	0.709	label-train
NaiveBayes	0.674	0.712	0.6451—0.7784	0.658	0.69	0.682	0.667	0.682	0.658	0.67	0.709	label-test
SVM	0.721	0.79	0.7606—0.8188	0.766	0.681	0.685	0.762	0.685	0.766	0.723	0.729	label-train
SVM	0.705	0.748	0.6842—0.8116	0.746	0.664	0.691	0.721	0.691	0.746	0.717	0.729	label-test
KNN	0.777	0.86	0.8374—0.8819	0.819	0.739	0.74	0.819	0.74	0.819	0.778	0.712	label-train
KNN	0.683	0.746	0.6848—0.8077	0.886	0.486	0.631	0.806	0.631	0.886	0.737	0.712	label-test
RandomForest	0.985	0.999	0.9979—0.9997	0.991	0.979	0.977	0.991	0.977	0.991	0.984	0.635	label-train
RandomForest	0.678	0.721	0.6555—0.7871	0.833	0.527	0.638	0.756	0.638	0.833	0.722	0.635	label-test
ExtraTrees	1	1	1.0000—1.0000	1	1	1	1	1	1	1	0.684	label-train
ExtraTrees	0.736	0.778	0.7186—0.8375	0.789	0.681	0.714	0.762	0.714	0.789	0.75	0.684	label-test
XGBoost	0.939	0.982	0.9750—0.9883	0.947	0.933	0.927	0.951	0.927	0.947	0.937	0.674	label-train
XGBoost	0.687	0.727	0.6611—0.7924	0.93	0.442	0.627	0.862	0.627	0.93	0.749	0.674	label-test
LightGBM	0.901	0.962	0.9509—0.9730	0.912	0.891	0.883	0.918	0.883	0.912	0.897	0.649	label-train
LightGBM	0.67	0.725	0.6594—0.7900	0.816	0.522	0.633	0.737	0.633	0.816	0.713	0.649	label-test
GradientBoosting	0.752	0.826	0.7993—0.8519	0.775	0.731	0.723	0.782	0.723	0.775	0.748	0.708	label-train
GradientBoosting	0.696	0.733	0.6672—0.7980	0.728	0.664	0.686	0.708	0.686	0.728	0.706	0.708	label-test
AdaBoost	0.686	0.744	0.7139—0.7742	0.633	0.733	0.682	0.688	0.682	0.633	0.657	0.710	label-train
AdaBoost	0.674	0.708	0.6447—0.7721	0.781	0.571	0.645	0.719	0.645	0.781	0.706	0.710	label-test
MLP	0.68	0.74	0.7082—0.7716	0.798	0.574	0.629	0.758	0.629	0.798	0.703	0.727	label-train
MLP	0.7	0.748	0.6840—0.8112	0.623	0.779	0.74	0.672	0.74	0.623	0.676	0.727	label-test

Back to article page

ISSN: 1471-2407

Contact us

Submission enquiries: bmccancer@biomedcentral.com
General enquiries: ORSupport@springernature.com