Skip to main content
. Author manuscript; available in PMC: 2022 Nov 17.
Published in final edited form as: IEEE Trans Med Imaging. 2022 Oct 27;41(11):3003–3015. doi: 10.1109/TMI.2022.3176598

TABLE II.

Performance Metrics for the 3-Label (Normal vs. Luad vs. Lscc) Classification Task. Mean Performance Metrics Are Reported Along with the Corresponding Values of Standard Deviation in Parentheses

(a) Precision, Recall/Sensitivity, find Specificity (Percentage (%) values are reported).

Method Data Precision Recall/Sensitivity Specificity
Normal LUAD LSCC Normal LUAD LSCC Normal LUAD LSCC

TransMIL [27] CPTAC 89.7(1.8) 81.0(3.1) 87.1(1.7) 90.4(1.9) 81.2(2.0) 85.9(3.7) 94.4(1.1) 90.8(1.9) 93.7(1.2)
TCGA 76.6(5.1) 64.3(4.1) 80.4(1.7) 87.3(3.6) 75.8(5.3) 55.6(7.7) 90.6(2.9) 73.0(5.6) 92.4(1.9)

TransMIL [27] CPTAC 89.7(1.8) 81.0(3.1) 87.1(1.7) 90.4(1.9) 81.2(2.0) 85.9(3.7) 94.4(1.1) 90.8(1.9) 93.7(1.2)
TCGA 89.1(4.1) 69.9(3.9) 81.4(3.0) 87.8(2.7) 79.4(2.4) 71.3(4.0) 94.9(2.0) 82.5(2.9) 91.4(1.7)

GTP* (only GCN) CPTAC 82.9(6.5) 81.6(6.5) 86.5(4.7) 93.6(5.0) 74.4(6.7) 80.3(4.8) 89.3(5.3) 91.6(3.9) 93.6(2.7)
TCGA 72.5(9.0) 69.1(3.2) 82.0(9.8) 90.7(8.3) 57.7(9.7) 69.9(9.5) 82.4(9.9) 85.7(8.4) 90.7(6.1)

GTP CPTAC 93.2(3.0) 88.4(3.9) 87.8(3.0) 95.9(2.2) 83.9(4.5) 89.2(4.0) 96.2(1.7) 94.7(1.9) 93.8(1.7)
TCGA 89.2(2.8) 74.4(2.7) 84.4(0.7) 92.6(2.7) 79.8(1.9) 75.2(1.6) 94.7(1.6) 86.0(2.3) 92.7(0.4)

(b) Accuracy and AUC (Percentage (%) values are reported).

Method Data Accuracy AUC

TransMIL [27] CPTAC 85.9(0.7) 96.1(0.3)
TCGA 71.6(2.3) 88.0(0.7)

AttPool [15] CPTAC 81.9(2.1) 92.5(1.6)
TCGA 79.3(2.3) 91.3(1.1)

GTP* (only GCN) CPTAC 83.0(2.7) 95.2(1.2)
TCGA 72.4(4.9) 86.6(3.8)

GTP CPTAC 91.2(2.5) 97.7(0.9)
TCGA 82.3(1.0) 92.8(0.3)

(c) DeLong’s algorithm for comparing the AUC values between GTP and other methods. log10(0.05) = −1.301.

Method Data log10(p-value)

TransMIL [27] CPTAC −1.578(0.853)
TCGA −5.627(2.263)

AttPool [15] CPTAC −2.305(1.250)
TCGA −2.068(1.339)

GTP* (only GCN) CPTAC −1.759(1.129)
TCGA −5.373(3.146)