Skip to main content
. 2026 Jul 20;16:1844810. doi: 10.3389/fonc.2026.1844810

Table 2.

The diagnostic performance of the five models on the training cohort, test cohort, and two external validation cohorts.

Cohorts Models AUC (95% CI) Sensitivity Specificity Accuracy PPV NPV
Training (n=585) PICG2 0.773(0.7359-0.8109) 0.770 (191/248) 0.638 (215/337) 0.694 (406/585) 0.610 (191/313) 0.790 (215/272)
Z-SSMNet 0.792(0.7554-0.8282) 0.798 (198/248) 0.659 (222/337) 0.718 (420/585) 0.633 (198/313) 0.816 (222/272)
CM 0.758(0.7193-0.7960) 0.827(205/248) 0.585(197/337) 0.687(402/585) 0.594(205/345) 0.821(197/240)
RTLM 0.894(0.8666-0.9221) 0.774(192/248) 0.911(307/337) 0.853(499/585) 0.865(192/222) 0.846(307/363)
CIFM 0.918(0.8952-0.9417) 0.855(212/248) 0.893(301/337) 0.877(513/585) 0.855(212/248) 0.893(301/337)
Test(n=310) PICG2 0.681(0.6207-0.7413) 0.656 (84/128) 0.626 (114/182) 0.639 (198/310) 0.553 (84/152) 0.722 (114/158)
Z-SSMNet 0.710(0.6517-0.7690) 0.781 (100/128) 0.577 (105/182) 0.661 (205/310) 0.565 (100/177) 0.789 (105/133)
CM 0.742(0.6866-0.7974) 0.734(94/128) 0.604(110/182) 0.658(204/310) 0.566(94/166) 0.764(110/144)
RTLM 0.848(0.8065-0.8886) 0.680(87/128) 0.731(133/182) 0.710(220/310) 0.640(87/136) 0.764(133/174)
CIFM 0.890(0.8547-0.9250) 0.727(93/128) 0.863(157/182) 0.806(250/310) 0.788(93/118) 0.818(157/192)
External validation Cohort 1(n=510) PICG2 0.611(0.5620-0.6595) 0.931 (244/262) 0.177 (44/248) 0.565 (288/510) 0.545 (244/448) 0.710 (44/62)
Z-SSMNet 0.621(0.5725-0.6694) 0.656 (172/262) 0.484 (120/248) 0.573 (292/510) 0.573 (172/300) 0.571 (120/210)
CM 0.726(0.6821-0.7693) 0.656(172/262) 0.702(174/248) 0.678(346/510) 0.699(172/246) 0.659(174/264)
RTLM 0.815(0.7802-0.8504) 0.687(180/262) 0.726(180/248) 0.706(360/510) 0.726(180/248) 0.687(180/262)
CIFM 0.828(0.7936-0.8627) 0.744(195/262) 0.758(188/248) 0.751(383/510) 0.765(195/255) 0.737(188/255)
External validation Cohort 2(n=94) PICG2 0.716(0.6043-0.8275) 0.227 (15/66) 0.929 (26/28) 0.436 (41/94) 0.882 (15/17) 0.338 (26/77)
Z-SSMNet 0.742(0.6413-0.8436) 0.561 (37/66) 0.786 (22/28) 0.628 (59/94) 0.860 (37/43) 0.431 (22/51)
CM 0.738(0.6312-0.8440) 0.894(59/66) 0.286(8/28) 0.713(67/94) 0.747(59/79) 0.533(8/15)
RTLM 0.827(0.7291-0.9257) 0.621(41/66) 0.821(23/28) 0.681(64/94) 0.891(41/46) 0.479(23/48)
CIFM 0.852(0.7760-0.9285) 0.818(54/66) 0.643(18/28) 0.766(72/94) 0.844(54/64) 0.600(18/30)