Back to Benchmarks

BC Survival Benchmark
pathology
survival
H&E

Multi-cohort benchmark by Gustafsson et al. (2026) evaluating 13 pathology foundation models on breast cancer survival prediction (RFS and PFS) across three independent Swedish clinical cohorts (N=5,434 patients). Models are assessed on all patients and on the ER+ & HER2− subgroup, using concordance index (C-index) as the primary metric.

13 models evaluated
4 tasks
Organs:
breast

Detailed Results

C-index with 95% bootstrap confidence intervals across 4 evaluation settings (RFS/PFS × all patients / ER+ & HER2− subgroup).

Model
Average
rank
Mean
C-index
RFS (All patients)
C-index (95% CI)
RFS (ER+ & HER2−)
C-index (95% CI)
PFS (All patients)
C-index (95% CI)
PFS (ER+ & HER2−)
C-index (95% CI)
1.500.680
0.678
(0.6560.700)
0.670
(0.6450.695)
0.702
(0.6660.737)
0.670
(0.6260.715)
3.250.675
0.676
(0.6530.698)
0.668
(0.6420.693)
0.692
(0.6550.728)
0.666
(0.6190.711)
4.500.670
0.664
(0.6420.686)
0.660
(0.6330.685)
0.686
(0.6510.722)
0.670
(0.6240.714)
4.750.671
0.660
(0.6370.683)
0.651
(0.6250.676)
0.695
(0.6600.729)
0.680
(0.6360.721)
5.000.667
0.667
(0.6440.689)
0.663
(0.6370.688)
0.683
(0.6470.719)
0.657
(0.6110.702)
5.250.669
0.675
(0.6530.698)
0.665
(0.6390.690)
0.681
(0.6450.718)
0.655
(0.6080.700)
6.000.667
0.656
(0.6330.679)
0.639
(0.6130.665)
0.695
(0.6600.729)
0.678
(0.6340.720)
6.750.663
0.663
(0.6410.685)
0.657
(0.6310.681)
0.675
(0.6390.710)
0.656
(0.6120.700)
7.500.659
0.663
(0.6400.686)
0.657
(0.6310.682)
0.673
(0.6360.710)
0.641
(0.5940.687)
10UNI
9.500.654
0.657
(0.6350.680)
0.648
(0.6220.674)
0.671
(0.6340.708)
0.638
(0.5920.683)
11.500.632
0.632
(0.6090.655)
0.627
(0.6010.652)
0.647
(0.6090.686)
0.622
(0.5730.670)
12.000.628
0.645
(0.6220.668)
0.636
(0.6090.662)
0.627
(0.5890.666)
0.604
(0.5550.653)
12.500.611
0.612
(0.5880.635)
0.600
(0.5720.627)
0.628
(0.5900.666)
0.606
(0.5570.655)