Back

Cracks in the Foundation: How Data-Hungry and Sensitive to Domain Shift are Vision Foundation Models for Computational Pathology?

2026-01-06 urology Title + abstract only
View on medRxiv
Show abstract

BackgroundVision Foundation Models (VFM) have emerged as a promising approach for computational pathology, offering scalable feature representations that may reduce labelled-data requirements and improve robustness to variation in tissue preparation and digitisation. However, VFM decoder and dataset size requirements as well as the performance under real-world domain shifts remain unclear. MethodsWe evaluated six contemporary VFMs on a protocol-variant Prostate Cancer (PCa) dataset comprising 3...

Predicted journal destinations

1
Scientific Reports
701 training papers
Top 4% 18.3%
2
npj Digital Medicine
85 training papers
Top 1% 9.3%
3
PLOS ONE
1737 training papers
Top 65% 8.1%
4
Nature Communications
483 training papers
Top 15% 6.6%
5
PLOS Digital Health
88 training papers
Top 6% 4.3%
6
eLife
262 training papers
Top 10% 4.3%
7
Computers in Biology and Medicine
39 training papers
Top 3% 2.0%
8
BMJ Open
553 training papers
Top 48% 2.0%
9
PLOS Computational Biology
141 training papers
Top 8% 1.6%
10
Cancers
57 training papers
Top 6% 1.4%
11
Frontiers in Oncology
34 training papers
Top 5% 1.2%
12
Communications Medicine
63 training papers
Top 4% 1.0%
13
Diagnostics
36 training papers
Top 4% 1.0%
14
Cureus
64 training papers
Top 16% 1.0%
15
Scientific Data
30 training papers
Top 3% 1.0%
16
Journal of Medical Internet Research
81 training papers
Top 16% 1.0%
17
JAMA Network Open
125 training papers
Top 23% 1.0%
18
JAMIA Open
35 training papers
Top 5% 1.0%
19
Science Translational Medicine
40 training papers
Top 4% 0.7%
20
Journal of Translational Medicine
21 training papers
Top 1% 0.7%
21
Human Brain Mapping
53 training papers
Top 10% 0.7%
22
eBioMedicine
82 training papers
Top 12% 0.7%
23
Journal of Biomedical Informatics
37 training papers
Top 6% 0.7%
24
Clinical Cancer Research
22 training papers
Top 4% 0.7%
25
Frontiers in Medicine
99 training papers
Top 23% 0.7%