A controlled benchmark of 17 QA-relevant anomaly types in multi-center DCE breast MRI shows unsupervised AD/OOD methods detect medium-far and far anomalies reliably but struggle with near-OOD cases such as implants and mastectomies.
Overcoming data scarcity in biomedical imaging with a foundational multi-task model
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
years
2026 5representative citing papers
A five-class CMR disease classifier built from LLM-derived report labels and three fine-tuned vision foundation models reached ensemble AUCs of 0.84–0.97 on a single-center test set.
A sparse-label framework using a frozen DINOv3 encoder and DeepLabV3+ decoder segments and quantifies three crack classes in hundred-megapixel SEM images of NMC cathodes with only 79 initial manual tiles.
Medical foundation models match a ResNet-50 but are outperformed by radiomics (AUC 0.88) on external validation for renal lesion stratification in CT.
citing papers explorer
-
Unsupervised Anomaly Detection for Image Dataset Quality Assurance in Multi-Center Breast MRI
A controlled benchmark of 17 QA-relevant anomaly types in multi-center DCE breast MRI shows unsupervised AD/OOD methods detect medium-far and far anomalies reliably but struggle with near-OOD cases such as implants and mastectomies.
-
Development of an automated, reliable, and clinically meaningful artificial intelligence (AI) tool for diagnosing cardiac disease from conventional cardiovascular magnetic resonance (CMR) images
A five-class CMR disease classifier built from LLM-derived report labels and three fine-tuned vision foundation models reached ensemble AUCs of 0.84–0.97 on a single-center test set.
-
Data-efficient crack quantification in lithium-ion cathodes using foundation model transfer
A sparse-label framework using a frozen DINOv3 encoder and DeepLabV3+ decoder segments and quantifies three crack classes in hundred-megapixel SEM images of NMC cathodes with only 79 initial manual tiles.
-
Benchmarking Foundation Models for Renal Lesion Stratification in CT
Medical foundation models match a ResNet-50 but are outperformed by radiomics (AUC 0.88) on external validation for renal lesion stratification in CT.
- NeuroBridge: Bridging Multi-Task MRI Knowledge for Neurodegenerative Disease Diagnosis