The study evaluates how well different medical foundation models agree on image embeddings by measuring mutual k‑nearest‑neighbor overlap across 750,000 images from 14 datasets and 18 public models. Results show that label‑supervised models converge the least, sharing only 6–10% of nearest neighbors, whereas models trained with self‑distillation, masked image modeling, or contrastive pretraining share up to 80%. The authors suggest that developers can select pretraining objectives to control convergence and recommend testing unlabeled images for nearest‑neighbor consistency before replacing a model.
arXiv:2607. 20274v1 Announce Type: cross Abstract: Medical image encoders from different groups are increasingly treated as interchangeable, on the assumption that scale and clinical supervision concentrate their representations onto a shared structure.
arXiv:2609.37848v1 Announce Type: cross Abstract: Biomedical machine learning papers often compress model performance into one headline number. That number can look like a property of the model even...
arXiv:2606. 17710v1 Announce Type: cross Abstract: Medical vision-language models report strong chest radiograph accuracy, and this is increasingly read as evidence that they use the image.
Med-AR introduces two autoregressive vision‑language models, Med‑AR‑8B and Med‑AR‑2B, pretrained on structured radiology reports, abnormality‑focused text, and region annotations to address long‑tailed chest X‑ray classification. The models outperform existing contrastive, self‑supervised, and supervised encoders—including Med‑CLIP, CheXFound, EVA‑Base, ARK, and BioViL‑T—across PadChest, MIMIC‑CXR, and CheXpert, achieving higher mean AUROC and AUPRC for head, medium, and tail findings and lower excess area under the risk‑coverage curve. Med‑AR also demonstrates improved selective‑prediction performance, with Med‑AR‑8B raising tail‑label mean AUPRC on MIMIC‑CXR from 0.1033 to 0.1441 and Med‑AR‑2B delivering the strongest discrimination on PadChest.
The paper introduces a label‑free method called AURCC for selecting the best foundational model for medical image classification when the target domain lacks labels. AURCC uses a pseudo‑label discrepancy computed by the SUDO framework to score models without fine‑tuning. Experiments on chest X‑ray data across three inter‑hospital shifts show that AURCC closely matches the true model ranking, outperforming simple source‑accuracy baselines especially when source data are limited.
arXiv:2606. 02276v1 Announce Type: cross Abstract: Vision-language models (VLMs) trained on paired chest radiographs and radiology reports learn a shared embedding space that can preserve instance-level image-report correspondence.
The paper evaluates federated learning with Low‑Rank Adaptation (LoRA) for fine‑tuning the BiomedCLIP vision‑language model on chest X‑ray classification across four international cohorts. Federated LoRA improves shared‑class AUC from 0.687 to 0.802, outperforming isolated single‑cohort training and approaching a centralized reference. The study shows that SVD‑based product‑space aggregation (FlexLoRA) is crucial for performance, while FedProx offers no advantage over FedAvg in this setting.
arXiv:2509. 19671v3 Announce Type: replace Abstract: Public datasets of Chest X-Rays (CXRs) have long been a popular benchmark for developing machine learning (ML) computer vision models in healthcare.
arXiv:2608. 16198v1 Announce Type: cross Abstract: Dermatology models face distribution shifts in teledermatology settings, where submitted images differ from the training data in lighting, angle, distance, focus, and framing.
arXiv:2607. 09305v1 Announce Type: cross Abstract: Chest radiography (CXR) remains the most widely used thoracic imaging modality, yet expert interpretation is constrained by a severe shortage of radiologists in Thailand and across Southeast Asia.
FedDRAW introduces a new server‑side aggregation technique for federated learning called Federated Dual Reputation Annealing Weighting. It blends a data‑size prior with cosine similarity between client and global model parameters, using two annealing schedules to shift weight from size to similarity and to gradually relax selectivity. Experiments on 12 simulated client partitions of CheXpert and ChestMNIST datasets show FedDRAW outperforms seven federated baselines in AUC and geometric mean of sensitivity and specificity, with statistically significant results.
OmniMed‑FL is a multimodal federated learning framework that fuses chest radiographs and synthetic patient notes to classify five clinical conditions. The study benchmarks eight fusion strategies, three initializations, and four missing‑text imputation rules across 3–20 hospital clients under non‑IID Dirichlet partitioning, showing that federated approaches (FedAvg, FedProx, SCAFFOLD‑AdamW) outperform local‑only training. Multimodal fusion consistently improves performance, achieving macro‑F1 scores up to 0.956 on the synthetic corpus and 0.906 on the radiograph corpus.