The field of medical imaging and analysis is rapidly advancing with the development of new foundation models, fine-tuning strategies, and applications of deep learning techniques. Recent research has focused on improving the accuracy and efficiency of medical image segmentation, classification, and analysis. Notably, the use of vision transformers and multimodal models has shown promising results in various medical imaging tasks. Additionally, the adaptation of generic foundation models for specific medical applications has demonstrated potential for improved performance and clinical utility. Overall, these advancements are expected to enhance the diagnosis, treatment, and monitoring of various diseases and conditions. Noteworthy papers include: MedDINOv3, which introduces a simple and effective framework for adapting DINOv3 to medical segmentation, achieving state-of-the-art performance across four segmentation benchmarks. Curia, a multi-modal foundation model trained on a large corpus of real-world radiology data, accurately identifies organs, detects conditions, and predicts outcomes in tumor staging, meeting or surpassing the performance of radiologists and recent foundation models.
Advancements in Medical Imaging and Analysis
Sources
A Modality-agnostic Multi-task Foundation Model for Human Brain Imaging
A Hybrid AI-based and Rule-based Approach to DICOM De-identification: A Solution for the MIDI-B Challenge
SurgLLM: A Versatile Large Multimodal Model with Spatial Focus and Temporal Awareness for Surgical Video Understanding
SegDINO: An Efficient Design for Medical and Natural Image Segmentation with DINO-V3
MedDINOv3: How to adapt vision foundation models for medical image segmentation?
Mix-modal Federated Learning for MRI Image Segmentation
The Transparent Earth: A Multimodal Foundation Model for the Earth's Subsurface
PRECISE-AS: Personalized Reinforcement Learning for Efficient Point-of-Care Echocardiography in Aortic Stenosis Diagnosis
Tabular foundation model for GEOAI benchmark problems BM/AirportSoilProperties/2/2025
Lightweight image segmentation for echocardiography
A Generative Foundation Model for Chest Radiography
A Foundation Model for Chest X-ray Interpretation with Grounded Reasoning via Online Reinforcement Learning
Chest X-ray Pneumothorax Segmentation Using EfficientNet-B4 Transfer Learning in a U-Net Architecture
Uncertain but Useful: Leveraging CNN Variability into Data Augmentation
Challenges in Deep Learning-Based Small Organ Segmentation: A Benchmarking Perspective for Medical Research with Limited Datasets
MedSeqFT: Sequential Fine-tuning Foundation Models for 3D Medical Image Segmentation
Does DINOv3 Set a New Medical Vision Standard?
BioLite U-Net: Edge-Deployable Semantic Segmentation for In Situ Bioprinting Monitoring
Curia: A Multi-Modal Foundation Model for Radiology
Leveraging Generic Foundation Models for Multimodal Surgical Data Analysis
RepViT-CXR: A Channel Replication Strategy for Vision Transformers in Chest X-ray Tuberculosis and Pneumonia Classification