contrastive learning

MaxCAT-Net: Hybrid Transformer-Based Skin Lesion Classification and Lightweight Segmentation with Calibration and Cross-Dataset Validation

MaxCAT-Net improves skin lesion classification and segmentation with hybrid architecture and calibration, showing performance on benchmark datasets but with preliminary results.

Fine-Grained and Reliable Cell Classification from Pathology Images

CellBridge improves cell classification in pathology images by transferring molecular information, achieving higher accuracy on fine-grained labels.

UniMedSeg: Unified Multi-Modal Medical Image Segmentation via Uncertainty Guided Encoding and Frequency-Spatial Dual-Prompt Decoding

UniMedSeg improves multi-modal medical image segmentation by integrating uncertainty-guided encoding and dual-prompt decoding for better domain generalization and spatial risk estimation.

Cross-cohort Mamography]{Representation-dependent domain adaptation for cross-cohort generalization of mammography diagnostic models

This study evaluates cross-cohort generalization of mammography diagnostic models using different image representations and domain-adversarial learning, finding that MedSigLIP embeddings provide the best transferability.

Discovering Latent Scientific Concepts through Discrete Representation Learning

This paper introduces Kodiak, a discrete representation learning method for scientific imaging that improves representation quality and enables interpretable, reusable concepts across domains.

Deep learning of fossil pollen morphology reveals 25,000 years of ecological change in East African grasslands

This paper uses deep learning to analyze fossil grass pollen morphology and reconstruct 25,000 years of ecological change in East African grasslands, showing reduced diversity and shifts in C3:C4 ratios during the last ice age.

Foveation as self-supervision signal: Blur masks beat blank masks for learning robust visual representations

This paper introduces blur and retinal masks as self-supervision signals for vision transformers, showing improved robustness and generalization in visual representation learning.

RNA-guided contrastive learning enhances patient-level prediction from histology

This paper improves histology-based molecular receptor prediction by integrating RNA-seq data during training, showing that gene selection strategy is key to model performance.

Beyond Marginal Coverage: Class-Conditional Conformal Prediction in Multi-Class Psychiatric Neuroimaging

This paper introduces class-conditional conformal prediction to address uneven coverage in multi-class psychiatric neuroimaging, showing improved coverage consistency across diagnostic classes.

AC-Diff: Anatomy-Contrast Disentangled Diffusion for Multi-Vendor Liver MRI Harmonization

AC-Diff improves liver MRI harmonization by separating anatomy from contrast, achieving better segmentation results.

Advancing Wildlife Image Analysis: A Graph Attention Contrastive Learning Approach for Region-Specific Mammal Classification

This paper develops a region-specific AI model for mammal classification using graph attention contrastive learning, achieving high accuracy in recognizing local species and improving data cleaning efficiency.

Self-Supervised Discovery of Discrete Local States in Noisy Image Sequences

This paper introduces a self-supervised framework to discover discrete local states in noisy image sequences without clean targets or labels, achieving high precision in particle tracking and separating biological structures from artifacts.

AMADEUS: Annotation-free multi-animal direction estimation using self-supervised learning

AMADEUS tracks multiple animals without annotations or markings, achieving better performance in identity tracking and keypoint localization across various species.

SHOT-CCR: Biologically guided adversarial training for test-time adaptation in cellular morphology

SHOT-CCR improves classification of genetic perturbations in Cell Painting datasets by reducing batch effects through biologically guided adversarial training.

Scalable Computational Phenomics of Nuclear Morphology

NucXplore converts H&E-stained nuclei into 129 explicit features, outperforming classical and deep learning methods in histopathology cohorts and revealing aging and exposure-related cellular changes.

SVD-SN2N: Self-Inspired Noise2Noise Learning for Denoising Log-Compressed SVD-Filtered Ultrasound Imaging

SVD-SN2N improves denoising of SVD-filtered ultrasound images by converting multiplicative speckle to additive noise, enhancing SNR and vessel resolution without requiring clean references or raw data.

DINOCT: robust skin surface localization for robotic noncontact dynamic optical coherence elastography

DINOCT improves skin surface localization in noncontact OCE by combining self-supervised pretraining with DINO and iBOT objectives, achieving high accuracy and robustness to artifacts.

Decoupling Detection and Classification to Improve Morphological Phenotype Analysis of Sickle Red Blood Cells in Full-Scope Microscopy

This paper improves sickle cell disease morphology analysis by combining a YOLO detector with a DenseNet classifier for accurate detection and classification of red blood cell morphologies in full-scope microscopy images.

An explainable AI latent space of brain dynamics reveals a cerebello-prefrontal signature of schizophrenia symptoms

This paper develops an interpretable AI framework to map brain dynamics onto a latent space, identifying a cerebello-prefrontal signature associated with schizophrenia symptoms.

UMITIC: An unsupervised framework for the joint characterization of cellular phenotypes and spatial neighborhoods in multiplex and hyperplex immunofluorescence imaging data

UMITIC is an unsupervised framework that characterizes cellular phenotypes and spatial neighborhoods in multiplex imaging data, showing improved accuracy over existing methods.

Unsupervised machine-learning identifies latent pyrenoid states linked to mitotic remodeling defects and CO2-dependent growth

The paper uses unsupervised machine learning to identify pyrenoid morphology mutants in Chlamydomonas reinhardtii, linking defects to CO2-dependent growth and mitotic remodeling.

Trustworthy in silico labeling via semantic visual interpretability of image-to-image translation

The paper introduces Mask Interpreter, a method to improve interpretability of image-to-image translation models, validating biological relevance and identifying artifacts in in silico labeling.

From Generation to Discrimination: Vision Foundation Models for Synthetic SEM Image Detection

This paper develops a deep learning framework to detect synthetic SEM images generated by GANs, using CLIP ViT-L-14 fine-tuned on real and synthetic datasets.

MAE-UNETR++: Masked Autoencoder Pretraining for 3-D Lung Nodule Segmentation

This paper proposes MAE-UNETR++ for 3-D lung nodule segmentation, showing that MAE pretraining outperforms random initialization and Decathlon transfer learning in low-data scenarios.

PlumageParts: A fine-grained avian plumage segmentation dataset and benchmark for ecological image analysis

PlumageParts provides a dataset of 4,705 bird images annotated for nine plumage regions and benchmarks segmentation models, achieving 84.01% mean Intersection over Union with a frozen DINOv3 encoder.

DCGAN-Based Synthetic MRI Augmentation for Data-centric Brain Tumor Segmentation

This paper uses DCGAN to generate synthetic MRI images for brain tumor segmentation, improving Dice coefficient and IoU metrics compared to a baseline model.

Machine-learning MRI stratification of genetic frontotemporal dementia for clinical trial enrichment

This paper uses machine learning to stratify genetic frontotemporal dementia patients into subtypes and provides progression scores that improve trial enrichment efficiency.

Multicenter self-supervised computational pathology identifies prognostic histomorphological phenotypes in colorectal cancer

This paper develops a self-supervised computational pathology framework to identify prognostic histomorphological phenotypes in colorectal cancer using H&E images and tile embeddings.

SST-MAE: Learning Spectral-Spatio-Temporal Representations from Plant Hyperspectral Time Series to Discover Complex Genotype-Phenotype Relations

NeuroVLM: A generative vision-language framework for human neuroimaging

NeuroVLM generates and evaluates vision-language models for human neuroimaging using 30,826 neuroimage-text pairs to perform tasks like text-to-image generation and image-to-text interpretation.