BCG-FM, the first foundation model for ambient BCG, achieves 3.26-year MAE on biological age estimation and discriminates 15 health conditions using frozen embeddings from participant-level contrastive pretraining on the largest raw biosignal corpus reported.
Mixed citations
LoDoPaB-CT, a benchmark dataset for low-dose computed tomography reconstruction,
Mixed citation behavior. Most common role is dataset (43%).
21
Pith papers citing it
Dataset
43% of classified citations
citation-role summary
background 4
dataset 3
citation-polarity summary
representative citing papers
VitaminP uses paired H&E-mIF data to train a model that transfers molecular boundary information, enabling accurate whole-cell segmentation directly from routine H&E histology across 34 cancer types.
Neural network corrects residual errors in isotopologue energy extrapolations for CO2 (MAE reduction in >87% of levels vs Marvel) and transfers patterns to improve CO predictions in >93% of samples.
Bayesian modeling of aneurysm surface displacements relative to the adjacent vessel provides probabilistic growth detection with AUC 0.86-0.87 and higher expert agreement than volumetric criteria.
New theoretical results on estimators and intervals for predicting unseen outcomes in additional samples from discrete distributions, with extensions to grouped incidence data.
A new schema-validated biomedical knowledge graph with rich properties and provenance, plus an LLM-agent validation showing 70% of sampled edges have literature support.
Bayesian active learning with SSCHA predicts phase transitions in materials like CsPbI3 using only 50-256 first-principles calculations.
An LLM-guided framework simulates physiological trajectories to provide interpretable early warnings for sepsis, achieving AUC scores of 0.861-0.903 on MIMIC-IV and eICU data.
GazeVaLM provides 960 gaze recordings from 16 radiologists on 60 chest X-rays (half synthetic) plus LLM predictions for diagnostic accuracy and real-fake detection under matched conditions.
Transfer learning from QM9 with transformers and uncertainty quantification predicts Hansen solubility parameters, dielectric constants, and limited-data Gutmann numbers to enable green solvent screening.
A four-mechanism framework produces auditable, evidence-grounded trait records from LLMs applied to over 400,000 tropical plant, aquatic, and pet species.
Open-weight LLMs match or exceed commercial API performance on 9 of 34 political science classification tasks, with average F1 differences under 0.02 and clearest API advantages on complex multi-label tasks.
Cascade classification improves macro F1 over single-stage for some models by allowing sensitivity control but reveals a large generalization gap on external clinical data.
Deep GLR combines graph Laplacian regularization with three lightweight CNN modules in a proximal optimization framework to reach 30.70 dB PSNR on LoDoPaB-CT using 5.8x fewer parameters and 30x less data per dB gain than typical deep methods.
A machine-learning framework maps soil salinity in Satkhira, Bangladesh, from field samples and Landsat indices, revealing expanding moderate-to-high salinity zones over the past decade.
PHES-ODM v3 is an enhanced open relational data model that adds tables for public health actions, external repository linkages, analytical workflows, and mapping tools to improve interoperability across wastewater surveillance programs.
SatBLIP fine-tunes a satellite-adapted BLIP model on GPT-4o-generated captions to predict county-level SVI from satellite tiles and uses SHAP to highlight key features like roof condition and vegetation.
Mistral uses careful lexical simplification to raise readability while keeping BERTScore at 0.91 comparable to humans, whereas QWen improves readability but shows a disconnect with its 0.89 BERTScore in biomedical text simplification.
High-throughput xTB screening of 747 experimental TADF molecules identifies D-A-D architectures and 50-90° torsional angles as favorable for small ΔE_ST, plus 127 candidates meeting ΔE_ST < 0.1 eV and f > 0.1.
DOI, ORCID, and ROR have evolved from metadata tags into machine-actionable links connecting research entities, with ORCID adoption varying 41-89% across German organizations and ongoing metadata quality issues.
A critical review of methods for estimating onshore wind energy potentials at multiple levels, with an attempt to derive best practice recommendations.
citing papers explorer
-
BCG-FM: A Foundation Model for Ambient Cardiac Health Sensing
BCG-FM, the first foundation model for ambient BCG, achieves 3.26-year MAE on biological age estimation and discriminates 15 health conditions using frozen embeddings from participant-level contrastive pretraining on the largest raw biosignal corpus reported.
-
VitaminP: cross-modal learning enables whole-cell segmentation from routine histology
VitaminP uses paired H&E-mIF data to train a model that transfers molecular boundary information, enabling accurate whole-cell segmentation directly from routine H&E histology across 34 cancer types.
-
Machine learning isotope shifts in molecular energy levels
Neural network corrects residual errors in isotopologue energy extrapolations for CO2 (MAE reduction in >87% of levels vs Marvel) and transfers patterns to improve CO predictions in >93% of samples.
-
Bayesian Aneurysm Growth Detection via Surface Displacement Modeling
Bayesian modeling of aneurysm surface displacements relative to the adjacent vessel provides probabilistic growth detection with AUC 0.86-0.87 and higher expert agreement than volumetric criteria.
-
The Unseen Species Problem Revisited
New theoretical results on estimators and intervals for predicting unseen outcomes in additional samples from discrete distributions, with extensions to grouped incidence data.
-
Unifying biomedical knowledge in a modern multimodal graph
A new schema-validated biomedical knowledge graph with rich properties and provenance, plus an LLM-agent validation showing 70% of sampled edges have literature support.
-
Predicting challenging phase transitions with Bayesian active learning
Bayesian active learning with SSCHA predicts phase transitions in materials like CsPbI3 using only 50-256 first-principles calculations.
-
Clinically Interpretable Sepsis Early Warning via LLM-Guided Simulation of Temporal Physiological Dynamics
An LLM-guided framework simulates physiological trajectories to provide interpretable early warnings for sepsis, achieving AUC scores of 0.861-0.903 on MIMIC-IV and eICU data.
-
GazeVaLM: A Multi-Observer Eye-Tracking Benchmark for Evaluating Clinical Realism in AI-Generated X-Rays
GazeVaLM provides 960 gaze recordings from 16 radiologists on 60 chest X-rays (half synthetic) plus LLM predictions for diagnostic accuracy and real-fake detection under matched conditions.
-
A green solvent screening tool for emerging materials via uncertainty aware, transformer enhanced transfer learning
Transfer learning from QM9 with transformers and uncertainty quantification predicts Hansen solubility parameters, dielectric constants, and limited-data Gutmann numbers to enable green solvent screening.
-
A Registry-Bound LLM Pipeline for Evidence-Grounded Trait Extraction across Tropical Plants, Aquatic Species, and Exotic Pets
A four-mechanism framework produces auditable, evidence-grounded trait records from LLMs applied to over 400,000 tropical plant, aquatic, and pet species.
-
Open-Weight LLMs Are Often Competitive with Commercial APIs for Political Science Text Classification
Open-weight LLMs match or exceed commercial API performance on 9 of 34 political science classification tasks, with average F1 differences under 0.02 and clearest API advantages on complex multi-label tasks.
-
Cascade Classification of Dermoscopic Images of Skin Neoplasms with Controllable Sensitivity and External Clinical Validation
Cascade classification improves macro F1 over single-stage for some models by allowing sensitivity control but reveals a large generalization gap on external clinical data.
-
Parameter-Efficient CT Reconstruction via Deep Graph Laplacian Regularization
Deep GLR combines graph Laplacian regularization with three lightweight CNN modules in a proximal optimization framework to reach 30.70 dB PSNR on LoDoPaB-CT using 5.8x fewer parameters and 30x less data per dB gain than typical deep methods.
-
A Dynamic Learning Observatory Reveals the Rapid Salinization of Satkhira, Bangladesh
A machine-learning framework maps soil salinity in Satkhira, Bangladesh, from field samples and Landsat indices, revealing expanding moderate-to-high salinity zones over the past decade.
-
The Public Health and Environmental Surveillance Open Data Model (PHES-ODM) Version 3: An Open, Relational Data Model and Interoperability Framework for Wastewater Surveillance
PHES-ODM v3 is an enhanced open relational data model that adds tables for public health actions, external repository linkages, analytical workflows, and mapping tools to improve interoperability across wastewater surveillance programs.
-
SatBLIP: Context Understanding and Feature Identification from Satellite Imagery with Vision-Language Learning
SatBLIP fine-tunes a satellite-adapted BLIP model on GPT-4o-generated captions to predict county-level SVI from satellite tiles and uses SHAP to highlight key features like roof condition and vegetation.
-
Making Knowledge Accessible: Divergent Readability-Accuracy Strategies of Mistral and QWen in Biomedical Text Simplification
Mistral uses careful lexical simplification to raise readability while keeping BERTScore at 0.91 comparable to humans, whereas QWen improves readability but shows a disconnect with its 0.89 BERTScore in biomedical text simplification.
-
Data-Driven Design Rules for TADF Emitters from a High-Throughput Screening of 747 Molecules
High-throughput xTB screening of 747 experimental TADF molecules identifies D-A-D architectures and 50-90° torsional angles as favorable for small ΔE_ST, plus 127 candidates meeting ΔE_ST < 0.1 eV and f > 0.1.
-
Beyond Metadata: The Role of DOI, ORCID, and ROR in Shaping Transparent and Interoperable Library Systems
DOI, ORCID, and ROR have evolved from metadata tags into machine-actionable links connecting research entities, with ORCID adoption varying 41-89% across German organizations and ongoing metadata quality issues.
-
Reviewing methods and assumptions for high-resolution large-scale onshore wind energy potential assessments
A critical review of methods for estimating onshore wind energy potentials at multiple levels, with an attempt to derive best practice recommendations.