MIMIC-CXR, a de-identified publicly available database of chest radiographs with free-text reports
PulseAugur coverage of MIMIC-CXR, a de-identified publicly available database of chest radiographs with free-text reports — every cluster mentioning MIMIC-CXR, a de-identified publicly available database of chest radiographs with free-text reports across labs, papers, and developer communities, ranked by signal.
5 day(s) with sentiment data
-
New research reveals flaws in AI benchmarking for medical image and EHR data
A new paper published on arXiv highlights significant issues with the evaluation of class-imbalance strategies in multimodal medical classification tasks. The research found that benchmarks using record-level splits on …
-
New framework improves chest X-ray report generation with data compression
Researchers have developed ICGF-CXRMate, a novel framework designed to enhance the generation of clinical reports from chest X-ray images. This system addresses the challenge of integrating auxiliary patient data by com…
-
New MVMAE framework enhances medical imaging analysis using structural and textual supervision
Researchers have developed a new self-supervised learning framework called Multiview Masked Autoencoder (MVMAE) designed for medical imaging analysis. This framework leverages the inherent multi-view structure of radiol…
-
New AI framework enhances chest X-ray reports with longitudinal context
Researchers have developed CPVB-RG, a novel framework designed to improve chest X-ray report generation by incorporating longitudinal context from prior examinations. This system utilizes a Cross-Prior Visual Bridge (CP…
-
New framework separates decision rules from report generation for radiology AI
Researchers have developed a framework to address the inference problem of converting sampled evidence into final decisions for generating radiology reports. Their study, conducted on MIMIC-CXR and IU-Xray datasets usin…
-
Biomedical imaging AI progress stalled by data silos, not model limitations
The primary challenge in advancing biomedical imaging AI is not the development of more sophisticated models, but rather the accessibility and usability of the data itself. Imaging data is often siloed within hospital s…
-
New MambaXray-PRB framework enhances AI-driven X-ray report generation
Researchers have developed a new benchmark and framework called MambaXray-PRB for X-ray report generation (RRG) using large models. This framework addresses limitations in standardized benchmarks and domain adaptation f…
-
AI research advances structured radiology reporting with LLM-driven templates
Two new research papers propose advanced methods for improving structured radiology reporting using large language models. The first, ProtoSR, leverages an LLM to extract knowledge from free-text reports to create visua…
-
New AI framework MedCertAIn enhances clinical risk prediction with uncertainty awareness
Researchers have developed MedCertAIn, a new framework designed to improve the reliability and uncertainty estimation of AI models used in clinical settings. This framework specifically addresses the challenge of integr…
-
Radiology AI report evaluation metrics sensitive to reporting variations
A new research paper highlights how variations in radiologist reporting practices can significantly impact the evaluation of AI-based radiology report generation (RRG) models. The study introduces a method called ReRef …
-
New CNN-BiLSTM framework offers factually grounded biomedical text summarization
Researchers have developed a new hybrid framework combining 1D-CNN and BiLSTM models for extractive summarization of biomedical and clinical texts. This approach aims to prevent factual inaccuracies common in abstractiv…
-
ModaLens audit reveals medical VLMs rely more on text than images
Researchers have developed ModaLens, a new audit method to measure the sensitivity of vision-language models (VLMs) in medical contexts. Using the MedGemma-27B model on MIMIC-CXR data, the study found that when radiolog…
-
New MEDMKG knowledge graph integrates visual and textual medical data
Researchers have developed MEDMKG, a novel multimodal knowledge graph designed to integrate visual and textual medical information. This graph combines data from MIMIC-CXR and the Unified Medical Language System (UMLS),…
-
New AI Model NeoRed Enhances Neonatal Respiratory Disease Diagnosis
Researchers have developed NeoRed, a novel multimodal large language model specifically designed for diagnosing neonatal respiratory diseases. This model addresses limitations in existing systems, such as the domain gap…
-
LLMs evaluated for radiology report accuracy and longitudinal data extraction · 2 sources tracked
Researchers are exploring the use of large language models (LLMs) for improving radiology report quality and extracting longitudinal information. One study compared domain-specific BERT models with open-weight LLMs like…
-
New discrete diffusion model enhances radiology report generation
Researchers have developed DRRG, a novel discrete diffusion framework for radiology report generation that moves beyond traditional autoregressive models. This new approach allows for iterative refinement of reports, mi…
-
New Causal Model Enhances Chest X-Ray Interpretation and Interpretability
Researchers have developed XpertCausal, a novel causal concept bottleneck model designed to enhance the interpretability of chest X-ray interpretation. This model explicitly models the generative process from disease to…
-
New benchmark CRS-Bench evaluates medical image encoder reliability
Researchers have developed CRS-Bench, a new benchmark designed to evaluate the reliability of medical image encoders. Unlike previous methods that focused solely on discrimination, CRS-Bench assesses encoders across fou…
-
New LLM frameworks automate radiology report generation and template creation · 2 sources tracked
Researchers have developed new methods for improving radiology report generation using large language models (LLMs). One approach, ASTAR, automates the creation of standardized radiology reporting templates from clinica…
-
New MLLMs enhance radiology AI with 3D context and uncertainty reasoning · 3 sources tracked
Researchers are advancing multimodal large language models (MLLMs) for radiology, moving beyond simple image analysis to complex reasoning. One paper introduces a framework that addresses the representational mismatch b…