electronic health records
PulseAugur coverage of electronic health records — every cluster mentioning electronic health records across labs, papers, and developer communities, ranked by signal.
6 day(s) with sentiment data
-
MILM model uses LLMs for multimodal irregular time series
Researchers have developed MILM, a Large Language Model designed to process multimodal irregular time series data. This model represents time-series data as XML triplets and employs a two-stage fine-tuning strategy. The…
-
New EHR model uses retrieval to pinpoint relevant patient history
Researchers have developed EHR-RAGp, a new retrieval-augmented foundation model designed to more effectively utilize historical patient data within Electronic Health Records (EHRs). This model employs a prototype-guided…
-
Clin-JEPA framework enhances EHR data prediction and risk assessment
Researchers have developed Clin-JEPA, a novel framework for joint-embedding predictive pretraining specifically designed for electronic health record (EHR) patient trajectories. This method addresses challenges in apply…
-
Whoop adds clinician video calls and AI coaching to its fitness tracker
Whoop is integrating in-app video consultations with licensed clinicians for its US users, starting this summer. The fitness tracker company is also adding Electronic Health Record (EHR) syncing to its platform. Additio…
-
New method enables generalizable neural scaling laws across domains
Researchers have developed a method to create generalizable neural scaling laws that can be applied across different domains. These laws predict the relationship between model performance and resources like data or comp…
-
AI model predicts chronic rhinosinusitis using nationwide EHR data
Researchers have developed a new method to predict chronic rhinosinusitis (CRS) using nationwide electronic health record (EHR) data. The approach leverages two years of pre-diagnostic history and a hybrid feature-selec…
-
Healthcare EHR crisis demands technical architecture diplomacy for system interoperability
The article argues that the technical architecture of healthcare systems is crucial for enabling interoperability and resolving the global EHR crisis. It posits that the ability of different electronic health record sys…
-
New methods detect conditional anomalies in patient management
Researchers have developed conditional anomaly detection methods to identify unusual patterns in patient management, specifically focusing on instance-based approaches that use distance metrics. These methods aim to fla…
-
Sunscreen's two-hour reapplication rule lacks scientific evidence, analysis finds
The common guideline to reapply sunscreen every two hours while in the sun is not based on strong scientific evidence. This recommendation appears to have originated from a 2007 proposed rule by the FDA, which cited stu…
-
Machine learning models improve patient mortality prediction using medical notes
Researchers have developed a new Deep Neural Network (DNN) model with a pooling mechanism to improve the prediction of patient mortality after hospital discharge. This model leverages unstructured medical notes, which o…
-
LLMs enhance medical concept representation with text-attributed knowledge graphs
Researchers have developed MedCo, a framework that uses large language models to enhance medical concept representation within knowledge graphs. This approach addresses limitations in existing medical ontologies by infe…
-
Academic medical center deploys LLM platform, saving $6M in first year
Researchers have developed ChatEHR, a system designed to integrate large language models (LLMs) into academic medical centers, addressing workflow friction by allowing LLMs to access patient timelines. The system suppor…
-
Federated GNNs boost GDM prediction with privacy-preserving semi-supervised learning
Researchers have developed a novel federated semi-supervised learning framework called FedTGNN-SS to predict Gestational Diabetes Mellitus (GDM) while preserving data privacy across hospitals. This approach addresses ch…
-
New LITT architecture aligns clinical event timing for precision medicine
Researchers have developed a new architecture called LITT (Individual-Level Time Transformation) to better analyze clinical time-series data. This model addresses limitations in current AI, such as transformers, which o…
-
SHIELD: A Diverse Clinical Note Dataset and Distilled Small Language Models for Enterprise-Scale De-identification
Researchers have introduced SHIELD, a new dataset comprising 1,394 clinical notes with over 10,000 identified Protected Health Information (PHI) spans. This dataset aims to address the limitations of older benchmarks by…
-
Claude AI accused of defending dictatorships and Trump's EPA plans
A user on Mastodon expressed frustration with Claude, an AI model, alleging it was trained to defend dictatorships. The user specifically mentioned an instance where Claude was asked about Donald Trump's alleged plans f…
-
Deep Kernel Learning stratifies glaucoma patient risk trajectories using EHR data
Researchers have developed a new deep kernel learning architecture to help stratify glaucoma patient risk using electronic health records. The model employs a transformer-based feature extractor with clinical-BERT embed…
-
AI agent for clinicians shows improved performance with continuous governance
Researchers have developed a comprehensive framework for the continuous governance of AI agents embedded within Electronic Health Records (EHRs). This system integrates rubric validation, real-time feedback, performance…
-
MAHA movement wins pesticide vote but loses ally for surgeon general
A movement advocating for healthier food and reduced pesticide use, known as "Make America Healthy Again" (MAHA), successfully influenced a House farm bill vote to remove language that would have limited lawsuits agains…
-
Databricks Genie enables timely intervention for hospital readmissions
Databricks has introduced Genie for Clinical Outcomes Intelligence, a tool designed to bridge the gap between predicting patient readmissions and enabling timely interventions. While readmission risk models are accurate…