Project Grant R01LM014604

Award Date 8/29/24
Completion Date 7/31/28
Dollars Obligated $376K
Federal Grant Program
93.879
Assistance Type
Project Grant
Place of Performance
Connecticut, USA
Similar Awards
This federal Project Grant award for $176,068 from the National Institute of Environmental Health Sciences (NIEHS) under the Medical Library Assistance (CFDA 93.879) program aims to develop and validate a clinically reliable and transparent large language model (LLM)-based question-answering system for clinical decision support. The key products and services to be delivered include: Developing novel clinically accurate LLMs (ClinicGPT) using a clinical-specific pre-training and fine-tuning...
This $729,821 Project Grant award from the National Institute of Environmental Health Sciences (NIEHS) under the Medical Library Assistance (CFDA 93.879) program supports the development of artificial intelligence (AI) and machine learning (ML) methods for real-time monitoring and updating of clinical decision support (CDS) systems. The goal is to reduce health disparities that may arise from the use of CDS tools. The key products of this work include: Fair ML models trained on retrospective...
This $348,227 federal Project Grant award, made by the National Science Foundation (NSF) under the Computer and Information Science and Engineering (CFDA 47.070) program, focuses on improving the scalability and effectiveness of using large language models (LLMs) in healthcare applications. The key objectives are to develop an evaluation framework to address issues like factual and faithfulness hallucinations in LLM outputs, and to introduce innovative reinforcement learning methods to align LLM...
The National Institute of Environmental Health Sciences (NIEHS) awarded a $727,580 Project Grant under the Medical Library Assistance (CFDA 93.879) federal grant program to the University of Texas at Austin (UT Austin) to develop novel natural language processing (NLP) technologies to improve the reliability of large language models (LLMs) for automatically simplifying medical documents. The project aims to create tools to detect errors in LLM-generated medical text simplifications and to...
This Project Grant award from the National Institute of Environmental Health Sciences (NIEHS) under the Medical Library Assistance (CFDA 93.879) program provides $351,000 to the University of Texas Health Science Center at Houston (UTHealth) to develop trustworthy and high-performance question answering capabilities for electronic health records (EHRs). The key objectives are to combine large language models with the UTHealth-developed QUEHRY system, which excels at providing trustworthy answers...
This federal Project Grant award from the National Institute of Environmental Health Sciences (NIEHS), under the Medical Library Assistance (CFDA 93.879) program, supports the development and evaluation of multi-modal clinical diagnostic reasoning models for automated diagnosis generation. The $498,042 award, starting on September 1, 2024, aims to address issues around diagnostic errors that affect millions of patients annually. The project will develop novel natural language processing (NLP)...
This $347,570 Project Grant award from the National Science Foundation's (NSF) Computer and Information Science and Engineering (CISE) program (CFDA 47.070) aims to develop a systematic framework called "Knowledge Graphs-Large Language Model Co-Learning (KG-LLM Co-Learning)" to address challenges in using large language models (LLMs) and knowledge graphs (KGs) for healthcare applications. The key products and services to be delivered include: Novel LLM-based methods for constructing...
This $249,000 federal Project Grant award from the National Institute of Environmental Health Sciences (NIEHS) under the Medical Library Assistance (CFDA 93.879) program supports the development of a novel Contrastive Feature Analysis (CFA) framework for reliable visualization and effective design of high-performance deep neural networks (DNNs) for medical image analysis. The key objectives of this 3-year project are to: 1) develop an efficient CFA visualization technique for high-dimensional...
This federal Project Grant award, funded by the National Institute of Environmental Health Sciences (NIEHS) under the Medical Library Assistance (CFDA 93.879) program, aims to develop and validate a novel informatics framework to extract computable population, intervention, comparison, and outcome (PICO) elements and summarize them from publications of interventional and observational studies. The total funding amount is $391,388.00, with an award date of September 18, 2024 and an ultimate...
This federal Project Grant award for $1,312,212 from the National Institute of Environmental Health Sciences (NIEHS) under the Medical Library Assistance (CFDA 93.879) program will support research to develop an advanced DNA language model. The primary goal is to use natural language processing techniques to better understand the role and functionality of DNA sequences, especially non-coding regions, which could lead to advancements in areas like genetic testing and personalized medicine. The...

This $375,964 Project Grant award from the National Institute of Environmental Health Sciences (NIEHS) under the Medical Library Assistance (CFDA 93.879) program aims to address the issues of factual inaccuracy and unfaithful reasoning in large language models (LLMs) applied to biomedical and healthcare domains. The key objectives are to: (1) establish a self-augmentation framework to enable LLMs to automatically select and use relevant biomedical digital resources; (2) develop an LLM curator that stimulates fact-checking and evidence attribution through multi-stage, multi-task instruction tuning; (3) introduce a step-level automated feedback-guided paradigm for LLMs to reflect and improve from their intermediate responses; and (4) evaluate these methods in downstream biomedical and healthcare use cases. The project will be carried out by Yale University, a leading research institution, over the award period from Aug 2024 to Jul 2028. The proposed work is expected to enhance the accuracy and trustworthiness of LLM-generated responses, advancing their application in biomedical discovery and healthcare.

Generated 7/1/25, 4:46 AM