The field of large language models (LLMs) is rapidly advancing, with a focus on improving their performance and reliability in various healthcare applications. Recent developments highlight the potential of LLMs to support clinical decision-making, automate medical text generation, and enhance patient-centered care. Notably, researchers are exploring the use of LLMs to generate structured and clinically grounded text, such as discharge summaries and progress notes, which can reduce the documentation burden on healthcare providers. Additionally, LLMs are being investigated for their ability to predict depression and retinopathy of prematurity risk, as well as to detect forced labor in supply chains. Overall, the field is moving towards more transparent, explainable, and human-centered AI systems that can augment clinical care and improve patient outcomes. Noteworthy papers in this regard include LCDS, which proposes a logic-controlled discharge summary generation system, and MedReadCtrl, which introduces a readability-controlled instruction tuning framework for personalized medical text generation. Furthermore, papers like DocCHA and MedThink-Bench demonstrate the potential of LLMs to support interactive online diagnosis and evaluate medical reasoning capability.
Advances in Large Language Models for Healthcare Applications
Sources
LCDS: A Logic-Controlled Discharge Summary Generation System Supporting Source Attribution and Expert Review
On the Semantics of Large Language Models
Empowering Healthcare Practitioners with Language Models: Structuring Speech Transcripts in Two Real-World Clinical Applications
MLlm-DR: Towards Explainable Depression Recognition with MultiModal Large Language Models
Divergent Realities: A Comparative Analysis of Human Expert vs. Artificial Intelligence Based Generation and Evaluation of Treatment Plans in Dermatology
Affective-ROPTester: Capability and Bias Analysis of LLMs in Predicting Retinopathy of Prematurity
Development and Evaluation of HopeBot: an LLM-based chatbot for structured and interactive PHQ-9 depression screening
Could the Road to Grounded, Neuro-symbolic AI be Paved with Words-as-Classifiers?
Exploring Task Performance with Interpretable Models via Sparse Auto-Encoders
Integrating Perceptions: A Human-Centered Physical Safety Model for Human-Robot Interaction
CLI-RAG: A Retrieval-Augmented Framework for Clinically Structured and Context Aware Text Generation with LLMs
Pre-Columbian Settlements Shaped Palm Clusters in the Sierra Nevada de Santa Marta, Colombia
Generating Multi-Table Time Series EHR from Latent Space with Minimal Preprocessing
Neurosymbolic Feature Extraction for Identifying Forced Labor in Supply Chains
SynthTextEval: Synthetic Text Data Generation and Evaluation for High-Stakes Domains
Medical Red Teaming Protocol of Language Models: On the Importance of User Perspectives in Healthcare Settings
MedReadCtrl: Personalizing medical text generation with readability-controlled instruction learning
SynthEHR-Eviction: Enhancing Eviction SDoH Detection with LLM-Augmented Synthetic EHR Data
Toward Real-World Chinese Psychological Support Dialogues: CPsDD Dataset and a Co-Evolving Multi-Agent System
Enhancing Vaccine Safety Surveillance: Extracting Vaccine Mentions from Emergency Department Triage Notes Using Fine-Tuned Large Language Models
SecureSpeech: Prompt-based Speaker and Content Protection
DocCHA: Towards LLM-Augmented Interactive Online diagnosis System
Performance and Practical Considerations of Large and Small Language Models in Clinical Decision Support in Rheumatology
Automating Expert-Level Medical Reasoning Evaluation of Large Language Models