Advancements in Clinical Text Analysis and Disease Detection

The field of clinical text analysis and disease detection is rapidly evolving, with a growing focus on leveraging large language models (LLMs) and artificial intelligence (AI) to improve diagnostic accuracy and streamline clinical workflows. Recent studies have demonstrated the potential of LLMs to extract relevant information from clinical notes, identify patterns and anomalies, and predict patient outcomes. Notably, ensemble-based approaches that combine the strengths of multiple models have shown promising results in improving the accuracy of disease detection and diagnosis. Furthermore, the integration of human expertise with AI-driven insights is emerging as a key strategy for optimizing clinical decision-making and reducing the reliance on manual labeling and annotation. Noteworthy papers in this area include:

  • ELM, which introduced an ensemble-based approach leveraging both small and large language models to predict tumor groups from pathology reports, achieving state-of-the-art results and significantly enhancing operational efficiencies.
  • IHC-LLMiner, which presented an automated pipeline for extracting tumour immunohistochemical profiles from PubMed abstracts using large language models, achieving 91.5% accuracy and outperforming existing models.

Sources

ELM: Ensemble of Language Models for Predicting Tumor Group from Pathology Reports

Refining Time Series Anomaly Detectors using Large Language Models

Leveraging LLMs for Predicting Unknown Diagnoses from Clinical Notes

Extracting Patient History from Clinical Text: A Comparative Study of Clinical Large Language Models

Comparing representations of long clinical texts for the task of patient note-identification

Multi-Task Learning for Extracting Menstrual Characteristics from Clinical Notes

Integrating Large Language Models with Human Expertise for Disease Detection in Electronic Health Records

Artificial Intelligence-Assisted Prostate Cancer Diagnosis for Reduced Use of Immunohistochemistry

IHC-LLMiner: Automated extraction of tumour immunohistochemical profiles from PubMed abstracts using large language models

Built with on top of