Using natural language processing to identify emergency department patients with incidental lung nodules requiring follow-up

2025-01-17 · Source: Explosion · Developer tools and consulting for AI, Machine Learning and NLP - Explosion.ai · Field: Health & Wellbeing — Medical Devices & Health Technology, Clinical Care & Medical Practice · Depth: Intermediate, quick

Summary

An NLP pipeline was developed to identify emergency department patients with incidental lung nodules requiring follow-up from CT reports. Medical doctors annotated these reports using Prodigy software to train the system. The pipeline operates in a stepwise manner, initially excluding cases with prior or known malignancy to ensure focus on new findings. Subsequently, it accurately determines the presence of a lung nodule within the CT scan data and then categorizes any recommended follow-up actions, such as further imaging or specialist consultation. This NLP system was built utilizing a RoBERTa large language model, integrated within the spaCy platform, to automate the detection and classification process, aiming to improve the timely management of incidental findings in a high-volume clinical setting.

Key takeaway

For NLP Engineers developing clinical decision support tools, this approach demonstrates a robust method for automating critical follow-up identification. You should consider a stepwise pipeline design, utilizing large language models like RoBERTa on platforms such as spaCy, to process unstructured medical reports. This can significantly enhance the efficiency of detecting incidental findings and ensuring timely patient management in high-volume environments like emergency departments.

Key insights

An NLP pipeline using RoBERTa and spaCy can automate incidental lung nodule follow-up identification from ED CT reports.

Principles

• Stepwise NLP improves diagnostic accuracy.
• MD annotation is crucial for training.

Method

A stepwise NLP pipeline first excludes malignancy, then identifies nodules, and finally categorizes follow-up recommendations using a RoBERTa LLM on spaCy.

In practice

• Automate incidental finding detection.
• Streamline patient follow-up.

Topics

• Natural Language Processing
• Clinical Decision Support
• RoBERTa
• spaCy
• Lung Nodule Detection
• Emergency Department

Best for: AI Scientist, NLP Engineer, Research Scientist

Related on AIssential

• End-to-end pipeline for automated heart failure diagnosis with clinical notes using SNOMED-CT — An end-to-end pipeline automates heart failure diagnosis from German clinical notes using zero-shot learning and SNOMED-CT.
• Intro to NLP with spaCy (1): Detecting programming languages — spaCy's linguistic features enable robust rule-based NLP systems beyond simple string matching.
• Evidential learning driven Breast Tumor Segmentation with Stage-divided Vision-Language Interaction — TextBCS enhances breast tumor segmentation in MRI by combining text prompts with evidential learning to address low contrast and blurred boundaries.
• AeroSpectra Sentinel: An Auditable LLM Prompt-Chaining Decision-Support Workflow for Acute Asthma Risk Assessment from Respiratory Sounds and Clinical Signals — An auditable LLM prompt-chaining workflow can enhance acute asthma risk assessment by integrating respiratory sounds and clinical data.
• Automated Detection of Dosing Errors in Clinical Trial Narratives: A Multi-Modal Feature Engineering Approach with LightGBM — Combining sparse lexical features with dense semantic embeddings improves clinical dosing error detection.
• Trusting Right Predictions for Wrong Reasons: A LIME Based Analysis of Deep Learning Interpretability in Lung Cancer Diagnosis — Deep learning models can achieve high predictive accuracy in medical imaging for different reasons, requiring independent interpretability validation.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Explosion · Developer tools and consulting for AI, Machine Learning and NLP - Explosion.ai.