Automated Extraction and Longitudinal Analysis of Ground Glass Opacity Features in Lung Cancer Patients Powered by Deep Learning-based Natural Language Processing (Preprint)

Abstract

Ground-glass opacities (GGOs) appearing in computed tomography (CT) scans may indicate potential lung malignancy. Proper management of GGOs based on their features can prevent lung cancer (LCA) development. Electronic health records (EHRs) are rich sources of information on GGO nodules and their granular features, but most of the valuable information is embedded in unstructured clinical notes

To develop, test, and validate a deep learning-based natural language processing (NLP) tool that automatically extracts GGO features to inform the longitudinal trajectory of GGO status from large-scale radiology notes.

We developed a bidirectional-long-short-term memory with a conditional-random-field-based deep-learning NLP pipeline to extract GGO and granular features of GGO retrospectively from radiology notes of 13,216 lung cancer patients. We evaluated the pipeline with quality assessments and cohort characterization was analyzed on the distribution of nodule features longitudinally to assess changes in size and solidity over time.

Our NLP pipeline, built upon the GGO ontology we developed, achieved 95-100% precision, 89-100% recall, and 92-100% F1 scores on different GGO features. We deployed this GGO NLP model to extract and structure comprehensive characteristics of GGOs from 29,496 radiology notes of 4,521 lung cancer patients. Longitudinal analysis revealed that size increased in 17.5% of patients, decreased in 15.1%, and remained unchanged in 67.4% in their last note compared to the first note. Among 1,127 patients who had longitudinal radiology notes of GGO status, 815 patients (72.3%) were reported to have stable status and 259 patients (23%) had increased/progressed status in the subsequent notes.

Our deep learning-based NLP pipeline can automatically extract granular GGO features at scale from EHRs when such information is documented in radiology notes and inform the natural history of GGO, which opens the way for a new paradigm in lung cancer prevention and early detection.