English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Radiologists produce unstructured data that could be valuable for clinicalcare when consumed by information systems. However, variability in style limitsusage. Study compares performance of system using domain-adapted language model(RadLing) and general-purpose large language model (GPT-4) in extracting commondata elements (CDE) from thoracic radiology reports. Three radiologistsannotated a retrospective dataset of 1300 thoracic reports (900 training, 400test) and mapped to 21 pre-selected relevant CDEs. RadLing was used to generateembeddings for sentences and identify CDEs using cosine-similarity, which weremapped to values using light-weight mapper. GPT-4 system used OpenAI'sgeneral-purpose embeddings to identify relevant CDEs and used GPT-4 to map tovalues. The output CDE:value pairs were compared to the reference standard; anidentical match was considered true positive. Precision (positive predictivevalue) was 96% (2700/2824) for RadLing and 99% (2034/2047) for GPT-4. Recall(sensitivity) was 94% (2700/2876) for RadLing and 70% (2034/2887) for GPT-4;the difference was statistically significant (P<.001). RadLing's domain-adaptedembeddings were more sensitive in CDE identification (95% vs 71%) and itslight-weight mapper had comparable precision in value assignment (95.4% vs95.0%). RadLing system exhibited higher performance than GPT-4 system inextracting CDEs from radiology reports. RadLing system's domain-adaptedembeddings outperform general-purpose embeddings from OpenAI in CDEidentification and its light-weight value mapper achieves comparable precisionto large GPT-4. RadLing system offers operational advantages including localdeployment and reduced runtime costs. Domain-adapted RadLing system surpassesGPT-4 system in extracting common data elements from radiology reports, whileproviding benefits of local deployment and lower costs.

General-Purpose vs. Domain-Adapted Large Language Models for Extraction of Data from Thoracic Radiology Reports