What question did this study set out to answer?

This study aims to evaluate whether pretrained large language models can effectively extract structured information from unstructured medical notes in radiation oncology.

March 12, 2026Clinical and Translational Radiation OncologyOpen Access

Leveraging large language models to Populate structured clinical case report forms from unstructured medical notes in radiation oncology

View Full Paper

Ask AI

Bookmark

Authors

MNMarcel NachbarUniversity Children's Hospital Tübingen NYNianzi YiUniversity of Tübingen MBMarcel BüttnerGerman Cancer Research Center

Discussion

Member takes

Implication

Demonstrates LLMs extract structured data from unstructured notes in radiation oncology, suggesting enhanced efficiency in clinical documentation.

Key Points

This study aims to evaluate whether pretrained large language models can effectively extract structured information from unstructured medical notes in radiation oncology.
Examined medical notes from 100 prostate cancer patients treated with the MR-Linac.
Utilized the Llama-3.1-8b model with prompts designed using chain-of-thought logic.
Conducted independent testing on 10 patients after developing prompts using 90 patients.
Performed data extraction and categorization of medical notes by time points.
Compared model outputs against ground truth data from the MOMENTUM database.
Achieved matching accuracies of 83.6% and 83.8% on development and testing datasets respectively.
Processing time averaged 16 seconds per medical note.
Discrepancies noted in 8.1% of development and 8.6% of testing cases.
Approximately 7.5% of routinely collected test data did not match reviewed values.

Cite This Study

Nachbar et al. (2026) studied this question.

synapsesocial.com/papers/69b25abe96eeacc4fcec8c4e https://doi.org/10.1016/j.ctro.2026.101143

View Full Paper

Ask AI

Bookmark

Implication

Demonstrates LLMs extract structured data from unstructured notes in radiation oncology, suggesting enhanced efficiency in clinical documentation.

Leveraging large language models to Populate structured clinical case report forms from unstructured medical notes in radiation oncology

Discussion

Member takes

Implication

Key Points

Cite This Study

Leveraging large language models to Populate structured clinical case report forms from unstructured medical notes in radiation oncology

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider