Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
April 10, 2026JCO Clinical Cancer Informatics

Toward Automating the Summarization of Cancer Pathology Reports Using Large Language Models to Improve Clinical Usability

View Full Paper
Ask AI
Bookmark
Share

Authors

YLYirong LiuJJJacob JohnSSSagnik Sarkar

Discussion

Loading...

Member takes

Overview

This evaluation compares LLM-generated summaries with physician summaries in oncology, suggesting improved clinical usability.

Key Points

  • This research aims to assess the effectiveness of large language models in summarizing complex cancer pathology reports.
  • Analyzed pathology reports from patients in a thoracic clinic from Jan 2019 to July 2023.
  • Extracted and anonymized original reports and physician summaries for comparison.
  • Employed six open-source large language models (LLMs) to generate summaries from original reports.
  • Conducted both objective and subjective evaluations against original reports as a benchmark.
  • LLM-generated summaries outperformed physician summaries in all objective evaluation metrics (P < .0001).
  • DeepSeek, Mistral, Llama 3.1, and Llama 3.2 scored higher for completeness in subjective evaluations (P values ranging from .017 to < .0001).
  • Correctness of LLM summaries was comparable to physician summaries (P = 1.000).
  • Additional evaluations confirmed the consistency of results for Llama 3.1.

Cite This Study

Liu et al. (2026) studied this question.

synapsesocial.com/papers/69d895be6c1944d70ce06e43https://doi.org/10.1200/cci-25-00284
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Synoptic Reporting by Summarizing Cancer Pathology Reports using Large Language Models2024 · 3 citations
  2. 2Configuring large language models to deliver patient-facing explanations of pathology reports2026 · 1 citations
  3. 3Generative Artificial Intelligence for Medical Summarization in Prostate Cancer: Comparative Evaluation by Physicians and Patient Advocates—A Pilot Study2026 · 1 citations
  4. 4Cancer Type, Stage and Prognosis Assessment from Pathology Reports using LLMs2025
  5. 5Performance of large language models for extracting clinical data from breast cancer pathology reports: a systematic review2026 · 3 citations