Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Editorial for "Integrating nnU-Net Segmentation and Clinical-Radiomics for Multicenter Prediction of Soft Tissue Sarcoma Grade and Ki-67 Expression".

Journal of magnetic resonance imaging : JMRI·2026

Same author

Hybrid Curriculum Learning for Data-Efficient Lung Nodule Detection with YOLOv11.

Diagnostics (Basel, Switzerland)·2026

Same author

De-Escalated Adjuvant Radiation Therapy in Patients With HPV-Positive Oropharyngeal Cancer.

JAMA network open·2026

Same author

A new chapter for JACMP: vision, article types, and new initiatives.

Journal of applied clinical medical physics·2026

Same author

Four-year experience with an in-house treatment management platform to streamline departmental operations in radiation oncology.

Journal of applied clinical medical physics·2026

Same author

Establishing prospective performance monitoring for real-world implementation of deep learning-based auto-segmentation in prostate cancer radiotherapy.

Physics and imaging in radiation oncology·2026

Search research articles

Related Experiment Video

Updated: Sep 18, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Improving the Precision of Deep-Learning-Based Head and Neck Target Auto-Segmentation by Leveraging Radiology Reports

Libing Zhu¹, Jean-Claude M Rwigema¹, Xue Feng²

¹Department of Radiation Oncology, Mayo Clinic, Phoenix, AZ 85058, USA.

|June 26, 2025

Summary

This study developed a deep learning auto-segmentation model for head and neck cancers, improving tumor delineation accuracy by eliminating false positives using clinical reports.

Keywords:

GTV auto-segmentation clinical diagnosis report head and neck large language model

More Related Videos

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

Related Experiment Videos

Last Updated: Sep 18, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

Area of Science:

Medical Imaging
Artificial Intelligence in Oncology
Radiation Therapy Planning

Background:

Accurate delineation of primary tumors (GTVp) and metastatic lymph nodes (GTVn) is critical for head and neck (HN) cancer radiation treatment.
Current segmentation methods are challenging and time-consuming, necessitating improved automated approaches.

Purpose of the Study:

To develop and validate a deep-learning-based auto-segmentation (DLAS) model for GTVp and GTVn in HN cancers.
To implement a false-positive elimination strategy using clinical diagnosis reports and large language models.

Main Methods:

A DLAS model was trained on a large multi-institutional dataset (882 cases).
Clinical diagnosis reports were used with ChatGPT-4 and Llama-3 to identify and rule out false-positive segmentations.
Performance was assessed using Dice Similarity Coefficient (DSC), Hausdorff distance (HD95), and precision.

Main Results:

ChatGPT-4 demonstrated superior performance in extracting tumor locations from reports compared to Llama-3.
False positives were identified in 15/44 cases, and their elimination improved mean DSC for GTVp and GTVn to 0.75.
Average HD95 for GTVn significantly decreased from 18.81 mm to 5.2 mm post-correction.

Conclusions:

False-positive ruling out using diagnostic reports significantly enhances DLAS precision for HN cancer segmentation.
The refined DLAS model accurately identifies tumor locations and detects false-negative errors, improving treatment planning efficiency.