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

Search research articles

Related Experiment Video

Updated: Jul 19, 2026

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules

Published on: October 13, 2023

Leaf-Specific Classification of Multi-Leaf Collimator Positioning Errors in Volumetric Modulated Arc Therapy Using a

Ju Yeol Shin^1,2, Chang Heon Choi^1,2, Jung-In Kim^1,2

¹Paprica Lab. Co., Ltd., Seoul 03123, Republic of Korea.

Journal of Clinical Medicine

|July 15, 2026

Summary

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

Dictionary-Augmented Large Language Model Postprocessing for Bilingual Code-Switched Medical Speech Recognition: Development and Evaluation Study.

Journal of medical Internet research·2026

Same author

Effects of Dielectric Interlayer on Polarization Switching and Rectifying Characteristics in Al<sub>0.8</sub>Sc<sub>0.2</sub>N/HfO<sub>2</sub> Ferroelectric Diodes.

Micromachines·2026

Same author

Machine learning approach to estimation of in vivo measured dose and treatment planning for total body irradiation.

Radiation oncology (London, England)·2026

Same author

Microdosimetric kinetic model for enhanced estimation of cell survival in aerobic human salivary gland cells under heavy ion irradiation.

Physics in medicine and biology·2026

Same author

Loss of ABCA3 disrupts lipid balance and leads to AMPK-dependent suppression of SREBP1 in glioblastoma stem cells.

Oncogene·2026

Same author

Plaque-adaptive flexible film dosimetry system for Ru-106 plaque brachytherapy commissioning and quality assurance.

Medical physics·2026

A new convolutional neural network (CNN) accurately identifies individual multi-leaf collimator (MLC) positioning errors in volumetric modulated arc therapy (VMAT). This AI framework precisely locates leaf errors, improving radiation dose accuracy and patient safety.

Area of Science:

Medical Physics
Radiotherapy Technology
Artificial Intelligence in Medicine

Background:

Multi-leaf collimator (MLC) positioning accuracy is crucial for dose delivery in volumetric modulated arc therapy (VMAT).
Current quality assurance (QA) methods using gamma analysis offer only plan-level pass/fail results, lacking leaf-specific error localization.
This limitation hinders precise identification and correction of individual MLC positional deviations.

Purpose of the Study:

To develop and validate a convolutional neural network (CNN) framework for classifying MLC leaf positioning errors.
To accurately determine the magnitude and direction of individual MLC leaf errors directly from fluence map data.
To enhance the precision of VMAT QA by enabling leaf-specific error identification.

Main Methods:

Keywords:

MLC error classification convolutional neural network deep learning fluence map multi-leaf collimator quality assurance volumetric modulated arc therapy

Related Experiment Videos

Last Updated: Jul 19, 2026

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules

Published on: October 13, 2023

A CNN was trained as a 121-class classifier using simulated fluence map data from three patient cohorts.
Systematic offsets (-5 mm to +5 mm) were applied to inner MLC leaf banks to generate error conditions.
The CNN model utilized two-channel inputs pairing reference and error-induced fluence map regions.
Performance was evaluated against tree-based baselines using cross-validation and tested on independent patient datasets.

Main Results:

The CNN achieved high accuracy: 97.00% on the internal test set and 96.54% across cross-validation folds.
Over 99.8% of predictions were within 1 mm of true offsets for both leaf banks, meeting AAPM TG-142 tolerances.
External validation demonstrated good accuracy (96.19% for prostate, 93.72% for H&N), indicating reproducibility and potential robustness.
The framework successfully identified both magnitude and direction of MLC positioning errors.

Conclusions:

The developed CNN framework effectively identifies individual MLC positioning errors in magnitude and direction from simulated fluence maps.
This AI-driven approach shows significant promise for improving VMAT QA precision.
Further research using physically measured fluence data is recommended for clinical implementation.