Objective: Automatic planning is a commonly used alternative to manual planning. This study evaluated the clinical performance of automatic plans available in commercial treatment planning systems for nasopharyngeal carcinoma (NPC) treatment by comparing automatic planning with manual planning.
Methods: A total of 14 patients with nasopharyngeal carcinoma were enrolled in the study. For each patient, three different sets of clinical goals were used to generate three hybrid automatic plans based on 3D dose distribution prediction and three automatic plans based on script, respectively, which were compared with the manual plans used in clinic.
Results: The dose coverage performance of the automatic planning based on 3D dose distribution prediction on the planning target volume (PTV) was comparable to that of the manual planning. Automatic planning based on 3D dose prediction achieved the level of manual planning in most organs at risk. However, automatic planning based on scripts did not perform well in the prediction of some organs at risk, especially the parotid gland.
Conclusion: The hybrid automatic plan based on 3D dose distribution prediction can reach the level of manual planning and have good robustness with the change of clinical objective.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.12455/j.issn.1671-7104.240159 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Deep reinforcement learning trajectory planning for robotic manipulator based on simulation-efficient training.
Sci Rep
March 2025
School of Information Science and Engineering, Northeastern University, Shenyang, 110819, China.
The paper proposes a new M2ACD(Multi-Actor-Critic Deep Deterministic Policy Gradient) algorithm to apply trajectory planning of the robotic manipulator in complex environments. First, the paper presents a general inverse kinematics algorithm that transforms the inverse kinematics problem into a general Newton-MP iterative method. The M2ACD algorithm based on multiple actors and critics is structured.
View Article and Find Full Text PDFMachine-learning based quantification of lung shunt fraction from 99mTc-MAA SPECT/CT for selective internal radiation therapy of liver tumors using TriDFusion (3DF).
EJNMMI Phys
March 2025
Departments of Medical Physics, Memorial Sloan Kettering Cancer Center, 1250 First Avenue, New York, NY, 10065, USA.
Background: Prior to selective internal radiotherapy of liver tumors, a determination of the lung shunt fraction (LSF) is performed using 99mTc- macroaggregated albumin (99mTc-MAA) injected into the hepatic artery. Most commonly planar but sometimes SPECT/CT images are acquired upon which regions of interests are drawn manually to define the liver and the lung. The LSF is then calculated by taking the count ratios between these two organs.
View Article and Find Full Text PDFInvestigation of Automated Cochlear Length and Cochlear Implant Insertion Angle Predictions with a Surgical Planning Platform.
Otol Neurotol
February 2025
Department of Otorhinolaryngology, Hannover Medical School, Hannover, Germany.
Introduction: Preoperative anatomical assessment is essential to optimize the outcome of individualized cochlear implantation. Algorithms based on cochlear diameters simplify this evaluation. The new version of a surgical planning platform is capable of performing this determination automatically.
View Article and Find Full Text PDFDosimetric evaluation of variation in Library of plans selection for cervical cancer radiotherapy.
Radiother Oncol
March 2025
Department of Radiation Oncology, Catharina Hospital, Eindhoven, the Netherlands.
Purpose: The aim was to determine whether differences in treatment plan selection methods from a Library-of-Plans (LoP) on cone-beam computed tomography (CBCT) leads to significant discrepancies in total accumulated dose.
Methods: Ten cervical cancer patients were retrospectively analyzed on plan selection and dose distribution. Dose accumulation was performed using two different plan selection methods.
Accurate fully automated assessment of left ventricle, left atrium, and left atrial appendage function from computed tomography using deep learning.
Eur Heart J Imaging Methods Pract
October 2024
Center for Medical Image Science and Visualization, Linköping University, SE-581 83 Linköping, Sweden.
Aims: Assessment of cardiac function is essential for diagnosis and treatment planning in cardiovascular disease. Volume of cardiac regions and the derived measures of stroke volume (SV) and ejection fraction (EF) are most accurately calculated from imaging. This study aims to develop a fully automatic deep learning approach for calculation of cardiac function from computed tomography (CT).
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!