Lung Tumor Motion Prediction by Neural Networks

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21220%2F12%3A00196637" target="_blank" >RIV/68407700:21220/12:00196637 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Lung Tumor Motion Prediction by Neural Networks

Popis výsledku v původním jazyce

In tracking radiotherapy, the respiratory motion of the lungs affects the precision of the radiation dose delivered into a moving lung tumor. Thus, it has been complicated to synchronize the radiation beam with the tumor position. For medical treatment,there is an objective to minimize inaccurate delivery of doses to healthy tissues surrounding the lung tumor during radiotherapy. In order to improve the efficiency of radiation tracking therapy, this thesis presents a study of neural-network predictiontechnique with retraining for real time prediction of lung tumor motion position. The study focuses not only on the classical perceptron model, but also on a class of higher-order neural network model, which has more attractive attributes regarding its optimization convergence and computational efficiency, which are discussed in detail in the thesis. The implemented static feed-forward neural architectures used the Levenberg-Marquardt algorithm, and the presented modification resulted in

Název v anglickém jazyce

Lung Tumor Motion Prediction by Neural Networks

Popis výsledku anglicky

In tracking radiotherapy, the respiratory motion of the lungs affects the precision of the radiation dose delivered into a moving lung tumor. Thus, it has been complicated to synchronize the radiation beam with the tumor position. For medical treatment,there is an objective to minimize inaccurate delivery of doses to healthy tissues surrounding the lung tumor during radiotherapy. In order to improve the efficiency of radiation tracking therapy, this thesis presents a study of neural-network predictiontechnique with retraining for real time prediction of lung tumor motion position. The study focuses not only on the classical perceptron model, but also on a class of higher-order neural network model, which has more attractive attributes regarding its optimization convergence and computational efficiency, which are discussed in detail in the thesis. The implemented static feed-forward neural architectures used the Levenberg-Marquardt algorithm, and the presented modification resulted in

Druh

O - Ostatní výsledky

CEP obor

JB - Senzory, čidla, měření a regulace

OECD FORD obor

—

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2012

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů