IN-VIVO MOTION TRACKING DEVICE

PROBLEM TO BE SOLVED: To improve accuracy and robustness of tracking a target portion moving in vivo.SOLUTION: An organ position estimation unit 50 finds an estimated position of an organ in an in-vivo image based on past movement of the organ. An organ contour estimation unit 52 searches for contou...

Full description

Saved in:
Bibliographic Details
Main Authors FUKUDA HIROYUKI, AZUMA TAKASHI, HOSOI IZUMU, TSUKIHARA HIROYUKI, KAYASUGA ATSUSHI, NUMATA KAZUJI, TOMITA KYOHEI, YOSHINAKA KIYOSHI, KOIZUMI NORIHIRO, SUGITA NAOHIKO, MATSUMOTO YOICHIRO, NISHIYAMA YU, HONMA YUKIO, MIYAZAKI HIDEYO, MITSUISHI MAMORU
Format Patent
LanguageEnglish
Japanese
Published 20.09.2018
Subjects
CALCULATING
COMPUTING
COUNTING
DIAGNOSIS
ELECTROTHERAPY
HUMAN NECESSITIES
HYGIENE
IDENTIFICATION
IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
MAGNETOTHERAPY
MEDICAL OR VETERINARY SCIENCE
PHYSICS
RADIATION THERAPY
SURGERY
ULTRASOUND THERAPY
Online AccessGet full text

Cover

More Information
Summary:PROBLEM TO BE SOLVED: To improve accuracy and robustness of tracking a target portion moving in vivo.SOLUTION: An organ position estimation unit 50 finds an estimated position of an organ in an in-vivo image based on past movement of the organ. An organ contour estimation unit 52 searches for contour points in response to each of multiple control points representing a contour shape of the organ in a region corresponding to the estimated position, and finds an estimated contour of the organ based on the contour points. An affected part tracking unit 54 finds the position of the target portion from the estimated contour based on sample data acquired previously concerning the positional relationship between the contour of the organ and a target portion in conjunction with the organ. The organ contour estimation unit 52 finds the estimated contour by repeating the processing to find a factor of the deformation mode of the linear combination based on only the point corresponding to the clear contour point among control points and the processing to correct the position of the control point given in the processing by the SNAKE method based on the energy minimization principle concerning the contour shape, in a shape model of representing the contour shape by the sum of a reference shape and the linear combination of multiple deformation modes.SELECTED DRAWING: Figure 1 【課題】生体内を運動する注目部分の追跡の精度及びロバスト性の向上を図る。【解決手段】臓器位置推定部５０は臓器の過去の動きに基づき生体像における臓器の推定位置を求める。臓器輪郭推定部５２は推定位置に対応した領域にて、臓器の輪郭形状を表す複数の制御点それぞれに対応して輪郭点を探索し、輪郭点に基づいて臓器の推定輪郭を求める。患部追跡部５４は、臓器の輪郭と当該臓器に連動する注目部分との位置関係についての予め取得された標本データに基づき、推定輪郭から注目部分の位置を求める。臓器輪郭推定部５２は輪郭形状を基準形状と複数の変形モードの線形結合との和で表す形状モデルにて、制御点のうち明瞭な輪郭点に対応するもののみに基づき線形結合の変形モードの係数を求める処理と、当該処理で与えられる制御点の位置を輪郭形状に関するエネルギー最小化原理に基づくＳＮＡＫＥ法より修正する処理とを繰り返して推定輪郭を求める。【選択図】図１
Bibliography:Application Number: JP20170040348