光コヒーレンス断層画像を用いた生体組織歪み分布計測法 : 第2報,隣接相互相関乗法を用いたスペックルノイズ除去

概要

論文の詳細を見る
Acute coronary syndromes are caused by the rupture of unstable plaques on coronary arteries. The stability of plaque, which depends on biomechanical properties of fibrous cap, should be diagnosed crucially. Recently, Optical Coherence Straingraphy (OCS), e.g. an estimator of micro bio-strain distribution, has been being improved on the basis of Optical Coherence Tomography. This is composed of the recursive cross-correlation technique, which can provide the displacement vector distribution with high resolution. However, biomechanical characteristics have never estimated exhaustively, because OCT images severely include the speckle noise occurred by backscattering light from tissue. In the present study, proposed is Adjacent Cross-correlation Multiplication (ACM) as speckle noise reduction method. Multiplying adjacent correlation maps can eliminate anomalies from speckle noise, and then enhanse S/N in the determination of maximum correlation coefficient. Error propagation can be further privented by recursively introducing to OCS algorithm. In this report, the proposed method was experimentally and numerically applied to compressed tissue samples to carry out the accuracy verifications. Consequently, it was quantitatively confirmed that its accuracy of displacement vectors could be enhanced by introducing ACM to OCS, thus, the proposed method can obtain the accurate strain map with high resolution.
社団法人日本機械学会の論文
2007-11-25