IEICE Transactions on Electronics
Online ISSN : 1745-1353
Print ISSN : 0916-8524
Noise-Robust Distorted Born Iterative Method with Prior Estimate for Microwave Ablation Monitoring
Yuriko TAKAISHIShouhei KIDERA
著者情報
ジャーナル 認証あり 早期公開

論文ID: 2020ECS6014

この記事には本公開記事があります。
詳細
抄録

A noise-robust and accuracy-enhanced microwave imaging algorithm is presented for microwave ablation monitoring of cancer treatment. The ablation impact of dielectric change can be assessed by microwave inverse scattering analysis, where the dimension and dielectric drop of the ablation zone enable safe ablation monitoring. We focus on the distorted Born iterative method (DBIM), which is applicable to highly heterogeneous and contrasted dielectric profiles. As the reconstruction accuracy and convergence speed of DBIM depend largely on the initial estimate of the dielectric profile or noise level, this study exploits a prior estimate of the DBIM for the pre-ablation state to accelerate the convergence speed and introduces the matched-filter-based noise reduction scheme in the DBIM framework. The two-dimensional finite-difference time-domain numerical test with realistic breast phantoms shows that our method significantly enhances the reconstruction accuracy with a lower computational cost.

著者関連情報
© 2020 The Institute of Electronics, Information and Communication Engineers
feedback
Top