首页> 外文会议>Conference on photons plus ultrasound: imaging and sensing >S-Sequence Patterned Illumination for Fixed-Point Iterative Multiple Illumination Photoacoustic Tomography
【24h】

S-Sequence Patterned Illumination for Fixed-Point Iterative Multiple Illumination Photoacoustic Tomography

机译:定点迭代多次照明光声层析成像的S序列模式照明。

获取原文

摘要

Fixed-point iteration shows promise for quantitative reconstruction of optical absorption in photoacoustic tomography. However, there are issues that prevent the technique from being practical including: non-uniqueness of scattering and absorption profiles, divergence with over-iteration, and sensitivity to noise. Multiple illumination has been proposed to deal with the first problem, and may help with the second. The issue of noise may be balanced out by increasing the regularization parameter at the expense of the exactness of the reconstruction. In a multiple-illumination setup with a circular geometry where fluence is abundant, using a patterned illumination with a decoding step may provide an alternative which will boost SNR. We present a simple sequence of patterned illuminations based on an S-sequence that serves to improve SNR. While the forward model of the iterative method may be applied directly to the patterned excitations, including the decoding step improves SNR in an individual image by a factor equal to the size of the S-sequence, thus greatly improving convergence for a given value of regularization and SNR. For example, with 15 illuminations, 50-60dB noise levels with S-sequence patterned illuminations gives similar simulated performance to the 70dB case with single-source illuminations. This technique will allow the application of fixed-point iteration techniques in a broader range of SNR conditions without resorting to averaging.
机译:定点迭代显示了在光声层析成像中定量重建光吸收的希望。但是,存在使该技术不实用的问题,包括:散射和吸收曲线的不唯一性,过度重复的发散以及对噪声的敏感性。已经提出了多重照明来解决第一个问题,并且可能有助于第二个问题。可以通过增加正则化参数来平衡噪声的问题,但要以重建的准确性为代价。在通量丰富的具有圆形几何形状的多重照明设置中,将带图案的照明与解码步骤一起使用可以提供提高SNR的替代方法。我们提出了一个基于S序列的图案化照明的简单序列,该序列可用于改善SNR。虽然迭代方法的前向模型可以直接应用于图案化激发,但包括解码步骤,可以将单个图像中的SNR提高等于S序列大小的因数,从而在给定正则化值的情况下大大提高了收敛性和SNR。例如,在15个照明条件下,使用S序列模式照明的噪声水平为50-60dB,与在单光源照明条件下的70dB情况下具有相似的模拟性能。该技术将允许定点迭代技术在较宽的SNR条件范围内应用,而无需求平均值。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号