Bayesian approach with the maximum entropy principle in image reconstruction from microwave scattered field data

The Maximum Entropy Method in Ultrasonic Non-Destructive Testing-Increasing the Resolution, Image Noise Reduction and Echo Acquisition Rate

The use of linear methods, for example, the Combined Synthetic Aperture Focusing Technique (C–SAFT), does not allow one to obtain images with high resolution and low noise, especially structural noise in all cases. Non-linear methods should improve the quality of the reconstructed image. Several examples of the application of the maximum entropy (ME) method for ultrasonic echo processing in order to reconstruct the image of reflectors with Rayleigh super-resolution and a high signal-to-noise ratio are considered in the article. The use of the complex phase-shifted Barker code signal as a probe pulse and the compression of measured echoes by the ME method made it possible to increase the signal-to-noise ratio by more than 20 dB for the image of a flat-bottom hole with a diameter of 1 mm in a model experiment. A modification of the ME method for restoring the reflector image by the time-of-flight diffraction (TOFD) method is considered, taking into account the change of the echo signal shape, depending on the depth of the reflector. Using the ME method, 2.5D-images of models of dangling cracks in a pipeline with a diameter of 800 mm were obtained, which make it possible to determine their dimensions. In the object with structural noise, using the ME method, it was possible to increase the signal-to-noise ratio of the reflector image by more than 12 dB. To accelerate the acquisition of echoes in the dual scan mode, it is proposed to use code division multiple access (CDMA) technology based on simultaneous emission by all elements of the array of pseudo-orthogonal signals. The model experiment showed the effectiveness of applying the ME method.

Adaptation and focusing of optode configurations for fluorescence optical tomography by experimental design methods

Fluorescence tomography excites a fluorophore inside a sample by light sources on the surface. From boundary measurements of the fluorescent light, the distribution of the fluorophore is reconstructed. The optode placement determines the quality of the reconstructions in terms of, e.g., resolution and contrast-to-noise ratio. We address the adaptation of the measurement setup. The redundancy of the measurements is chosen as a quality criterion for the optodes and is computed from the Jacobian of the mathematical formulation of light propagation. The algorithm finds a subset with minimum redundancy in the measurements from a feasible pool of optodes. This allows biasing the design in order to favor reconstruction results inside a given region. Two different variations of the algorithm, based on geometric and arithmetic averaging, are compared. Both deliver similar optode configurations. The arithmetic averaging is slightly more stable, whereas the geometric averaging approach shows a better conditioning of the sensitivity matrix and mathematically corresponds more closely with entropy optimization. Adapted illumination and detector patterns are presented for an initial set of 96 optodes placed on a cylinder with focusing on different regions. Examples for the attenuation of fluorophore signals from regions outside the focus are given.

Regularization for ultrasonic measurements of tissue displacement vector and strain tensor

In this report, we present the regularization results of displacement vector measurement by the multidimensional cross-spectrum phase gradient method (MCSPGM), multidimensional autocorrelation method (MAM), and multidimensional Doppler method (MDM) without lateral modulation. In addition to the results of spatially variant regularization in displacement vector measurement using displacement variances, the results of displacement component-dependent regularization are presented. That is, according to the measurement accuracy of displacement components, the regularization is properly applied to the respective displacement components. For instance, only the regularization on the lateral displacement is effective. That is, for the lateral nonmodulation case, the accuracies and stabilities of lateral/elevational displacement measurements significantly increase. In conjunction, the convergence speed of phase matching also increases. The demonstrated measurements of the displacement vector distributions in experiments using inhomogeneous shear modulus agar phantoms confirm that displacement-componentdependent regularization enables better strain tensor measurement and shear modulus reconstruction from the viewpoints of accuracy and stability. The contrast-to-noise ratio (CNR) is useful to set regularization parameters properly for the displacement regularization, whereas the CNR is useless for shear modulus regularization in the sense that the accuracy of reconstruction value cannot be evaluated.

Regularization of tissue shear modulus reconstruction using strain variance

An effective setting method (that is, a method using the variances of strain tensor component measurements) is described for the properly spatially varied regularization parameters for our shear modulus reconstruction. At each position, the respective strain variances can be experimentally evaluated using plural field measurements or single field measurement, for example, when using all crosscorrelation- based methods, by using the Ziv-Zakai Lower Bound (ZZLB). The demonstrated regularization by the single field measurement using the cross-spectrum phase gradient method (MCSPGM) in experiments confirms that the use of the axial strain variance estimated by the echo signal-to-noise ratio and correlations (the combined SNRc) effectively stabilizes the 1-D reconstruction on an agar phantom and a human in vivo liver carcinoma during interstitial microwave thermal treatment. The regularization yields a spatially uniform stability in reconstruction.

Higher-order (nonlinear) diffraction tomography: reconstruction algorithms and computer simulation

The usual propagation transform of diffraction tomography is generalized into higher-order (nonlinear) propagation transforms via use of the Born series as the data-generating model in scattering experiments. Nonlinear tomographic reconstruction algorithms are developed for inversion of scattered field data modeled up to an arbitrarily large (possibly infinite) number of terms in the Born series. A computer simulation study is included to illustrate the performance of the algorithms for the case of scattering objects with cylindrical symmetry.

Image restoration and reconstruction with a Bayesian approach

We have extended Johnson's Bayesian method for image restoration and reconstruction by introducing diagonal line sites, using symmetric neighborhood configurations, and employing an additional hyperparameter for estimation of line sites. A general formulation for arbitrary neighborhood configurations was derived. The major part of this paper deals with the conduct of computer simulations intended to examine the effect of the hyperparameters, the diagonal line sites, and the size of the neighborhood configuration on the performance of the proposed Bayesian method. We show that, for optimal performance, distinct hyperparameters should be used for the intensity sites and line sites. The results also suggest that a large neighborhood configuration should be used. By comparing the near-optimal restored images, we demonstrated that the use of diagonal line sites, along with the symmetric configurations thus made possible, can effectively remove the blocky edge artifacts and produce images of better quality. When the method was applied to positron emission tomography (PET) image reconstruction, our results showed that the quality of the reconstructed images was improved for both computer-simulated and real patient PET data.