Experimental Validation of Microwave Tomography with the DBIM-TwIST Algorithm for Brain Stroke Detection and Classification

Abstract

We present an initial experimental validation of a microwave tomography (MWT) prototype for brain stroke detection and classification using the distorted Born iterative method, two-step iterative shrinkage thresholding (DBIM-TwIST) algorithm. The validation study consists of first preparing and characterizing gel phantoms which mimic the structure and the dielectric properties of a simplified brain model with a haemorrhagic or ischemic stroke target. Then, we measure the S-parameters of the phantoms in our experimental prototype and process the scattered signals from 0.5 to 2.5 GHz using the DBIM-TwIST algorithm to estimate the dielectric properties of the reconstruction domain. Our results demonstrate that we are able to detect the stroke target in scenarios where the initial guess of the inverse problem is only an approximation of the true experimental phantom. Moreover, the prototype can differentiate between haemorrhagic and ischemic strokes based on the estimation of their dielectric properties.

Expansion of the Nodal-Adjoint Method for Simple and Efficient Computation of the 2D Tomographic Imaging Jacobian Matrix

“…For the tomographic approach, numerous groups have produced simulation studies [ 18 , 19 , 20 , 21 ] on both classic conical structures and with data from actual MR exams [ 22 ]. Several approaches have advanced to phantom and animal studies [ 12 , 23 , 24 , 25 , 26 ] and there have been several patient studies by the Dartmouth College group—particularly in breast cancer neoadjuvant chemotherapy monitoring [ 2 ]. This paper focuses entirely on the tomography approaches since these are the ones that require construction of a Jacobian matrix as part of their image reconstruction process.…”

“…This is due primarily to heavy resources needed for the multiple forward solution calculations at each iteration. These times have generally ranged from several minutes for some 2D implementations [ 27 ] to many hours and even days for larger 3D reconstructions that often require parallelized multi-processors and graphical processing units (GPUs) to accelerate the computation speed [ 18 , 23 , 24 ]. In fact, the EMTensor group is exploring massively parallel, cloud computing to solve the computational issues for their portable and computationally expensive system [ 28 ].…”

“…The technique has been shown to work well with low contrast and small scatter problems when the starting property distribution estimate is that of the surrounding background. This approach is subsequently utilized by Shea et al [ 18 ] and further by Karadima et al [ 23 ]. Zakaria et al [ 36 ] used the finite element contrast source inversion method for microwave imaging and reported in detail on the gradients of the cost function.…”

Microwave Detection of Brain Injuries by Means of a Hybrid Imaging Method

“…Nevertheless, having a pointwise indication about the material of the inhomogeneity inside brain may represent a concrete aid in the diagnosis of these pathologies. This is the application field of quantitative imaging techniques [17]- [21]. The progress in this area is continuous, and some recent advances have been obtained by adopting deep-learning-aware inverse scattering methods [22]- [26].…”

Simulation and Analysis of Variable Antenna Designs for Effective Stroke Detection

“…This concept is now explicitly applied for mapping the human head phantom with the idea to locate the stroke at early stages with clear visualization of its shape, size, and most importantly their numbers with each one's *Corresponding Author 238 | P a g e www.ijacsa.thesai.org precise location. Many prototype devices are now been proposed and developed based on this concept [15][16][17][18][19][20][21][22][23][24][25]. Some are now even popularly applied as a commercial solution [18,23,24].…”