A deep neural network for estimating the bladder boundary using electrical impedance tomography.

OBJECTIVE: Accurate bladder size estimation is an important clinical parameter that assists physicians, enabling them to provide better treatment for patients who are suffering from urinary incontinence. Electrical impedance tomography (EIT) is a non-invasive medical imaging method that estimates organ boundaries assuming that the electrical conductivity values of the background, bladder, and adjacent tissues inside the pelvic domain are known a priori. However, the performance of a traditional EIT inverse algorithm such as the modified Newton-Raphson (mNR) for shape estimation exhibits severe convergence problems as it heavily depends on the initial guess and often fails to estimate complex boundaries that require greater numbers of Fourier coefficients to approximate the boundary shape. Therefore, in this study a deep neural network (DNN) is introduced to estimate the urinary bladder boundary inside the pelvic domain. APPROACH: We designed a five-layer DNN which was trained with a dataset of 15 subjects that had different pelvic boundaries, bladder shapes, and conductivity. The boundary voltage measurements of the pelvic domain are defined as input and the corresponding Fourier coefficients that describe the bladder boundary as output data. To evaluate the DNN, we tested with three different sizes of urinary bladder. MAIN RESULTS: Numerical simulations and phantom experiments were performed to validate the performance of the proposed DNN model. The proposed DNN algorithm is compared with the radial basis function (RBF) and mNR method for bladder shape estimation. The results show that the DNN has a low root mean square error for estimated boundary coefficients and better estimation of bladder size when compared to the mNR and RBF. SIGNIFICANCE: We apply the first DNN algorithm to estimate the complex boundaries such as the urinary bladder using EIT. Our work provides a novel efficient EIT inverse solver to estimate the bladder boundary and size accurately. The proposed DNN algorithm has advantages in that it is simple to implement, and has better accuracy and fast estimation.

An oppositional biogeography-based optimization technique to reconstruct organ boundaries in the human thorax using electrical impedance tomography.

Electrical impedance tomography (EIT) is a non-invasive imaging modality which has been actively studied for its industrial as well as medical applications. However, the performance of the inverse algorithms to reconstruct the conductivity images using EIT is often sub-optimal. Several factors contribute to this poor performance, including high sensitivity of EIT to the measurement noise, the rounding-off errors, the inherent ill-posed nature of the problem and the convergence to a local minimum instead of the global minimum. Moreover, the performance of many of these inverse algorithms heavily relies on the selection of initial guess as well as the accurate calculation of a gradient matrix. Considering these facts, the need for an efficient optimization algorithm to reach the correct solution cannot be overstated. This paper presents an oppositional biogeography-based optimization (OBBO) algorithm to estimate the shape, size and location of organ boundaries in a human thorax using 2D EIT. The organ boundaries are expressed as coefficients of truncated Fourier series, while the conductivities of the tissues inside the thorax region are assumed to be known a priori. The proposed method is tested with the use of a realistic chest-shaped mesh structure. The robustness of the algorithm has been verified, first through repetitive numerical simulations by adding randomly generated measurement noise to the simulated voltage data, and then with the help of an experimental setup resembling the human chest. An extensive statistical analysis of the estimated parameters using OBBO and its comparison with the traditional modified Newton-Raphson (mNR) method are presented. The results demonstrate that OBBO has significantly better estimation performance compared to mNR. Furthermore, it has been found that OBBO is robust to the initial guess of the size and location of the boundaries as well as offering a reasonable solution when the a priori knowledge of the conductivity of the organs is not very accurate.