Search

Multi-frequency Electrical Impedance Tomography (mfEIT) is an emerging biomedical imaging modality to reveal frequency-dependent conductivity distributions in biomedical ap-plications. Conventional model-based image reconstruction meth-ods suffer from low spatial resolution, unconstrained frequency correlation and high computational cost. Deep learning has been extensively applied in solving the EIT inverse problem in biomed-ical and industrial process imaging. However, most existing learning-based approaches deal with the single-frequency setup, which is inefficient and ineffective when extended to the multi-frequency setup. This paper presents a Multiple Measurement Vector (MMV) model based learning algorithm named MMV-Net to solve the mfEIT image reconstruction problem. MMV-Net considers the correlations between mfEIT images and unfolds the update steps of the Alternating Direction Method of Multipliers for the MMV problem (MMV-ADMM). The non-linear shrinkage operator associated with the weighted l2,1 regularization term of MMV-ADMM is generalized in MMV-Net with a cascade of a Spatial Self-Attention module and a Convolutional Long Short-Term Memory (ConvLSTM) module to better capture intra- and inter-frequency dependencies. The proposed MMV-Net was validated on our Edinburgh mfEIT Dataset and a series of comprehensive experiments. The results show superior image quality, convergence performance, noise robustness and computational efficiency against the conventional MMV-ADMM and the state-of-the-art deep learning methods.

The intrinsic issue of low spatial resolution of Electrical Impedance Tomography (EIT) is a long-standing challenge that hinders the capability of performing quantitative analysis based on EIT image. Our recent work demonstrates an impedance-optical dual-modal imaging framework and a deep learning model named Multi-Scale Feature Cross Fusion Network (MSFCF-Net) to realize information fusion and high-quality EIT image reconstruction. However, this framework’s performance is limited by the accuracy of the mask image obtained from an auxiliary imaging modality. This paper further proposes a two-stage deep neural network, which is the enhanced version of the MSFCF-Net (named En-MSFCF-Net), to automatically improve mask image and conduct information fusion and image reconstruction. Compared to MSFCF-Net, En-MSFCF-Net demonstrates the superior ability to correct the inaccurate mask image, leading to a more accurate conductivity estimation. Furthermore, the En-MSFCF-Net also maintains the best shape preservation and conductivity prediction accuracy among given learning-based and model-based algorithms. Both qualitative and quantitative results indicate that En-MSFCF-Net could make dual-modal imaging more robust in real-world situations.

Real-time, non-destructive and label-free imaging of 3-D cell culture process using miniature Electrical Impedance Tomography (mEIT) is an emerging topic in tissue engineering. Image reconstruction of mEIT for cell culture is challenging due to weak sensing signals and increased sensitivity to sensor imperfection.Conventional regularization based image reconstruction methods can not always achieve satisfactory performance in terms of image quality and computational efficiency for this particular setup. Recent advances of deep learning have pointed out a promising alternative. However, with a single neural network, it is still difficult to reconstruct multiple objects with varying conductivity levels, which cases are widespread in the application of cell imaging. Aiming at this challenge, in this paper we propose a deep learning and group sparsity regularization based hybrid algorithm for cell imaging with mEIT. A deep neural network is proposed to estimate the structural information in form of binarymasks given the limited amount of data sets. Then the structural information is encoded in group sparsity regularization to obtain the final estimation of conductivity. The proposed approach is validated by both simulation and experimental data on MCF-7 human breast cancer cell aggregates, which demonstrates its superior performance and generalization ability compared witha number of existing algorithms.

A 3-D electrical impedance tomography (EIT) is an emerging technique for real-time and non-destructive 3-D cell culture imaging. This paper presents a pioneering study of scaffold-based 3-D cell culture imaging using a miniature planar EIT sensor. A 17-electrode miniature-planar EIT sensor equipped with a regular-shape 3-D printed scaffold was manufactured, modeled, and characterized. In addition, an efficient 3-D image reconstruction method based on 3-D isotropic total variation and ${l}$ 1 joint regularization was proposed. The numerical simulation on scaffold phantoms and the experimental study on time-varying distribution of MCF-7 cancer cell suspension within the scaffold were performed. Both the simulation and experiment results suggest that using the miniature EIT sensor and the developed 3-D image reconstruction algorithms are able to achieve high quality, non-destructive scaffold-based 3-D cell culture imaging.

Electrical impedance tomography (EIT) has the potential to be an alternative technique to microscopy for many biological applications. It is a novel approach to monitor cell viability, proliferation, or drug response with high temporal resolution in a label-free, non-toxic, and non-destructive way. To limit measurement errors, only time-difference EIT is currently used for biological imaging. Still, the need for constant background reference limits its uses in long-term imaging. A novel equivalent circuit model was developed to analyze the errors in frequency-difference EIT (FDEIT) measurements. Based on the circuit model, the calibrated FDEIT (CFDEIT) method has been derived to compensate the measurement errors. The simulation results show that the voltage variations introduced by the leakage current are correctly modeled by the equivalent circuit. Significant improvements in the image quality are observed in both simulation and experimental results after the application of CFDEIT. It indicates that this paper improves the accuracy of FDEIT and makes it feasible for 3D tissue culture monitoring.

We aimed to study the time course decrease of human retinal pigment epithelium (RPE) barrier function when exposed to blue light. To this end, we cultured ARPE-19 cells on Electrical Cell-substrate Impedance Sensing (ECIS) multi-well arrays. Using an ad hoc light emitting diode (LED) array illumination system together with a set of neutral density filters and a 3-dimensional (3D) printed filter holder, cells were exposed to a gradient of irradiances of blue-light with a measured peak at 468 nm. The electrical resistance between 4 kHz and 64 kHz was recorded during the exposure. Blue light exposure induced a dose-dependent decrease in the resistances at 4 kHz, however the time course resistance at 64 kHz did not show any decrease before t = 52 h. Quantification of the barrier function using mathematical model integrated in the ECIS software showed that blue-light exposure induced a dose-dependent decrease in the barrier function associated with tight junction formation (P 0.05). This was confirmed by the immunostaining of the tight-junction associated structural protein, Zonula occludens-1 (ZO-1). The detection of reactive oxygen species by carboxy-H2DCFDA confirmed that the blue light induced dose-dependent decrease in the barrier function is mediated by oxidative stress. On a separate experiment, blue-light exposed ARPE-19 cells were treated with 100 nM Protein Kinase C zeta (PKC-ζ) pseudo substrate inhibitor to identify underlying pathway for blue-light induced damage on the barrier function. The treatment with 100 nM PKC-ζ pseudo substrate inhibitor induced faster recovery of the barrier function compared to no treatment. Altogether our results document that blue LED light exposure decreased RPE barrier function in-vitro in a dose-dependent manner, before any cell death occurred. This damage induced by blue-light on tight junctions is mediated by oxidative stress through PKC-ζ activation. The quantification of the healing effect observed by inhibition of PKC-ζ might lead to development of high throughput wound healing assays through ECIS in the future.

Real-time quantitative imaging is becoming highly desirable to study nondestructively the biological behavior of 3-D cell culture systems. In this paper, we investigate the feasibility of quantitative imaging/monitoring of 3-D cell culture processes via electrical impedance tomography (EIT), which is capable of generating conductivity images in a non-destructive manner with high temporal resolution. To this end, a planar miniature EIT sensor amenable to standard cell culture format is designed, and a 3-D forward model for the sensor is developed for 3-D imaging. Furthermore, a novel 3-D-Laplacian and sparsity joint regularization algorithm is proposed for enhanced 3-D image reconstruction. Simulation phantoms with spheres at various vertical and horizontal positions were imaged for 3-D performance evaluation. In addition, experiments on human breast cancer cell spheroid and a triangular breast cancer cell pellet were carried out for experimental verification. The results have shown that the stable measurement on high conductive cell culture medium and the significant improvement of image quality based on the proposed regularization method are achieved. It demonstrates the feasibility of using the miniature EIT sensor and 3-D image reconstruction algorithm to visualize 3-D cell cultures, such as spheroids or artificial tissues and organs. The established work would expedite real-time quantitative imaging of 3-D cell culture for assessment of cellular dynamics.

Here we demonstrate the ability to characterize microporous scaffolds and evaluate cell concentration variation via the utilization and interpretation of complex permittivity measurements (CP), a direct and nondestructive method. Polymer-based microporous scaffolds are of importance to tissue engineering, particularly in the promotion of cell adhesion, proliferation, and differentiation in predefined shapes. Chitosan gel scaffolds were seeded with increasing concentrations of macrophages to simulate cell growth. Complex permittivity measurements were performed using a dielectric probe and a vector network analyzer over a frequency ranging from 200 MHz to 2 GHz. An effective medium theory was applied to interpret the data obtained; respectively, Looyenga and Maxwell-Wagner-Hanai functions were used to retrieve the porosity and the variation of the cell concentration from the CP measurements. Calculated porosities were in agreement with experimental evaluation—porosity ranged from 81–96%. Changes in cell concentration inside the scaffolds upon injection of differing cell concentrations into the scaffold were detected distinguishably. Variations resulting from the cumulative injection of 400–1800 μL of 106 cells/mL solution into the scaffold were monitored. Results suggest that CP measurements in combination with an appropriate effective medium approximation can enable on-line monitoring of cell growth within scaffolds. © 2003 Wiley Periodicals, Inc. Biotechnol Bioeng 84: 343–350, 2003.

Establishing a relationship between perfusion rate and fluid shear stress in a 3D cell culture environment is an ongoing and challenging task faced by tissue engineers. We explore Doppler optical coherence tomography (DOCT) as a potential imaging tool for in situ monitoring of local fluid flow profiles inside porous chitosan scaffolds. From the measured fluid flow profiles, the fluid shear stresses are evaluated. We examine the localized fluid flow and shear stress within low-and high-porosity chitosan scaffolds, which are subjected to a constant input flow rate of 0.5 ml l min(-1). The DOCT results show that the behavior of the fluid flow and shear stress in micropores is strongly dependent on the micropore interconnectivity, porosity, and size of pores within the scaffold. For low-porosity and high-porosity chitosan scaffolds examined, the measured local fluid flow and shear stress varied from micropore to micropore, with a mean shear stress of 0.49 +/- 0.3 dyn.cm(-2) and 0.38 +/- 0.2 dyn.cm(-2), respectively. In addition, we show that the scaffold's porosity and interconnectivity can be quantified by combining analyses of the 3D structural and flow images obtained from DOCT. (C) 2009 Society of Photo-Optical Instrumentation Engineers. [DOI: 10.1117/1.3130345]