Abstract
Multifrequency electrical impedance tomography (mfEIT) is an emerging biomedical imaging modality that exploits frequency-dependent electrical properties. The mfEIT image reconstruction problem for cell imaging is particularly challenging due to weak signals from miniaturized sensors and high sensitivity to modeling errors. The existing approaches are primarily based on the linearized model and few are applied to the miniaturized setup. Here, we report a mask-guided spatial-temporal graph neural network (M-STGNN) to reconstruct mfEIT images in cell culture imaging. The M-STGNN captures simultaneously spatial and frequency correlations, and the spatial correlation is further constrained by geometric structures from auxiliary binary masks, such as computed tomography (CT) or microscopic images. We validate the mfEIT approach through numerical simulations and experiments on MCF-7 human breast cancer cell aggregates. The results demonstrate the superiority of M-STGNN over the state-of-the-art with an improvement of approximately 10.7% under the experimental setup. It can be readily extended to multimodal biomedical imaging applications.
Original language | English |
---|---|
Article number | 4505610 |
Journal | IEEE Transactions on Instrumentation and Measurement |
Volume | 71 |
Early online date | 10 Aug 2022 |
DOIs | |
Publication status | E-pub ahead of print - 10 Aug 2022 |
Keywords / Materials (for Non-textual outputs)
- Computer architecture
- Conductivity
- Deep learning
- Electrical impedance tomography
- Frequency measurement
- Graph neural networks
- Image reconstruction
- Imaging
- electrical impedance tomography (EIT)
- graph neural network (GNN)
- image reconstruction
- multifrequency