An Automated System for Frequency-Differential Data Collection in EIT

Abstract Electrical Impedance Tomography (EIT) is an innovative imaging modality that offers significant advantages for real-time monitoring due to its safety, low cost, and portability. In domains such as tumor detection, frequency-difference EIT (fdEIT) is the preferred EIT modality but there is a lack of large non-simulated datasets. To address this gap, we introduce a novel automated system that utilizes a repurposed 3D printer for accurate, automated positioning of a target within an experimental tank. We collect a frequency-differential dataset with a potato target and assess the feasibility and performance of our setup, using a GREIT-type matrix derived from our data for image reconstruction. We evaluate the system’s performance using GREIT figures of merit and compare the results to time-differential EIT (tdEIT) reconstructions. Our results show that while fdEIT presents a more complex challenge than tdEIT particularly in the central regions of the imaging field, the system’s performance is robust, with potential improvements achievable through alternative electrode configurations or advanced algorithms. This study demonstrates the capabilities of our new system in generating large, real-world fdEIT datasets.