2023 年 Annual61 巻 Abstract 号 p. 146_2
A synchronized measurement and time-varying noise cancellation in multi-node electrical impedance tomography (mnEIT) has been proposed for muscle compartment visualization under electrical muscle stimulation (EMS). mnEIT has two features which are 1) network hardware for synchronized multi-node measurement and 2) embedded nodal fast Fourier transform (enFFT) for cancelling time-varying noise. Under the experiments protocol, the muscle compartments response in the upper-arm and thigh of eight healthy subjects are imaged by mnEIT under four voltage intensity levels of EMS 𝑙. According to the result of pre- and during EMS training, the reconstructed images of conductivity distribution 𝛔 trend of the upper arm 𝛔𝑎𝑟𝑚 and thigh 𝛔𝑡ℎ𝑖𝑔ℎ are enlarged along to the increment of 𝑙. Here, it is shown that, during EMS training the magnitude of 𝛔𝑎𝑟𝑚 response in the biceps is stronger than triceps in particular. This phenomenon occurs because the upper arm’s biceps are dominated by white muscle fibers. On the other hand, the 𝛔𝑡ℎ𝑖𝑔ℎ at biceps was equivalent to the hamstrings compartment. Further, as the hardware error evaluation of the mnEIT, the normalized mean impedance error 𝜀𝑓 is
relatively proportional to the frequency increment, while the average error 〈𝜀⟩ is 7. 521% respectively. In this case, the mnEIT has the best result at 𝜀𝑓<〈𝜀⟩ which occurred at 𝑓≤ 2500 Hz that agreed the frequency selection 𝑓1 = 500 Hz and 𝑓2 = 1000 Hz to obtain the best visualization of 𝛔.
