ストッキング圧迫下における脹脛内生理的むくみのスパースベイズ学習導入型電気インピーダンストモグラフィ法（SBL-EIT）による画像化

抄録

Sparse Bayesian learning implemented into electrical impedance tomography (SBL-EIT) has been used to image physiological swelling inside the human calf under stocking compression, aiming to evaluate the treatment effect of various compression pressures. Three categories of stockings, categorized as Strong, Weak, and Control according to the net pressure measured by the pressure sensor, were utilized to assess the calves of six subjects each during prolonged standing. The results revealed that 1) the spatial-mean conductivity 〈σ〉^α2 showed an increased trend for all subjects in the sequence of Control, Weak, and Strong, 2) 〈σ〉^α2 exhibited a significant negative correlation with conventional impedance Z^BIA by bioelectrical impedance analysis (BIA), and 3) SBL-EIT adequately reconstructed the conductivity distribution Δσ and exhibited varied responses to physiological swelling resulting from three different stocking compression pressures.