ウェアラブル電気インピーダンストモグラフィに有用な胸郭形状推定法

抄録

Evaluation of lung function is a typical application of electrical impedance tomography (EIT) and a wearable EIT has been developed to give easy use and reliable EIT information in an intensive care medicine. However, chest shape information was important factor to decide measurement accuracy and it was very difficult to measure from a patient on a bedside in an intensive care unit. Therefore, we proposed a novel estimation method of a chest shape using curvature data and chest circumference that would be possible to measure with our wearable EIT. Measurement accuracy was evaluated by applying the proposed method to an ellipse shape with 20cm in X-axis and 10cm in Y-axis, the same ellipse with 20% random noise and two CT images of thorax. To evaluate accuracy of estimated shape, we defined eccentricity ‘e’ that was aspect ratio of estimated shape and compared it between actual and estimated shapes respectively. Error in eccentricity ‘e’ was approximately 3% in ellipse shape without noise, 7% in ellipse shape with noise and 7% in two CT images respectively. The result was considered that our proposed method was useful to estimate shape for accurate EIT measurement using a wearable EIT.