Analysis of Elt Image of the Lungs by Fuzzy Black Box Back Propagation Intelligent Technique

Abstract

Electrical Impedance Tomography (EIT) is a functional maging method that is being developed for bedside use in critical care medicine. Aiming at improving the chest anatomical resolution of EIT images, we have developed a fuzzy Black Box back propagation Technique (BBT) based on EIT’s high temporal resolution and the functional information contained in the pulmonary perfusion and ventilation signals. EIT data from an experimental model were collected during normal ventilation and apnea while an injection of hypertonic saline was used as a reference. The fuzzy model was elaborated in three parts: a modeling of the heart, a pulmonary map from ventilation images and a pulmonary map from perfusion images. Image segmentation was performed using a threshold method and a ventilation/perfusion map was generated using Intelligent Black box Back Propagation Technique. EIT images treated by the fuzzy model were compared with the hypertonic saline injection method and CT-scan images, presenting good results in both qualitative (the image obtained by the model was very similar to that of the CT-scan)and quantitative (the ROC curve provided an area equal to 0.97) point of view. Undoubtedly, these results represent an important step in the EIT images area, since they open the possibility of developing EIT-based bedside clinical methods, which are not available nowadays. These achievements could serve as the base to develop EIT diagnosis system for some life-threatening diseases commonly found in critical care medicine.