نوع مقاله : مقاله کامل پژوهشی
نویسندگان
گروه مهندسی پزشکی، پژوهشکده برق و فناوری اطلاعات، سازمان پژوهشهای علمی و صنعتی ایران، تهران
کلیدواژهها
عنوان مقاله English
نویسندگان English
Fluid overload in patients with end-stage renal disease is a major factor contributing to increased cardiovascular complications, intradialytic hypotension, and mortality. Therefore, accurate determination of dry weight and monitoring of body fluid volume changes during hemodialysis are of considerable clinical importance. In this study, a multifrequency bioimpedance system was used to record bioimpedance from different body segments of hemodialysis patients. Data were collected from 11 patients during hemodialysis. Bioimpedance was measured in five body segments, including the right arm, left arm, right leg, left leg, and trunk, at frequencies of 50, 150, 250, 500 and 750 kHz. After extracting impedance amplitude and phase, normalized impedance parameters were calculated, and their relationships with ultrafiltration volume were investigated. The results showed that the impedance features obtained from the right arm segment at 50 and 750 kHz exhibited the highest correlations with ultrafiltration volume. The selected parameters were then used as inputs to a feedforward neural network, while the ultrafiltration volume was considered the model output. Model performance was evaluated using indices such as the correlation coefficient (more than 90%) and root mean square error (RMSE) (less than 0.01). The preliminary findings suggest that multifrequency bioimpedance can be considered a noninvasive and low-cost approach for monitoring fluid status in hemodialysis patients. However, given the limited number of patients included in this study, further validation in larger-scale investigations is required.
کلیدواژهها English