A bio-impedance System for Fluid Monitoring in Dengue Disease

Abstract

Dengue disease is a mosquito borne viral fever widespread in tropical and subtropical regions of the world. The incidence of the disease has rapidly grown over the last decade. There is no specific course of medication to treat dengue. But determining the intensity of fluid leakage is essential in managing the progression of the disease. Bioimpedance technology is an inviting tool for safe, inexpensive and non-invasive monitoring of body fluids. However existing systems are not attractive for continuous and long term assessments. In addition the system performance is also susceptible to variations in posture and motion related artifacts. This work presents a motion aware bioimpedance measurement system (AMBICA) for fluid monitoring using AD5940 and accelerometers to overcome the mentioned limitations. The primary aim of this thesis is to characterize the performance of the developed device with a commercial device(BIACorpus). Additionally a framework to detect the variation in posture and motion is presented.Bioimpedance and movement data was collected from 14 healthy volunteers. The impedance values measured by the device showed high correlation (r>0.98) while the reactance values showed weak correlation (0.10 < r < 0.67). The bias and limits of agreement estimated for impedance measurements are -4.50_13.01 and [21.0,-30.0] respectively. The estimated bias and limits of agreement for reactance measurements are 14.84_19.64, the limits of agreement are [53.36,-23.68]. The precision scores indicated a higher variability in measurements for BIACorpus device compared to AMBICA. The proposed framework was able to identify the active and inactive periods and gross postures like supine, sitting and standing with an accuracy of >95% and 94.6% respectively. The results in this study showed that the impedance and reactance values followed a typical trend agreeable with bioimpedance measurements. However, the values measured by AMBICA agree more closely with BIACorpus for impedance measurements than for reactance measurements. Inconsistencies are observed for both devices and they are attributed to parasitics and imbalance of interface impedances at the inputs of the AD5940. Due to the difference in technologies used, the values from the devices do not measure the exact same values. Therefore values above or below the reference values may still have clinical significance. AMBICA can be used in the future for fluid assessments in dengue affected subjects. Furthermore motion related information gathered using the framework proposed can be further incorporated to develop contextual bioimpedance monitoring applications.