Date of Graduation


Document Type


Degree Name

Master of Science in Electrical Engineering (MSEE)

Degree Level



Electrical Engineering


Robert Saunders

Committee Member

Wu, Jingxian

Second Committee Member

Jensen, Morten


dehydration assessment, embedded system, peripheral venous pressure, volume assessment


The severe side effects of acute dehydration and blood loss may be prevented if assessed and treated quickly. As amplifier technology has improved, small peripheral venous pressure (PVP) signals collected using a transducer on an intravenous catheter can be analyzed to monitor patient volume status. This thesis details the development of an embedded system prototype that quickly and accurately assesses volume status using PVP analysis. The volume prediction algorithm classified 10 seconds of PVP data as “dehydrated” or “resuscitated”. The hardware design consisted of four primary areas: analog signal conditioning, processing, user interaction, and power. After data collection, the signal was amplified and passed through a low pass Butterworth filter before 14-bit analog-to-digital conversion. One step of the analog-to-digital converter was equivalent to 0.007867mmHg at a gain of 2048V/V. The processor sampled the input signal at 819.2Hz. The processing block on the prototype device was realized using a system on module with a processor that met the memory specifications and had pre-existing Fast Fourier Transform (FFT) software. Predicted patient volume status was displayed via bicolor LEDs after 10 seconds of data collection. The volume status LEDs turned red if a 10 second window of PVP data window was predicted as dehydrated and green if predicted as resuscitated. Real patient PVP data from pediatric patients with hypertrophic pyloric stenosis converted from millimeters of mercury to voltage was utilized to test the algorithm implemented on the prototype using an arbitrary waveform generator. The software on the prototype device successfully recreated 10 seconds of data and output the frequency-domain samples for a sinusoid input and for an isolated 10 second window of patient PVP data input from a dehydrated and a resuscitated patient. The complete continuous data set from each patient was loaded to the device 10 times per patient, and the volume status output recorded. The prototype predicted dehydration at the window-level (sensitivity 93.27%, specificity 91.44%) and at the patient-level (sensitivity 97%, specificity 100%). The accuracy of the prototype was slightly less than that of the MATLAB prediction algorithm on the window-level (sensitivity 97.94%, specificity 93.07%) and the patient-level (sensitivity 100%, specificity 100%). The accuracy of the MATLAB and prototype volume prediction algorithms varied between patients. Evidence supported that a sensitivity or specificity of the prototype below 95% for a given patient was dependent on the input patient PVP data rather than systemic error. The patients with a greater mean and standard deviation in the number of errors also showed window-level dehydration probabilities closer to the decision level of 0.5. Data sets from one resuscitated and one dehydrated patient were combined to create a waveform with an active change in volume status. The active change in volume status was shown on the LED display. Future work includes updating the prototype software for anomaly detection. Also, the accuracy of volume status prediction could be improved by using the majority decision rule on overlapping 10 second windows.

Available for download on Saturday, February 07, 2026