A Power Efficient Analog Front End for Continuous Ultrasound Imaging of the Bladder

Abstract

Continuous bladder monitoring is important for the monitoring of bedridden patients. One method to continuously monitor the bladder is to capture ultrasound images and use machine learning processing to measure the bladder volume from these images. Circuits for implementing these functions can be integrated onto a wearable device, and each of these functions can be integrated onto a single chip. In this thesis, we analyze ultrasound imaging in the context of the bladder to come up with algorithms and hardware to perform continuous bladder monitoring. We first assemble a discrete setup which can form ultrasound images. Using this setup, we describe a new algorithm for generating an ultrasound image by to power gate the hardware during the imaging process to save additional power when capturing the image. We combine these concepts into a single Analog Front End (AFE) chip that can capture images in a power efficient manner.