Nikolaos Bourbakis (Advisor), Soon M. Chung (Committee Member), Catherine Marco (Committee Member), Konstantina Nikita (Committee Member), Yong Pei (Committee Member)
Doctor of Philosophy (PhD)
Ultrasound Imaging (USI) or Medical Sonography (MS), as it is formally called, has been widely used in biomedical applications over the last decades. USI can provide clinicians with a thorough view of the internal parts of the human body, making use of sound waves of higher frequencies than humans can perceive. USI systems are considered highly portable and of low-cost, compared to other imaging modalities. However, despite those advantages, Ultrasound Systems (US) and especially 3D ones, have not been yet extensively utilized for Point-of-Care (POC) applications, due to numerous restrictions and artifacts that they currently present.
Hardware complexity and real-time requirements are considered to be the major restrictions for portable, 3D (volumetric) USI. Volumetric transducers consist of thousands of piezoelectric elements that make the signaling and the networking of the system extremely complex. Additionally, regions of the internal body require significantly long time to be scanned in the three dimensions. Consequently, real-time applications are considered prohibited. Last, but yet equivalently important, most of the low-cost, portable systems manifest artifacts that degrade the quality of ultrasound image. It is obvious that researchers' concern and major challenge is to successfully address those problems and manage offset the strong trade-offs that exist.
Given the aforementioned challenge, the current research work presents a novel low-cost, portable 3D Ultrasound system design, composed of four volumetric transducers. The system has been designed for POC applications in a way to manifest extended imaging capabilities. The use of multiple (four), simple 2D phased array transducers is adopted in order for the system to provide enhanced field of view, as well as automatic scanning of the Region of Interest (ROI) (radiologist intervention-free). In order to deal with the high complexity of the system, the transducers were designed with limited number of elements (256 each) and were integrated to a single FPGA board. To compensate for the image degradation caused by transducers of fewer elements, a new image enhancement methodology was proposed. The methodology targets to image de-speckling and image resolution improvement, given the redundant information provided by the multiple transducers. It uses a combination of spatial and frequency compounding techniques along with a Super-Resolution (SR) algorithm. In order to vindicate the selection of the techniques that were used for the proposed methodology, a parametric study regarding the performance of numerous de-noising and SR techniques was conducted. The performance of the methodology was firstly tested using typical 1D phased array transducers and the results in the 2D images offered promising insights its advantages.
Having verified the effectiveness of the proposed methodology for the case of 2D ultrasound images, the methodology was extended to volumetric images. The final de-noised B-mode images manifested increased Contrast Noise Ratio (CNR) and Signal to Noise Ratio (SNR) compared to various other ultrasound image de-speckling techniques, while at the same time image resolution improvement was observed.
The novel low-cost, portable 3D Ultrasound system design that is proposed, combined with the new image enhancement technique implemented, successfully addresses the existing challenges, in regards to the trade-off between system complexity and image quality. In fact, it not only develops a system of significantly lower complexity but the same time tackles the disadvantages that such a system could have, by integrating in the design the component of image enhancement.
Department or Program
Department of Computer Science and Engineering
Year Degree Awarded
Copyright 2015, all rights reserved. This open access ETD is published by Wright State University and OhioLINK.