ECE Home
About ECE
Why Study Electrical & Computer Engineering at NJIT?
Our People
Academics
Undergraduate Majors
Research
Learning Beyond the Classroom
Careers in ECE & Industry Partners
News & Events
Info for Students
Contact Us
BS in Computer Engineering (CoE)
CoE Mission Statement
CoE Program Objectives
BS in Electrical Engineering (EE)
EE Mission Statement
EE Program Objectives
Senior Design Projects
Advising
Graduate Degree Programs
Ultrasound Measurement Resolution Enhancement with Low Transient Pulse Shaping and Model Reference Baseband Adaptive Processing - PhD Dissertation Defense
Biao Cheng, NJIT
Date : January 09, 2007 (Tuesday)
Time : 11:00 AM - 12:30PM
Location : ECE 202, NJIT
Abstract
Due to its non-invasive, non-ionizing, portable, and low-cost nature, ultrasound has become an indispensable part of various measurement systems such as medical diagnostics and non-destructive testing. However, ultrasound measurements suffer from an intrinsic artifact, caused by the limited bandwidth of the transducer and the inherent phase interference of acoustic waves. Such artifact degrades the measurement resolution and the target information.
The Low Transient Pulse (LTP) method proposed in this work is an innovative technique to produce a short duration and low transient acoustic pulse by means of pre-shaping the excitation signal, resulting into a higher resolution measurement than the conventional methods. To further improve the robustness and computational efficiency of the LTP method, a Model Reference Baseband Adaptive (MRBA) processing algorithm is also proposed. Recognizing that most ultrasonic transducers operate at their passbands, frequency translation technique based on the Hilbert transform is applied to obtain an equivalent baseband model. The MRBA then adaptively estimates the time-of-flight from the baseband model. The use of baseband model results in faster computation while the model reference control improves the overall robustness.
Finally, to fully realize the LTP/MRBA’s efficiency and performance, advanced Digital Signal Processors (DSPs) and microcontrollers are considered with the goal of small size and low cost. The present experimental setup consists of a DSP and a Field Programmable Gate Array (FPGA) for parallel processing with the FPGA generating the LTP signal and the DSP performing the signal processing. It is the final objective of this work to implement the entire software system with on a single microcontroller such as the Freescale MPC566.
Advisor:
Dr. Timothy Chang, Professor, ECE, NJIT
Committee Members:
Dr. Ali Abdi, Assistant Professor, ECE, NJIT
Dr. John Carpinelli, Associate Professor, ECE, NJIT
Dr. Edward Hou, Associate Professor, ECE, NJIT
Dr. Wen-Ching Liu, Assistant Professor, Department of Radiology, UMDNJ