Pilot Tone Signal Optimization for Cardiac Magnetic Resonance Imaging
Loading...
Date
2021-05
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
The Ohio State University
Abstract
Pilot Tone (PT) is an emerging technology in the field of cardiac magnetic resonance imaging (MRI). Its ability to effectively encode physiological motions has made it an attractive choice for cardiac MRI, where respiratory and cardiac motions are invariably present. PT is advantageous in cardiac MRI because of its high sampling rate and its ability to simultaneously encode cardiac and respiratory motions without interrupting the acquisition process. Typically, the cardiac signal extracted from PT data is of good quality; however, PT data can be excessively noisy and susceptible to electromagnetic interference and bulk motion artifact. To increase the robustness and reliability of the extracted cardiac signal from the PT data, it is imperative to enhance and optimize the data processing procedure.
This thesis focuses on the optimization of cardiac signal extraction and explores the hypothesis that by constructing bandwidth-optimized synthetic (BOSyn) channels from the measured raw PT data, the quality of the extracted cardiac signal can be significantly improved. By localizing the cardiac motion energy in the first few synthetic channels, BOSyn provides an avenue to perform channel pruning. These synthetic channels are generated from a linear combination of the measured PT channels. The weights that combine the measured PT channels into synthetic channels maximize the ratio of energy in the cardiac frequency range, called region-of-interest (ROI), to the energy outside the ROI. The synthetic channels from BOSyn are sorted, with the first synthetic channel representing the highest relative energy in the ROI. After this procedure, the synthetic channels with small relative energy in the ROI are discarded to improve signal quality. The resulting synthetic channels can be processed further for cardiac motion extraction.
A study performed on 14 healthy volunteers' PT data, with electrocardiogram (ECG) data as reference, highlights the merit of the BOSyn-based channel pruning for cardiac motion extraction. In summary, the proposed method was effective in improving the accuracy and precision of the extracted cardiac triggers.
Description
Keywords
Pilot Tone, Data Optimization, Cardiac, Magnetic Resonance Imaging