News Column

Patent Issued for System and Method for Improved Real-Time Cine Imaging

August 18, 2014



By a News Reporter-Staff News Editor at Cardiovascular Week -- From Alexandria, Virginia, NewsRx journalists report that a patent by the inventors Ding, Yu (Columbus, OH); Simonetti, Orlando P. (Columbus, OH); Chung, Yiu-Cho (Columbus, OH), filed on March 11, 2013, was published online on August 5, 2014 (see also The Ohio State University Research Foundation).

The patent's assignee for patent number 8798351 is The Ohio State University Research Foundation (Columbus, OH).

News editors obtained the following quote from the background information supplied by the inventors: "In magnetic resonance imaging (MRI), the requirements for acquisition speed (or temporal resolution), spatial resolution and signal-to-noise ratio (SNR) often compete with each other. This competition is especially apparent in real-time MRI where spatial resolution and SNR performance may be sacrificed to achieve the temporal resolution needed to capture dynamic physiological events. For example, parallel imaging techniques such as SMASH [1] and SENSE [2] greatly reduce scan time, facilitating dynamic real-time imaging of rapid physiological processes such as cardiac motion. However, these techniques come with a SNR penalty and may compromise the diagnostic value of the images.

"Filters based on linear transforms, such as the Fourier transform, are often used for image denoising. Spatial low-pass filters are commonly used to improve SNR of individual images [3]. In dynamic imaging, SNR may be further improved by temporal filtering. Averaging and spectral filtering are two commonly used Fourier transform based temporal filtering techniques [4].

"Linear transform based filters have three common steps: (i) transform image/images to a linear combination of 'modes'; (ii) truncate the insignificant 'modes' and, (iii) inverse transform the remaining significant modes to reconstruct the filtered image/images. The optimal number of 'modes' truncated in (ii) must balance the SNR improvement with image sharpness and fidelity. An optimal linear transform concentrates information into fewer 'modes', allowing more irrelevant 'modes' to be truncated, and gives SNR gains with minimal information loss. The selection of the optimal transform and optimal filter cutoff are primary considerations in the design of any linear transform filters for image denoising.

"Dynamic imaging of physiological processes generates series of images that usually show a high degree of temporal correlation. Physiological motion or signal changes are often periodic (e.g., cardiac or respiratory motion), or slowly varying (e.g., contrast agent-induced signal changes) compared to the temporal resolution of the imaging technique. Typically, either condition leads to series of images with substantially similar features in the temporal dimension."

As a supplement to the background information on this patent, NewsRx correspondents also obtained the inventors' summary information for this patent: "Exemplary embodiments of the present invention may be directed to a filter based on the Karhunen-Loeve Transform (KLT) along the temporal direction for denoising real-time dynamic magnetic resonance (CMR) images.

"In general, the KLT is an optimal signal compression method in the least squares sense. The KLT is a linear transform that exploits signal correlations to reduce a high dimensional data set (a large number of modes) into lower dimensions [3] (with information concentrated into a smaller number of modes) by transforming the original data into a set of orthogonal eigenimages. Using this transform, most of the variance is contained in the first few eigenimages. By removing the eigenimages associated with low variance, noise may be reduced.

"Noise reduction by KLT filter was first described by Sychra et al. [5]. It has also been proposed for SNR improvement in 20 images [6, 7], including cardiac nuclear scintigraphy imaging [8, 9, 10]. In prior methods using noise reduction by KLT filters for cardiac imaging applications, images of the heart were limited to one single cardiac cycle. While the images within one heartbeat are correlated, redundancy is limited because the heart changes shape from one image frame to the next. Conversely, when images span multiple heartbeats, redundancy is increased and the KLT may achieve higher levels of noise reduction. KLT filtering has also been proposed for CMR perfusion imaging [11].

"Although KLT filters have been proposed for other medical imaging applications, the details of KLT image-filter design have not been well addressed. More specifically, the choice of eigenimage cutoff and its effect on SNR gain and image sharpness have not been fully understood. The lack of such knowledge may result in a suboptimal use of an optimal transform filter.

"The Karhunen-Loeve Transform Filter improves signal to noise ratio in dynamic imaging. The Karhunen-Loeve (KL) transform (also called PCA), when applied to a series of dynamically changing images, reduces noise while introducing minimal or no blurring to individual images. This KL Transform based filter (herein referred to as KLT filter) exploits the fact that dynamic images have much stronger correlation than the noise in the temporal dimension. The KLT filter works best in periodic, quasi-periodic, and near-stationary slow varying dynamic images. It may be applied to cardiac magnetic resonance imaging (MRI). Dynamic cardiac MR images acquired over multiple cardiac cycles can take best advantage of the KLT filter properties. It is very important to note that the application of KLT need not to be confined to MRI dynamic images. The KLT filter can be applied to any dynamic data set, including, but not limited to perfusion imaging of other organs such as kidney, breast, or real-time kinematic MRI, or images acquired by other modalities such as ultrasound, X-ray fluoroscopy, computed tomography, or nuclear scintigraphy.

"Exemplary embodiments of the cine imaging filter for use in real-time cardiac imaging may include a denoising image-filter based on the Karhunen-Loeve transform along the temporal direction to take advantage of the high temporal correlation among images. Exemplary embodiments of the cine imaging filter may include the application of a simple formula describing the quantitative noise reduction capabilities of the KLT filter as a function of eigenimage cutoff. Additionally, exemplary embodiments may validate its accuracy in numerical simulation and in in-vivo real time cine images. Furthermore, exemplary embodiments of the cine imaging filter may employ a technique to automatically select the optimal eigenimage cutoff to maximize noise reduction with minimal effect on image information. Additionally, some exemplary embodiments of the KLT filter and method of use, with no need for a priori information, may reduce noise in dynamic real-time cardiac magnetic resonance images acquired with high parallel acceleration rates while preserving image quality. Exemplary embodiment of the cine imaging filter may be feasible and practical to overcome the low SNR often encountered in highly accelerated real-time dynamic cardiac MRI."

For additional information on this patent, see: Ding, Yu; Simonetti, Orlando P.; Chung, Yiu-Cho. System and Method for Improved Real-Time Cine Imaging. U.S. Patent Number 8798351, filed March 11, 2013, and published online on August 5, 2014. Patent URL: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=8798351.PN.&OS=PN/8798351RS=PN/8798351

Keywords for this news article include: Cardiology, Magnetic Resonance, The Ohio State University Research Foundation.

Our reports deliver fact-based news of research and discoveries from around the world. Copyright 2014, NewsRx LLC


For more stories covering the world of technology, please see HispanicBusiness' Tech Channel



Source: Cardiovascular Week


Story Tools






HispanicBusiness.com Facebook Linkedin Twitter RSS Feed Email Alerts & Newsletters