Method of Spatially-encoded MRS Using Preselected Compartments an Linear Algebraic Modeling

Case ID:
C11462
Disclosure Date:
3/11/2011

C11462: Novel Method of Performing Localized Magnetic Resonance Spectroscopy

Novelty:

This is a reduced-to practice method for optimizing both scan time and signal-to-noise (SNR) ratio in Magnetic Resonance Spectroscopy metabolic studies.

Value Proposition:

Scan time and SNR are major problems for magnetic resonance spectroscopy (MRS) of low-concentration metabolites. Model-based MRS reconstruction methods, which can reduce scan time, have not been realized in humans. Matching voxels to anatomical features prior to acquisition yields intrinsically higher signal to noise ratio (SNR) than summing chemical-shift-imaging (CSI) voxels post-acquisition. Localized spectroscopy using linear algebraic modeling (SLAM) yields spectra approximating the average of pre-selected compartments, with dramatic SNR and speed gains vs. standard CSI. Advantages include:

• 400% faster scan than CSI while maintaining quality .
• SNR gains of 30%-200% over CSI with same scan time .
• First ever implementation of optimized gradient sets that are fractional .
• Inter-compartmental bleeding controllable by selecting different phase-encode gradients

Technical Details:

Johns Hopkins researchers have developed a method of performing spatially localized magnetic resonance with enhanced scan-time and/or signal-to-noise ratio. After an initial scout MRI scan, the image is segmented into compartments of interest which will generate MRS signals, used to inform the voxel size. Then a number of phase-encoding steps (equivalent to averaging) greater or equal to the number of compartments are chosen to maximize the desired properties of the reconstruction, such as SNR or the level of compartmental leakage. Typically these phase-encoding steps are close to the center of the CSI k-space with high SNR. SLAM then solves a subset of simultaneous linear equations from standard CSI to reconstruct the spectra. This eliminates unneeded phase-encoding steps to achieve comparable quality in much less time.

Looking for Partners:

To develop and commercialize the technology for performing clinical or research localized magnetic resonance spectroscopic (MRS) studies on MRI and or NMR equipment. The technology will enable incorporation of a sped-up diagnostic MRS exam, as a part of a clinical brain MRI study, for example.

Stage of Development:

Prototype

Data Availability:

Method reduced to humans in a clinical setting

Publications/Associated Cases:

J Magn Reson. 2012 May;218:66-76.

Patent Information:
Title App Type Country Serial No. Patent No. File Date Issued Date Expire Date Patent Status
Method of Spatially-encoded MRS Using Preselected Compartments an Linear Algebraic Modeling PCT: Patent Cooperation Treaty European Patent Office 12760705.9 2688471 3/22/2012 6/21/2023 3/22/2032 Granted
SYSTEM AND METHOD OF PERFORMING MAGNETIC RESONANCE SPECTROSCOPIC IMAGING PCT: Patent Cooperation Treaty Japan 2014-501253 6070963 3/22/2012 1/13/2017 3/22/2032 Granted
Method of Spatially-encoded MRS Using Preselected Compartments an Linear Algebraic Modeling PCT: Patent Cooperation Treaty France 12760705.9   3/22/2012     Pending
Method of Spatially-encoded MRS Using Preselected Compartments an Linear Algebraic Modeling PCT: Patent Cooperation Treaty Germany 12760705.9   3/22/2012     Pending
Method of Spatially-encoded MRS Using Preselected Compartments an Linear Algebraic Modeling PCT: Patent Cooperation Treaty United Kingdom 12760705.9   3/22/2012     Pending
Method of Spatially-encoded MRS Using Preselected Compartments an Linear Algebraic Modeling PCT: Patent Cooperation Treaty United States 14/006,069 10,209,330 9/18/2013 2/19/2019 5/7/2036 Granted
Inventors:
Category(s):
Get custom alerts for techs in these categories/from these inventors:
For Information, Contact:
Louis Mari
lmari3@jhu.edu
410-614-0300
Save This Technology:
2017 - 2022 © Johns Hopkins Technology Ventures. All Rights Reserved. Powered by Inteum