Patent Grant
11372068
The present invention relates to signal processing and quantitation in Magnetic Resonance Spectroscopy. Particularly, the invention provides a comprehensive method that offers both automated and interactive signal processing and affords increased accuracy in signal quantitation.
Magnetic resonance spectroscopy (MRS) is an important non-invasive technique for in vivo detection of a wide range of brain neurochemicals. MRS technique has been commonly used for diagnostic purposes, and the utility of MRS is extended to diagnosing chronic and severe low back pain, as disclosed in one of the prior art patents.
The effectiveness of MRS technique is based on the ability to quantify characteristic metabolite spectra, which has various limitations such as small signal to noise ratio, overlapping spectra, distorted baseline due to physical reasons and engineering related concerns, not limited thereto.
Studies have been conducted and have indicated that the choice of quantification methods may affect the interpretation of results and their statistical significance. As such, different processing algorithms may be suitable for processing different MR data types. Indeed, it is preferable to analyze each spectrum separately using two or more quantification methods in time and frequency domain, to establish the robustness and consistency of observations. Therefore, a comprehensive MRS quantitation method with a provision of choosing different quantitation algorithms in time and frequency-domain as well as in-built features like segmentation and quantitation is much needed in MRS data processing.
A variety of data processing packages are available that attempt to address specific limitations and are optimized for either proton (1H) or phosphorous (31P) MRS and Incorporate a specific baseline estimation method and peak fitting method. For instance, various models like LCModel and AQSES are designed specifically for 1H MRS data; MIDAS, JSIPRO, SIVIC are optimized specifically for multi-voxel magnetic resonance spectroscopy imaging (MRSI) data; GANNET is designed for batch analysis of GABA edited spectra, and proFIT is designed for J-resolved spectra only.
in order for realizing the research and clinical potential of MRS technique, a comprehensive MRS quantitation tool is required that offers various features such as customization to both clinical and research use, capacity to be optimized for multiple MR detectable nuclei e.g. 1H, 31P. It is capable of Implementing all MRS preprocessing, quantitation, and post-processing techniques in single platform thereby reducing dependencies on multiple software, and implements robust and reliable quantitation method.
While existing packages cater to some of these requirements, none are designed to provide a complete solution. Certain vendor specific packages are intended for clinical use and offer close to complete automation features with accessible visualization of results, but often restrict user customization of the processing workflows preventing optimization of parameters for particular metabolites. In contrast, certain tools developed for research laboratories aim for processing control and flexibility, but do not focus as much on data visualization, automation, and post processing features such as segmentation.
Further, most of the currently available MRS signal processing packages implement a single quantitation algorithm that relies on a given set of underlying assumptions regarding the nature of the baseline and the treatment of signals from macromolecules. As such, they have been individually designed to be used for a specific data types and restrict the user to either time- or frequency-domain analysis.
The principal object of this invention is to offer a comprehensive solution that meets all the aforementioned feature requirements. One key object of the invention is to provide an interface that enables both automatic signal processing and quantitation of metabolites for clinical use as well as customizable modules that enable processing parameters to be dynamically varied to achieve the optimal post-processed results for research setting.
Another object of the invention is that the method provides segmentation and absolute quantitation interface along with signal processing.
Yet another object of the invention is to enhance the accuracy of spectral fitting by use of an iterative baseline estimation and peak fitting method to attempt to minimize the residual error of the signal fit. The method offers the option of utilization of this iterative algorithm with various non-parametric baseline estimation and peak fitting methods.
A final object of the invention is that it can be utilized for both proton (1H) and phosphorous (31P) MRS data.
The present invention relates to a comprehensive method for versatile processing and improved quantitation accuracy of magnetic resonance spectra using an Iterative baseline estimation and fitting procedure that can be implemented with a variety of baseline estimation and peak fitting methods. It is an integrative MATLAB-based platform for visualization, preprocessing and quantitation of MRS data acquired using single voxel and MRSI and special type (e.g., MEGA-PRESS) acquisition methods. The method integrates both time- and frequency-domain signal processing methods on a single platform and implements non-parametric modeling of baseline through a variety of available methods. Apart from providing a comprehensive and flexible signal processing Interface, the novelty lies in its approach to signal quantitation.
According to an embodiment of the present invention, a method for interactive and automatic processing of signals in MRS is disclosed. The method comprises providing metabolite and image data Input to a magnetic resonance spectroscopy system, segregating a set of voxels containing data having a minimum signal-to-noise ratio based on given input. Further, the spectral parameters are pre-processed from the remaining voxels. The individual spectral peak area from each voxel is converted into metabolite concentration and a metabolite concentration map registered with a magnetic resonance image is obtained as output.
According to another embodiment of the invention, metabolite and image data input includes MRS imaging data, reference MRS imaging data, and MRI Image.
According to another embodiment of the invention, the method is optimized for both proton (1H) and phosphorous (31P) MRS data.
According to another embodiment of the invention, the MRS imaging data is from SV, MRSI, and MEGA-PRESS acquisition.
According to another embodiment of the invention, the reference MRS imaging data is unsuppressed water, an internal peak, or in-vitro phantom signal.
According to still another embodiment of the invention, the step of pre-processing from the remaining voxels comprises correcting for phase and frequency relating to reference MRS imaging data. Subsequently, water suppression is applied with Hankel Singular Value Decomposition (HSVD). In order to improve the spectral resolution, zero-filling is done and a filtered signal is obtained by temporal apodization method.
According to another embodiment of the invention, the metabolite concentration map is corrected for partial volume effect at each voxel and displayed.
According to still another embodiment of the invention, an iterative approach for baseline correction and peak fitting is used.
According to yet another embodiment of the invention, absolute quantitation of metabolite concentration and segmentation of MRI image is done.
The detailed description is described with reference to the accompanying figures. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. The same numbers are used throughout the drawings to reference like features and components.
The following discussion provides a brief, general description of a suitable environment in which various embodiments of the present disclosure can be implemented. The aspects and embodiments are described in the general context of computer executable mechanisms such as routines executed by a device. The embodiments described herein can be practiced with other system configurations, including Internet appliances, hand held devices, multi-processor systems, microprocessor based or programmable consumer electronics, network PCs, mini computers, mainframe computers and the like. The embodiments can be embodied in a special purpose computer or data processor that is specifically programmed configured or constructed to perform one or more of the computer executable mechanisms explained in detail below.
Exemplary embodiments now will be described with reference to the accompanying drawings. The disclosure may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey its scope to those skilled in the art. The terminology used in the detailed description of the particular exemplary embodiments illustrated in the accompanying drawings is not Intended to be limiting. In the drawings, like numbers refer to like elements.
The specification may refer to “an”, “one” or “some” embodiment(s) in several locations. This does not necessarily imply that each such reference is to the same embodiment(s), or that the feature only applies to a single embodiment. Single features of different embodiments may also be combined to provide other embodiments.
As used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless expressly stated otherwise. It will be further understood that the terms “includes”, “comprises”, “including” and/or “comprising” when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being “connected” or “coupled” to another element, it can be directly connected or coupled to the other element or intervening elements may be present. Furthermore, “connected” or “coupled” as used herein may include wirelessly connected or coupled. As used herein, the term “and/or” Includes any and all combinations and arrangements of one or more of the associated listed items.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The figures depict a simplified structure only showing some elements and functional entities, all being logical units whose implementation may differ from what is shown. The connections shown are logical connections; the actual physical connections may be different. It is apparent to a person skilled in the art that the structure may also comprise other functions and structures. It should be appreciated that the functions, structures, elements and the protocols used in communication are irrelevant to the present disclosure. Therefore, they need not be discussed in further detail here.
In addition, all logical units described and depicted in the figures include the software and/or hardware components required for the unit to function. Further, each unit may comprise within itself one or more components, which are implicitly understood. These components may be operatively coupled to each other and be configured to communicate with each other to perform the function of the said unit.
The present invention is directed to an integrative platform for visualization, preprocessing and quantitation of MRS data acquired using single voxel and multi-voxel MRSI and MEGA-PRESS acquisition methods. The method is MATLAB-based, integrates both time-domain and frequency-domain signal processing methods on a single platform and implements non-parametric modeling of baseline through a variety of known techniques. Apart from providing a unique comprehensive and flexible signal processing interface, novelty of the method lies in its approach to signal quantitation.
Signal quantitation through various time-domain and frequency-domain methods has been achieved through application of an iterative baseline extraction and peak fitting method which allows for maximum accuracy and reliability of signal quantitation. The following terms have been used in abbreviated form in the subsequent paragraphs, SV: Single voxel; MV: Multi voxel i.e. multiple spectra; MRSI: Magnetic Resonance Spectroscopic Imaging; MRI: Magnetic Resonance Imaging; MRS: Magnetic Resonance Spectroscopy; MEGA-PRESS: MEshcher-GArwood Point-RESolved Spectroscopy; PSF: Point Spread Function; CSF: Cerebero-spinal fluid; LCModel: Unear combination model; AQSES: automatic quantitation of short echo time spectra; MIDAS: Metabolite Imaging and Data Analysis System; JSIPRO: Java Spectroscopic imaging Processing; SIVIC: Spectroscopic Imaging, Visualization, and Computing; GANNET: GABA-MRS Analysis Tool; proFit: Prior-Knowledge Fitting; TDFD: time-domain frequency-domain; FD: Frequency Domain; TD: Time Domain; NLLS: Non-linear least square; SSA: Singular Spectrum Analysis; JMRUI: Java-based graphical user interface for the Magnetic Resonance User interface; TARQUIN: Totally Automatic Robust Quantitation in NMR; 3DICSI: 3D Interactive Chemical Shift Imaging; HSVD: Hankel Singular Value Decomposition; VARPRO: Variable Projection.
This invention is useful in both clinical and research settings. Specifically, in vivo MRS is beneficial for non-invasive acquisition of biochemical information of various tissues in the human body related to many diseases such as cancer, epilepsy, and Parkinson's disease and as such this invention has tremendous applicative value.
The invention provides following key features over the prior art. The method is optimized for a wide variety of data that includes SV, MRSI and edited spectra such as MEGA-PRESS in contrast to most of the existing packages, which are optimized for particular type of data, for instance, LCModel and AQSES are designed specifically for 1H MRS data; MIDAS, JSIPRO, SIVIC are optimized specifically for MRSI data; GANNET is designed for batch analysis of GABA edited spectra, and proFIT is designed for J-resolved spectra only. It also provides a variety of peak fitting methods under one integrated platform. The user has an advantage of spectral fitting In time-domain, frequency-domain, or mixed domain.
Further, the method provided by the present invention implements an iterative baseline estimation and peak fitting algorithm with all methods leading to better quantitation accuracy. In prior art, this algorithm has only been implemented for TDFD fitting in combination with baseline estimation using wavelet shrinkage method and has been shown to yield more accurate results as compared to conventional method of modelling baseline and metabolite signal together to form non-linear least square (NLLS) cost function. However, the present method extends and improves upon this invention by implementing iterative estimation procedure with VARPRO, time-domain NLLS, and FDpeak fitting methods, and in combination with various baseline estimation methods, i.e. Singular Spectrum Analysis (SSA), spline, and wavelet decomposition.
The invention also provides various post-processing features such as absolute quantitation, segmentation, and metabolite map generation, where several existing packages lack e.g. JMRUI, LCModel, TARQUIN, 3DICSI, AQSES.
The peak area values from all voxels are then converted to metabolite concentrations using water or a similar metabolite as an internal reference. In order to correct for partial volume effects, MRI image is segmented and percent cerebrospinal fluid (CSF) is calculated within each MRSI voxel by convolving CSF mask image with MRSI point spread function. Subsequently, metabolite concentration map Is corrected for partial volume effects at each voxel and displayed as an overlay onto anatomical MRI images. The final output of the pipeline is metabolite concentration map co-registered with the MRI image as well as a tabulated listing of voxel-wise metabolite concentrations in institutional units along with estimated spectral parameter values.
Abbreviations used in
Following references have been used: a 2D JPRESS; b MEGA-PRESS; ** Command line; *** Segmentation only.
The present Invention provides an intuitive user interface that enables three different approaches for MRS data processing. Primarily, it permits the expert users to specify processing parameters at each stage and test the performance of different quantitation algorithms on their data in order to choose the best suited processing pipeline. Second for non-expert users, it provides customizable processing modules optimized separately for SV, MRSI, and MEGA-PRESS that allow complete quantitation with just a few clicks. Third, a batch processing interface is provided for automated analysis of large datasets without user interaction. The batch processing interface lets the user load and customize any default processing pipeline or define a new processing pipeline and run it for bulk data.
This invention ensures reliable and robust spectral quantitation by using an Iterative baseline estimation and peak fitting approach, whereby the process of baseline estimation and parametric peak fitting is alternated until the improvement in fitting residual in consecutive iterations is lower than a pre-defined threshold level. Further, it offers the option of utilization of this iterative process with various baseline estimation methods along with commonly used peak fitting methods in time- and frequency-domain. The robustness and accuracy of the invention in metabolite quantitation is well-matched to the established packages for MRS quantitation i.e. LCModel and JMRUI.
The abovementioned description of the embodiments of the invention has been presented for the purposes of illustration and description. It is not Intended to be exhaustive or to limit the invention to the precise form disclosed. Many modifications and variations are possible in light of the above teaching. It is Intended that the scope of the invention be limited not by this detailed description, but rather by the claims appended hereto.
As will be appreciated by one skilled in the art, the present invention may be embodied as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, a software embodiment or an embodiment combining software and hardware aspects all generally referred to herein as a “circuit” or “module.” Furthermore, the present invention may take the form of a computer program product on a computer-usable storage medium having computer-usable program code embodied in the medium.
It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program Instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.
Instructions may also be loaded onto a computer or other programmable data processing apparatus like a scanner/check scanner to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.
In the drawings and specification, there have been disclosed exemplary embodiments of the invention. Although specific terms are employed, they are used in a generic and descriptive sense only and not for purposes of limitation, the scope of the invention being defined by the following claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 201611001944 | Jan 2016 | IN | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/IB2016/054978 | 8/19/2016 | WO | 00 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2017/125800 | 7/27/2017 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 7966053 | Balchandani | Jun 2011 | B2 |
| 9234953 | Labadie | Jan 2016 | B2 |
| 20080272782 | Lin | Nov 2008 | A1 |
| 20090091322 | Posse | Apr 2009 | A1 |
| 20140085304 | Kim et al. | Mar 2014 | A1 |
| 20140191755 | Bauer | Jul 2014 | A1 |
| Number | Date | Country | |
|---|---|---|---|
| 20210190891 A1 | Jun 2021 | US |
