Patent Application
20050249398
1. Field of the Invention
The present invention relates generally to the field of processor-based imaging, and, more particularly, to a rapid and robust 3D/3D registration method.
2. Description of the Related Art
Registration of two volumetric data sets (i.e., a fixed volume and a moving volume) generally involves three steps: (1) computing a similarity measure and/or a difference measure quantifying a metric for comparing the volumes; (2) performing an optimization scheme, which searches through the parameter space (e.g., a six-dimensional rigid body motion) to maximize the similarity measure or to minimize the difference measure; and (3) performing a volume warping method, which applies the latest computed set of parameters to the moving volume to transform the moving volume closer the fixed volume.
In multi-modal volume registration, a key issue is involves choosing a well-behaved similarity measure that can robustly characterize a metric for the volumes. Additionally, the time for computing the similarity measure and applying the similarity measure to the moving volume is of substantial importance. Current methods for computing and applying a similarity measure suggest that, for a typical three-dimensional volume data sets, the computation time is in order of minutes. Most of this time is spent on computing the similarity measure over the whole volume set, and transforming (i.e., changing the coordinate frame) of the moving volume.
Some researchers have suggested random sampling of the volume data sets and performing computations only based on these random samples to decrease the computation load. Other researchers have proposed a hybrid technique, which selects a set of high-interest points (i.e., landmarks) within the volume and attempts to perform registration based on only the set of high-interest points. In both cases, a primary goal is to reduce the computational complexity of a full-blown volumetric matching by reducing the space in which the registration is performed, while preserving the accuracy and robustness as much as possible. For many practical applications, especially for interventional scenarios, the time used for performing a registration can be quite crucial.
In one aspect of the present invention, a method is provided. The method includes the steps of generating a two-dimensional signature for each of a plurality of three-dimensional volumes; and performing registration on the two-dimensional signatures.
In another aspect of the present invention, a program storage device readable by a machine, tangibly embodying a program of instructions executable on the machine to perform method steps, is provided. The method includes the steps of generating a two-dimensional signature for each of a plurality of three-dimensional volumes; and performing registration on the two-dimensional signatures
In yet another aspect of the present invention, a program storage device readable by a machine, tangibly embodying a program of instructions executable on the machine to perform method steps, is provided. The method includes the steps of generating a lower-dimension signature for each of a plurality of data sets of a given dimension; and performing registration on the lower-dimension signatures.
The invention may be understood by reference to the following description taken in conjunction with the accompanying drawings, in which like reference numerals identify like elements, and in which:
Illustrative embodiments of the invention are described below. In the interest of clarity, not all features of an actual implementation are described in this specification. It will of course be appreciated that in the development of any such actual embodiment, numerous implementation-specific decisions must be made to achieve the developers' specific goals, such as compliance with system-related and business-related constraints, which will vary from one implementation to another. Moreover, it will be appreciated that such a development effort might be complex and time-consuming, but would nevertheless be a routine undertaking for those of ordinary skill in the art having the benefit of this disclosure.
While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof have been shown by way of example in the drawings and are herein described in detail. It should be understood, however, that the description herein of specific embodiments is not intended to limit the invention to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined by the appended claims.
It is to be understood that the systems and methods described herein may be implemented in various forms of hardware, software, firmware, special purpose processors, or a combination thereof. In particular, at least a portion of the present invention is preferably implemented as an application comprising program instructions that are tangibly embodied on one or more program storage devices (e.g., hard disk, magnetic floppy disk, RAM, ROM, CD ROM, etc.) and executable by any device or machine comprising suitable architecture, such as a general purpose digital computer having a processor, memory, and input/output interfaces. It is to be further understood that, because some of the constituent system components and process steps depicted in the accompanying Figures are preferably implemented in software, the connections between system modules (or the logic flow of method steps) may differ depending upon the manner in which the present invention is programmed. Given the teachings herein, one of ordinary skill in the related art will be able to contemplate these and similar implementations of the present invention.
We propose a novel method for rigidly registering two multi-modal or mono-modal volumetric data sets. A primarily bottleneck in the computational effectiveness of traditional volumetric registration algorithms is volume transformation and similarity measure estimation over the volumes. In the approach proposed herein, we downgrade the three-dimensional problem to a two-dimensional one. We consider two-dimensional signatures of the volumes (e.g., multiple maximum intensity projections (“MIPs”) or digitally reconstructed radiographs (“DRRs”)) as the working space where the similarity measures are computed. Therefore, we achieve much greater computational efficiency by reducing the dimension of the working space.
Multiple two-dimensional signatures can robustly represent a volume, depending on the way the signatures are generated. An easy way to conceptualize the idea is to understand the motion of an object by looking only at three perpendicular shadows of the object. By also considering see-through shadows (i.e., DRRs), we can achieve even greater robustness in our registration method. Referring now to
We propose using multiple two dimensional signatures of three-dimensional volumes to perform registration. The number, orientation, and type of the two dimensional signature may vary depending on application. Exemplary two-dimensional signatures include, but are not limited to, planar projective (i.e., DRR), maximum intensity, and multi-planar reconstruction. The two-dimensional signature may also be non-planar (i.e., two-dimensional manifolds).
Two-dimensional signatures may have either one value or multiple values at each pixel position (x, y). Having multiple values at each pixel position encodes greater information about the represented volume than having only one value. In one exemplary embodiment of having multiple values at each pixel location, a set of two-layer projection images have minimum and maximum voxel intensities of the parallel rays intersecting the volume.
Transformation parameters, the use of which the two-dimensional projections are transformed, is inherently smaller than that of three-dimensional volumes. For example, if the volume is transformed about an axis, the orthographic projection along the same axis remains unchanged. Furthermore, the effect of two rotational transformations about the axes in the plane perpendicular to the projection axis is minimized and the small rotations can be ignored. Therefore, the number of parameters to be changed and optimized for alignment of projection images is three. Collective registration of the projection images should cover the whole space of rigid transformation comprising six parameters (i.e., three rotations and three translations). Collective registration can be done using an iterative method during which a subset of parameters is optimized. The result here is the optimization process becomes more robust and the capture range increases solely because the number of parameters decrease.
Steering of the parallel rays in the space of the search parameters (for example, in the case of the rigid body motion, three translations and three rotations) to generate moving volume two-dimensional signatures can be implemented using graphics hardware accelerated technique as described in Guering et al., Flexible DRR Generation using Programmable Computer Hardware, Ser. No. ______ and filed on ______, and assigned to the assignee of the present application. The full disclosure of the above-referenced patent application is hereby incorporated herein by reference. Steering of the parallel rays in the space of the search parameters to generate moving volume two-dimensional signatures can be implemented using the standard rapid orthographic volume rendering technique (e.g., sheer-warp method.)
By segmenting the two-dimensional signatures, we can isolate the area of the interest with in the volume and perform selective matching on that area only. Piecewise rigid matching can be implemented by segmenting the two-dimensional signatures into multiple areas, where each area will then have its own set of parameters to perform matching/registration. Parametric non-rigid matching of the two volumetric data sets can also be performed. As the number of parameters modeling the non-rigid motion grows, a greater number of two-dimensional signatures is generally required.
Referring now to
The particular embodiments disclosed above are illustrative only, as the invention may be modified and practiced in different but equivalent manners apparent to those skilled in the art having the benefit of the teachings herein. Furthermore, no limitations are intended to the details of construction or design herein shown, other than as described in the claims below. It is therefore evident that the particular embodiments disclosed above may be altered or modified and all such variations are considered within the scope and spirit of the invention. Accordingly, the protection sought herein is as set forth in the claims below.
This application claims priority to U.S. Provisional Application No. 60/564,145, which was filed on Apr. 21, 2004, and which is fully incorporated herein by reference.
| Number | Date | Country | |
|---|---|---|---|
| 60564145 | Apr 2004 | US |
