WO2007139697A2 - Apparatus and method for rendering for display forward-looking image data - Google Patents

Abstract

An apparatus and method for rendering for display forward-looking image data. The apparatus includes rendering for display image data from a forward-looking conical section of tissue collected by an image data collection device, comprising a data input unit configured to receive the image data; and an image data processing unit configured to rasterize the image data onto a projected three-dimensional geometric model to produce a rasterized image, thereby giving the appearance of three-dimensions when displayed The method includes a method of rendering for display image data from a forward-looking conical section of tissue collected by an image data collecting device, comprising receiving the image data; and rasterizing the image data onto a projected three-dimensional geometric model to produce a rasterized image, thereby giving the appearance of three dimensions when displayed.

Description

APPARATUS AND METHOD FOR RENDERING FOR DISPLAY FORWARD- LOOKING IMAGE DATA

This application claims priority to the filing date of U.S. Application Serial No.

11/437,687, filed May 22, 2006, which is hereby incorporated by reference.

BACKGROUND OF THE INVENTION

1. Field of the Invention

[0001] The invention is directed to an apparatus and method for rendering for

display forward-looking image data.

2. Background of the Related Art

[0002] In the field of Interventional Medicine, catheters, laparoscopes, and

endoscopes are frequently used to create images of tissues. These images can be used, for

example, to diagnose and treat disease. Further, it may be advantageous in certain cases to

employ medical images to guide and/ or monitor progress during therapeutic treatment of a

medical condition. Ultrasonic imaging and Optical Coherence Tomography (OCT) are

imaging techniques that are well suited for such applications. For example, they can produce

tomographic images in real-time from the end of, for example, a catheter, endoscope, or

laparoscope. Images from these modalities tend to be displayed in planar tomographic

formats. For example, when a mechanically or manually steered device is employed to rotate

a transducer, the images are usually displayed as circular images. Although most medical

Ultrasound and OCT images are scanned by mechanical rotation of the transducer, they can

also be electronically scanned by steering the ultrasonic or optical beam from the end of the

catheter or other device. New devices now incorporate manually rotated Intravascular

UltraSonic (TVUS) imaging. [0003] When using, for example, a catheter, endoscope, or laparoscope in the

confines of a vessel or body cavity for guiding and monitoring therapies, one frequently

would like to visualize what lies ahead of the catheter, endoscope, or laparoscope tip. This is

especially true when the therapeutic device is located on a distal most end of the catheter,

endoscope, or laparoscope. Forward visualization can also be beneficial when one wishes to

collect biopsy specimens from specific tissues in a minimally invasive fashion. Mechanically

steered ultrasound and OCT devices are excellent at making images that are substantially

perpendicular to an axis of the catheter, endoscope or laparoscope. However, for the

purpose of guiding therapy, in some applications, it is preferable to have, for example, a

transducer or optical fiber sweep out a conical forward-looking surface that depicts tissue

that lies distal to a tip of the catheter, endoscope, or laparoscope.

[0004] The field of Interventional Medicine arose to apply less invasive

technologies to the treatment of disease. For diagnostic and therapeutic guidance tools,

Cardiovascular Interventionalists had fluoroscopy systems that allowed them to image the

blood pool by injecting a radio-opaque contrast dye into the blood stream and then use a

real time x-ray imaging technique to watch as that contrast dye passed through the vascular

tree. A major limitation with this technology was that the fluoroscope did not, in fact, image

the vessel tissue but rather the blood pool inside the vessel lumen. In an attempt to obtain

an image of the vessel wall and not just the blood pool, ultrasonic imaging transducers were

mounted at the distal end of catheters and positioned in the arteries. These transducers were

typically mounted on catheters that were approximately 1 mm in diameter. The ultrasonic transducers typically had apertures of about 65 mm - .75 mm. This approach did indeed

allow the visualization of tissues that comprise the artery walls.

[0005] Optical Coherence Tomography (OCT), as described in U.S. Patent

No. 5,321,501, which is hereby incorporated by reference, is similar to ultrasonic imaging but

uses reflected infrared light to create images instead of echoes from sound waves. As with

ultrasound technology, it can be incorporated into an ~ 1 mm diameter catheter and make

images of the vessel wall in addition to the vessel lumen.

[0006] Although some mechanically steered IVUS and catheter-based OCT

imaging systems create images that are a few degrees off of the perpendicular to the axis of

the catheter, these systems are constructed in this way so as to minimize the reflections from

the catheter sheath. The size of the reflection is greatest when it encounters the catheter

sheath exactly perpendicular to the direction of propagation. Typically, these catheters image

~ +5 to -5 degrees from the perpendicular to the catheter axis. Since these images are nearly

planar, they are displayed as two-dimensional images on a video monitor. However, in order

to see the tissue distal to the tip of the catheter a greater angle from perpendicular must be

employed. This results in a conical forward-looking surface that is swept out in front of the

transducer or optical fiber. U.S. Patent Application No. 11/053,141, which is hereby

incorporated by reference, teaches a number of methods for displaying such images using

hue, saturation, and intensity of image data to give the impression of a non-planar image.

[0007] An excellent review of prior art imaging systems can be found in Bom

et. al. "Early and Recent Intraluminal Ultrasound Devices", International Journal of Cardiac

Imaging, Vol. 4 1989, pp. 79-88 (hereinafter "Bom article"), which is hereby incorporated by reference. The Bom article discusses different approaches to making intraluminal images that

have been tried over the years. With respect to IVUS imaging specifically, many patents

exist. For example, US. Patent No. 4,794,931 to Yock, which is hereby incorporated by

reference, teaches both an intravascular ultrasonic imaging device, and a combined imaging

and therapeutic device, and its use for guiding therapy. The catheters taught create images

that are planar in nature and are intended to help guide the removal of atherosclerotic tissue

by means of a circular cutting element. Continuations of this patent cover other forms of

therapy guidance.

[0008] Commercially available IVUS products today create two-dimensional

images that are approximately perpendicular to a longitudinal axis of the catheter. With the

addition of a motorized mechanical pullback sled as discussed in U.S. Patent No. 5,361,768,

which is hereby incorporated by reference, the catheter can collect these planar images as the

catheter is translated (withdrawn) from the vessel. The catheter pullback may also be done

by hand This, in effect, results in a spiral sampling of three-dimensional space that can then

be displayed in various ways. Once the data has been collected, the three-dimensional

volume can be re-sectioned along the longitudinal axis into what is often referred to as a

Longitudinal Mode (L-Mode) image, which is just a longitudinal slice through the vessel. The

primary purpose, however, of these three-dimensional IVUS images is to measure the total

tissue volume of the atherosclerotic plaque. The segmentation of the imaging data into

various tissue and plaque components may now be done automatically by a computer

algorithm, thereby saving the operator from the tedious task of tracing the borders of the

various tissue components, as discussed in, "Preintervention Lesion Remodeling Affects Operative Mechanisms of Balloon Optimized Directional Coronary Atherectomy

Procedures: A Volumetric Study With Three Dimensional Intravascular Ultrasound," Heart

2000; 83, 192-197, which is hereby incorporated by reference.

[0009] In some applications, however, it is advantageous to view the tissue in

front of, or distal to, the catheter tip. Two-dimensional forward-looking IVUS imaging has

been documented by a number of authors. In Evans et aL, "Arterial Imaging with a New

Forward-Viewing Intravascular Ultrasound Catheter, I, Initial Studies," Circulation, vol. 89,

No. 2, pp. 712-717, Feb. 1994, which is hereby incorporated by reference, a mechanically

wobbled transducer sweeps out a forward-looking, tomographic, two-dimensional image

which is displayed using scan conversion techniques common in the ultrasound industry.

[0010] A companion paper authored by Ng and titled, "Arterial Imaging With

a New Forward-Viewing Intravascular Ultrasound Catheter, H Three-Dimensional

Reconstruction and Display of Data," which is hereby incorporated by reference, addresses

the display of the image information. By using a forward-looking transducer and rotating it

on the catheter axis, a full three-dimensional conical data set is obtained. From the data set,

C-Scan (two-dimensional cross-sections parallel to the transducer face) or vessel cross-

sectional images are then reconstructed. After manually tracing the vessel wall borders, and

spatially orienting the component two-dimensional images, they rendered a three-

dimensional surface perspective of the vesseL

[0011] Another paper titled, "Three-Dimensional Forward-Viewing

Intravascular Ultrasound Imaging of Human in Vitro" by Gatzoulis et aL, which is hereby

incorporated by reference, describes a similar mechanically wobbled transducer at the end of a catheter using post processing software to display sections in both the complete three-

dimensional rendered volumes, B-Scan (two-dimensional cross-sections perpendicular to the

face of the transducer) format sections, and C-scan format sections.

[0012] Further, U.S. Patent No. 5,651,366 to Liang et aL, which is hereby

incorporated by reference, patented a device for making forward-looking images distal to the

tip of an intravascular catheter. Not only is the diagnosis of disease envisioned, but the

guidance of therapy as welL The Liang catheter employs a mirror in a distal tip of the

catheter that reflects an ultrasound beam from a transducer that is in relative motion to the

mirror. ^"With the specified geometry, an approximately sector format is created to image the

tissue in front of the catheter.

[0013] In a 1991 paper titled, "Investigation of a Forward-Looking IVUS

Imaging Transducer" by Lee and Benekeser, which is hereby incorporated by reference, the

concept of a forward-looking intravascular imaging transducer is described. In this case, the

transmitted sound is reflected off a conical mirror to create a conical scan pattern in front of

the catheter. No discussion of how one should accurately display the three-dimensional

images is included in the manuscript. Presumably, the conical section would be displayed as a

conventional IVUS two-dimensional image.

[0014] US. Patent No. 6,066,096 to Smith et aL, which is hereby incorporated

by reference, teaches an electronically steered two-dimensional phased array transducer on

the end of a catheter that can incorporate both diagnostic and therapeutic devices. This

transducer, in one embodiment, is a forward-looking imaging system. This patent cites prior art that employs C-Scan displays and B-Scan displays to present any plane in the field of

view.

[0015] US. Patent Nos. 6,457,365 and 6,780,157, which are hereby

incorporated by reference, teach a combination side-looking and forward-looking catheter

based imaging system. These patents propose two techniques for displaying the forward-

looking three-dimensional image data: B-Scans and C-Scans.

[0016] Optical Coherence Tomography (OCI) is analogous to ultrasonic

imaging but uses light waves instead of sound waves to make images. Since the speed of light

is much faster than the speed of sound, simple electronic gating techniques are not adequate

for separating the reflections that emanate from the tissues at varying depths. Instead, an

interferometer is used, but the rest of the signal processing is much the same as in ultrasonic

imaging. Both OCT and IVUS can provide images from the distal tip of a catheter and both

are able to guide interventional therapeutic procedures. The display options for intravascular

ultrasonic imaging can all be readily applied to intravascular OCT imaging by one of ordinary

skill in the art.

[0017] Three-dimensional rendering of both medical and non-medical images

is common. Three-dimensional image rendering is a process of converting the collected

image data in a manner that is amenable to human understanding, while preserving the

integrity, intensity, and geometric accuracy of the image information.

[0018] In medical imaging, displays tend to fall into two general categories.

The first is a surface rendering technique, where an organ surface, or a tissue surface, is

detected from the image data and the surface of the tissue is displayed rather than the tissue itself. In these surface renderings, in addition to creating a three-dimensional perspective,

one frequently adds lighting, shading, and/or texture to enhance the perception that one is

observing a three-dimensional object.

[0019] A second technique involves simply selecting any two-dimensional slice

from the three-dimensional volume and displaying that image plane on a monitor. In the

case of ultrasonic imaging, these two-dimensional planes are referred to as B-Scans if they

are in the plane that the sound waves traveled or C-Scans if they are perpendicular to the

plane that the sound waves originally traveled. For other medical imaging modalities, they are

simply referred to as cut planes or re-sampled planes.

[0020] An alternative three-dimensional display technique that allows the

observer to view a three-dimensional object from the perspective of inside the object is

taught in U.S. Patent No. 5,606,454 to Williams et al., which is hereby incorporated by

reference. This approach may have particular appeal for intraluminal imaging, where the

structure to be displayed is in fact hollow.

[0021] The above references are incorporated by reference herein where

appropriate for appropriate teachings of additional or alternative details, features and/or

technical background.

SUMMARY OF THE INVENTION

[0022] An object of the invention is to solve at least the above problems

and/or disadvantages and to provide at least the advantages described hereinafter. [0023] The invention is directed to an apparatus and method for rendering for

display forward-looking image data.

[0024] Additional advantages, objects, and features of the invention will be set

forth in part in the description which follows and in part will become apparent to those

having ordinary skill in the art upon examination of the following or may be learned from

practice of the invention. The objects and advantages of the invention may be realized and

attained as particularly pointed out in the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The patent or application file contains at least one drawing executed in

color. Copies of this patent or patent application publication with color drawing(s) will be

provided by the Patent Office upon request and payment of the necessary fee.

[0026] The invention will be described in detail with reference to the

following drawings in which like reference numerals refer to like elements wherein:

[0027] Figure 1 is a schematic diagram of a IVUS imaging system according to

an embodiment of the invention;

[0028] Figure 2A is a block diagram of a rendering apparatus according to an

embodiment of the invention;

[0029] Figure 2B is a flow chart of a rendering method according to an

embodiment of the invention;

[0030] Figure 3 is a perspective view of a catheter and transducer assembly

along with a depiction of a forward-scanning cone; [0031] Figure 4A is an oblique parallel projection of an exemplary non-

nutated, conical polyhedron;

[0032] Figure 4B is an oblique parallel projection of another exemplary non-

nutated, conical polyhedron;

[0033] Figure 5 illustrates a perspective projection of an exemplary nutated

version of the conical polyhedron of Figure 4A;

[0034] Figure 6 illustrates an exemplary data table according to the invention;

and

[0035] Figures 7A-7F are still frames taken from an exemplary display utilizing

an apparatus and method according to the embodiments of the invention.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

[0036] The invention is directed to rendering for display a conical forward-

looking image on a two-dimensional video monitor in a manner that conveys to an operator

the three-dimensional nature of the information. The invention provides a novel imaging

modality that allows visualization of a forward-looking section of, for example, a vessel for

the purpose of guiding a catheter or other based interventions. The invention may not only

reduce the number of emergency by-pass surgeries necessary to repair artery perforations

but, more importantly, may facilitate minimally invasive treatment of a higher percentage of

coronary artery lesions, thereby avoiding a far more expensive coronary bypass operation.

[0037] Further, certain embodiments of the invention are directed to an

apparatus and a method that geometrically accurately display conical forward-looking image data so that an observer can immediately assimilate the information being presented and

make more efficient and effective diagnostic and therapeutic decisions. The display is neither

a B-Scan format nor a C-Scan format nor even a three-dimensional surface rendering. The

image data is displayed as though it was on a surface of a three-dimensional geometric

model, in one embodiment a cone. As the image is re-scanned, new image data may replace

the previous image data in the correct geometric location on the surface of the three-

dimensional geometric model, in one embodiment the conical surface.

[0038] A nutating motion may be employed to enhance visualization of the

image data by an observer. In one embodiment where the three-dimensional geometric

model is a cone, the nutating motion would give an observer the impression that the display

is, in fact, a conical surface. lighting and/or shading may also be used to further enhance the

impression of a three-dimensional object. The nutation may be at a fixed rotation rate or it

may be matched to a rotation of the image device so that when image data is updated the

nutation angle is, for example, ~ 180 degrees out of phase with the newest image data.

Alternatively, the new image data may be exactly in phase with the nutation angle.

[0039] In one embodiment in which the three-dimensional geometric model is

a cone, an angle between an axis of the cone and a line perpendicular to the image display is

made sufficiently small with respect to the cone angle to prevent the nutation of the entire

cone from blocking the observer's view of the back side of the cone. This has the advantage

of avoiding the computational burden of incorporating a hidden line removal algorithm.

[0040] In the case where the apparatus and/or method according to

embodiments of the invention is used to assist in positioning a diagnostic and/or therapeutic tool or device, a representative marking may be superimposed on the surface of the three-

dimensional geometric model to inform the observer where the tool or device will be active.

[0041] Further, certain embodiments of the invention are directed to a

method and apparatus for rendering for display image data from forward-looking intracavity

transducers that acquire the image data in a three-dimensional conical scan format- The data

may be displayed in real time and may be nutated about an axis of the device's rotation to

provide visual cues to the three-dimensional conical nature of the image. A nutation angle

can be controlled either automatically based on an angle of the newest incoming data,

automatically based on a periodic time-varying sweep of the nutation angle, or manually

based on user input, for example, from a pointing device.

[0042] A three-dimensional geometric model of the conical scan format may

be constructed, then transformed into two-dimensional space via rotation, translation,

scaling, and projection. A nutation angle may be used to create a composite rotation

sequence, which may be applied to the three-dimensional conical model and results in the

longitudinal axis of the cone being tilted in the desired direction. The image data may be

scan-converted (Le. rasterized) in accordance with the transformed model The resulting

scan-converted image is a two-dimensional representation of a three-dimensional conical

scan format.

[0043] Although a primary visual cue that the image is three-dimensional is the

nutation angle which is varied, additional cues maybe added. For example, lighting, shading,

and/or texture may be added to provide additional information pertaining to the orientation

of the surfaces of the cone. For example, surfaces whose normals point roughly in the direction of the viewer will "reflect" more light and appear brighter than surfaces whose

normals point away from the viewer.

[0044] Additionally, certain embodiments of the invention include a method

and apparatus for rendering for display image data from a forward-looking intracaviry

transducer using rendering techniques that provide visual cues to a three-dimensional conical

scan format of the acquired image data. Forward-looking transducers are useful in

visualizing what lies ahead of, for example, a catheter, endoscope, or laparoscope, which is

especially important when a therapeutic modality is located at a distal-most end of such a

device. The visual cues may include, but are not limited to, nutating an axis of transducer

rotation, either automatically or manually, such that the conical scan format of the image is

apparent. Using this method and apparatus to render such an image data set avoids the

geometric distortion that occurs when data from a forward-looking transducer is rendered

with a traditional planar scan-conversion algorithm.

[0045] The invention relates generally to the field of medical devices and more

specifically to the field of interventional medicine. The invention provides for an accurate

depiction of three-dimensional spatial orientation of all points in an image, thereby allowing

for more accurate diagnosis and treatment of disease. One specific application that will be

used for the purpose of explaining the invention is Interventional Cardiology, however, the

invention may be utilized for any application for which rendering for display collected image

data to assist a user in visualizing the image data would be useful or advantageous.

[0046] The imaging modality frequently chosen for Interventional Cardiology

is IVUS; however, the image may be utilized with other image data collection modalities or techniques. While IVUS is a well known diagnostic imaging technique, it can also serve as a

therapeutic guidance tool. For example, expansion of an angioplasty balloon or deployment

of a stent can be visualized in real-time to improve clinical outcome. Other therapeutic

techniques and devices would also benefit from an operator's ability to visualize therapy as it

progresses. In some cases, therapeutic action takes place at the very distal tip of a catheter.

Such a situation can occur when using, for example, laser ablation, radio-frequency ablation,

biopsy forceps, and biopsy needles. To optimally guide these therapies and diagnostic

procedures it is beneficial to have an accurate depiction of the tissue involved which can lie

distal to the catheter tip.

[0047] Figure 1 is a schematic diagram of an IVUS imaging system according

to an embodiment of the invention. The system 1 includes a catheter 10. The distal end or

tip 28 is adapted to be inserted into a patient and is constructed to be navigable through the

patient's vasculature to advance the distal end 28 to an area or site of interest. The distal tip

28 of the catheter 10 carrier an ultrasound transducer (not shown). Further, the distal end

28 may also carry an ablation electrode (not shown) adapted to ablate an obstructed portion

of a vessel or other body cavity. The proximal end of the catheter 10 is designed to remain

outside of the patient where it can be associated with an angle encoder 16 and can be

manipulated by an operator, such as an interventionalist or physician.

[0048] The system 1 further includes an electronics module 18 that includes

circuitry and software for generating signals for operating the system 1 and for receiving and

processing signals from resulting ultrasound echoes, as well as for generating an RF ablation

signal if an ablation electrode is included in the system. A central processing unit 20 constructs images from the received ultrasound signals and displays the images on a monitor

21. The images may be generated on demand and may be refreshed in response to operator

rotation of the catheter 10. The images may be caused to fade after a predetermined time as

a reminder to an operator to refresh the image by rotating the catheter 10. The central

processing unit 20 may comprise, for example, a laptop or desktop computer or a dedicated

embedded processor. Cables 22, 24 may be connected between the angle encoder 16 and

the electronics module 18. In one embodiment, the cable 22 carries incremental angle

information that is sensed by the angle encoder 16 and cable 24 provides power and ground

Separate cables may run from the catheter 10 to the electronics module 18 and carry

ultrasound signals and also RF energy if an ablation electrode is included in the system. In

an alternate arrangement (not shown), transducer and RF cables from the catheter 10 may

plug into a connector integrated into the angle encoder 16 and then, after pre-amplϋying the

ultrasound signals, pass the signals through a second connector on the angle encoder 16 to

the electronics module 18. This alternate arrangement allows for a shorter catheter cable

and, potentially, reduces environmental noise pick-up.

[0049] The catheter 10 may be rotated and manipulated entirely under manual

control of the operator. Similarly, in the case where an ablation electrode is included in the

system 1, initiation of the ablation pulse may be determined by the operator independently

of any direct connection with the catheter or the system for sensing catheter rotation. It

should be understood that reference to "manual" with respect to control over the

application of ablation energy includes any arrangement by which the operator, based on

judgment as to the proper location of the ablation electrode, initiates the ablation sequence- Thus, "manual" operation may include a variety of arrangements, including, mechanically

controlled switches, for example, a foot switch, or a voice operated control or other means

by which the operator can trigger an ablation cycle, for example, by manipulation of pedal

19.

[0050] Figures 2A and 2B disclose a block diagram of a rendering apparatus

and a flow chart of a rendering method, respectively, according to embodiments of the

invention. As shown in Figure 2A, the rendering apparatus 140 includes a data input unit

180 in which image data is received or input from an imaging system, such as the IVUS

system shown in Figure 1. The data input unit 180 may organize the image data into a data

table, such as that shown in Figure 6. The data input unit 180 may also receive parameters

set by a user or set and/or calculated by the imaging system. Such parameters may include

the forward scanning angle, the radius of the base, the observer distance, the magnitude of

nutation, the samples per line, the lines per frame.

[0051] The rendering apparatus 140 further includes a three-dimensional

geometric model initialization unit 145, a projection matrix initialization unit 155, and a

nutation matrix initialization unit 175. The three-dimensional geometric model initialization

unit 145 outputs a three-dimensional geometric model, in one embodiment, a conical

polyhedron, while the projection matrix initialization unit 155 and the nutation matrix

initialization unit 175 output a projection matrix and a nutation matrix, respectively. The

present application discloses using a conical polyhedron as the three-dimensional geometric

model, for example, however, other three-dimensional geometric models may be appropriate

based on the particular application. The rendering apparatus 140 further includes a three- dimensional geometric model axis nutation unit 150, which receives the three-dimensional

geometric model, in one embodiment a conical polyhedron, and the nutation matrix from

the three-dimensional geometric model initialization unit 145 and the nutation matrix

initialization unit 175, respectively, and outputs a nutated, three-dimensional geometric

model, in one embodiment a nutated conical polyhedron. The nutated, three-dimensional

geometric model is forwarded to a three-dimensional geometric model projection unit 160,

which also receives the projection matrix from the projection matrix initialization unit 155.

The three-dimensional geometric model projection unit 160 outputs a nutated, projected

three-dimensional geometric model, in one embodiment a nutated, projected, conical

polyhedron, to an image data rasterization unit 165.

[0052] As set forth above, image data collected by the system 1 is received by

the data input unit 180 and arranged in a data table, such as that shown in Figure 6, including

angle and image data for each scan line. The image data is forwarded to the image data

rasterization unit 165 by the data input unit 180. The image data is then raterized by the

image data rasterization unit 165 onto the nutated, projected three-dimensional geometric

model, in one embodiment a nutated, projected, conical polyhedron output by the three-

dimensional geometric model projection unit 160. The data input unit 180 also provides the

leading scan line angle to the nutation matrix initialization unit 175 to be used in calculating

the nutation matrix.

[0053] As set forth above, Figure 2B is a block diagram of a rendering method

according to an embodiment of the invention. First, parameters, such as the forward

scanning angle, the radius of the base, the observer distance, the magnitude of nutation, the samples per line, the lines per frame, and the image data are received input by a user or an

imaging system, such as that shown in Figure 1, in step S202. In step S204, the image data is

arranged in a data table, such as that shown in Figure 6, including angle and image data for

each scan line. Then, using the received parameters, a three-dimensional geometric model,

projection matrix, and nutation matrix are initialized in steps S204, S208, and S210,

respectively. The order of the steps S204-S210 is not important and thus the order of these

steps may be changed. Next, using the three-dimensional geometric model, a conical

polyhedron in one embodiment, created in step S206 and the nutation matrix initialized in

step S210, an axis of the three-dimensional geometric model, in one embodiment a conical

polyhedron, is nutated to create a nutated, three-dimensional geometric model, in one

embodiment a nutated, conical polyhedron, in step S212. In step S214, the three-

dimensional geometric model is projected using the three-dimensional geometric model

projection matrix initialized in step S208. In step S216, the image data is rasterized onto the

nutated, projected, three-dimensional geometric modeL The various processing steps of

Figure 2B will be discussed in further detail below.

[0054] The rendering apparatus 140 of Figure 2A and rendering method of

Figure 2B are essentially a pipeline for processing three-dimensional coordinates as well as

image data. Referring to Figure 2A, the three-dimensional coordinates originate in the three-

dimensional geometric model initialization unit 145, and are subsequently transformed into

two-dimensional coordinates by the three-dimensional geometric model axis nutation unit

150 and the three-dimensional geometric model projection unit 160. The transformed coordinates are then sent to the image data rasterization unit 165 which rasterizes the image

data onto the nutated, projected, three-dimensional geometric model for scan conversion.

[0055] The overall pipeline depicted in Figures 2A and 2B is a variation of a

three-dimensional rendering pipeline as practiced in the field of computer graphics. The use

of homogeneous coordinates is well-established in this field, and the invention uses such

coordinates. In the invention, three-dimensional coordinates are "homogenized" in step

S212 by adding a fourth coordinate w, which is initialized to one. Thus, three-dimensional

coordinates (x, y, z) are promoted to homogeneous coordinates (x, y, z, 1). These

homogeneous coordinates are demoted back to three-dimensional coordinates at the

conclusion of step S214 such that coordinates of the form (x, y, z, w) are renormalized to

(x/w, y/w, z/w). This demotion or renormalization can also be thought of as the projection

of homogeneous coordinates of the form (x, y, z, w) into the w = K plane, where K is a

constant, often one.

[0056] Transformations of objects from so-called "object space" to "image

space" are accomplished by multiplying object-space coordinates by one or more

transformation matrices representing object rotation, scaling, translation, observer location

and orientation, and three-dimensional to two-dimensional projection. The transformation

matrices may be premultiplied into a . single matrix for efficiency. Because matrix

multiplication is not commutative, there are two different conventions for transforming

coordinates. The traditional convention as taught in mathematics and computer graphics

textbooks is to postmultiply the transformation matrices by column coordinate vectors. The

other convention is to premultiply the transformation matrices by row coordinate vectors. The transformation matrices must be transposed to go from one convention to the other, as

is known to practitioners skilled in the field of computer graphics. Embodiments of the

invention use the traditional convention of postmultiplying transformation matrices by

column coordinate vectors. This is illustrated by the following equation (1).

13 '14

Equation (1)

*34

where

Ui through t44 constitute aggregate transformation coefficients x, y, and z are coordinates of the point in three-dimensional space x\ y\ z\ andrø'are transformed homogeneous coordinates

[0057] In order to obtain the final two-dimensional coordinates in image

space, the transformed coordinates are renormalized to the form (x'/w', y'/w', z'/w^), and

then the z'/w' terms are simply dropped, yielding two-dimensional coordinates of the form

(x'/w', y'/V). These may have to be further and trivially mapped to physical device

coordinates suitable for addressing video monitor pixels.

[0058] The proper representation and manipulation of three-dimensional

objects requires the selection and consistent use of either a right-handed or left-handed

coordinate system. In a right-handed coordinate system, the (x, y, z) axes are arranged such

that a rotation of the x-axis into the y-axis (i.e. clockwise through 90°) corresponds to the

rotation of a right-handed screw advancing in the positive z direction. In the context of a

video monitor, when an observer is looking at the monitor, the x-axis points to the right, the y-axis points upward, and the z-axis comes "out" of the monitor toward the observer. This

means that the z-coordinates of most objects in three-dimensional space are negative, Le.

"inside" the monitor. Because of this somewhat awkward situation, three-dimensional

computer graphics systems, such as software libraries, often use left-handed coordinate

systems in order to locate objects in positive z-coordinate space. Nonetheless, mathematics

and computer graphics textbooks traditionally teach right-handed coordinate systems, and as

such, the discussion of the embodiments of the invention will also use a right-handed

coordinate system. A practitioner skilled in the field of computer graphics will be able to

easily convert between the two coordinate systems.

[0059] In step S206 in Figure 2B, a three-dimensional geometric model is

created of the scanning surface, in this exemplary embodiment a conical scanning surface.

In one embodiment, this cone is approximated by a series of connected planar, isosceles

triangular surfaces which define a polyhedron, as shown in Figure 4A. In one particular

alternative embodiment, the apex of the cone is open because the scan lines do not begin on

the longitudinal axis of the imaging catheter; rather, the scan lines begin at a radial offset

from this axis. Figure 4B shows an exemplary open conical polyhedron. In this alternative

embodiment, the open cone is approximated by a series of connected planar, regular

trapezoidal surfaces which define a polyhedron. In additional alternative embodiments, the

cone may be approximated by various types of parametric bicubic surface patches, including

but not limited to Hermite surfaces, Bezier surfaces, and B-Spline surfaces. The cone may

also be modeled exactly by using either implicit or explicit surface equations, including the

implicitly defined family of quadric surfaces. A survey of several surface representations is presented by Foley et al. in "Computer Graphic Principles and Practice," Second Edition,

Addison-Wesley Publishing Company, 1990, which is hereby incorporated by reference.

[0060] It is convenient, but not necessary, to choose the number of triangles

to be equal to the number of scan lines comprising a complete — 360° rotation of the

catheter. If this is the case, then each scan line represents a shared edge of two adjacent

triangles in the polyhedron. Furthermore, two adjacent scan lines represent two equal sides

of a single isosceles triangle. All scan lines emanate from the apex of the polyhedron and

diverge as they approach the base.

[0061] As the number of triangles increases, the polyhedron approaches a true

cone, but at the cost of more computational and storage resources required to realize the

apparatus. Conversely, defining the polyhedron with fewer triangles results in a rougher

approximation of a cone, but the computational and storage resources required are reduced.

As an illustrative example only, Figure 3 shows a catheter and transducer assembly along

with a depiction of a forward-scanning polyhedron. Figure 4A, discussed above, shows just

the polyhedron with 12 isosceles triangles corresponding to 12 scan lines per revolution.

Similarly, Figure 4B, discussed above, shows just the polyhedron with 12 regular trapezoids

corresponding to 12 scan lines per revolution; the hole at the apex of the polyhedron in

Figure 4B is an area adjacent the catheter and transducer assembly. The scan lines and

triangles are spaced at — 30° angular increments. This example is illustrative only because

approximating a cone with just 12 isosceles triangles is generally too rough an approximation

to produce a realistic-looking cone. Further, 12 scan lines per catheter revolution would

dramatically undersample the image with respect to the inherent lateral resolution of the transducer, resulting in poor image quality. One embodiment of the method and apparatus

according to the invention uses 400 isosceles triangles corresponding to 400 scan lines per

catheter revolution. This results in scan lines and triangles spaced at — 0.9° angular

increments.

[0062] When assigning three-dimensional coordinates to the geometric model,

its size, orientation, and location must be specified. First, it is convenient to define a small

number of variables to define the size of the geometric modeL Let

ψ = forward scanning angle R = radius of cone at base H = height of cone = R*t3ii(ψ)

In order to orient the geometric model such that the apex points directly toward the

observer, when the image is not being nutated, its longitudinal axis may be defined to be

coincident with the z-axis, with the apex of the model having a more positive z-coordinate

than the z-coordinates of the points on the base.

[0063] To locate the geometric model, the origin may be defined to be the

center of the base of the geometric model, i.e. the point along its longitudinal axis in the

plane of the base. This point has the three-dimensional coordinates (x, y, z) = (0, 0, 0).

Note that in order for nutations to be symmetrical about the longitudinal axis of the model,

the x- andy-coordinates must be equal to zero. However, the z-coordinate may be redefined

to vary the visual effect of nutation. For example, defining the origin at the apex results in

the apex remaining at a fixed location and the base gyrating when the image is nutated, while

defining the origin at the center of the base results in the apex gyrating and the base gyrating relatively less. When projecting the model into two dimensions for display, the choice of the z-coordinate for the origin has an effect on the total amount of screen area required to

display the image under all possible nutation angles. The value of this coordinate that

minimizes the screen area depends on the parameters of the geometric model, but the screen

area is never minimized when the origin is defined at the apex.

[0064] With the center of the base of the geometric model at (0, 0, 0), the

location of the apex is (0, 0, +H). Note that the observer must be located further "back"

than +H to avoid being "inside^* the geometric model. This is accomplished by a simple

translation of -H, plus any additional desired observer distance. The remaining coordinates

are located along the circular perimeter of the base of the polyhedron. Let

L = number of scan lines per complete catheter rotation

A = angle subtended per scan line = 360°/Z. n = variable used to represent the scan line number, ranges from 0 to 1-1 θ = scan line angle = n * A

Then, the position of the endpoint for scan line n may be provided by equation (2) as

follows:

P_n (x, V, Z) = (R cos(0), R sin(0), 0) Equation (2)

An example is provided below:

R = LO

^ = 30°

/f = 1.0*tan(30°) = 0.577

Apex of cone PA (X, y, z) = (0, 0, 0.577)

1 = 12

A = 360°/12 = 30°

The endpoints for each scan line are specified in the following table (1) below.

Table (1)

[0065] The three-dimensional polyhedron must be projected from three-

dimensional object space into two-dimensional image space. In one embodiment of the

method and apparatus according to the invention, a perspective projection is used to provide

depth cues to the observer. The simplest form for a perspective projection matrix where the

center of projection is at the origin (0, 0, 0) and the projection plane is normal to the z-axis

and at a distance z = d may be provided by equation (3) as follows:

1 0 0 0 0 1 0 0

P = Equation (3) 0 0 1 0 0 0 \ld 0

With d = -1, the projection matrix then simplifies as shown in exemplary equation (4) as

follows: 1 0 0 0 1 0

P = Equation (4) 0 0 1

0 0 - 1 0

An alternative form for a perspective projection matrix places the projection plane normal to

the z-axis at z=0 and the center of projection at (0, 0, -d) and may be provided by equation

(5) as follows:

1 0 0 0

0 1 0 0

P'= Equation (5)

0 0 0 0

0 0 Vd 1

This form allows the distance d from the center of projection to the projection plane to tend

to infinity. As the distance d approaches infinity, the perspective projection becomes a

parallel projection. In an alternative embodiment of the method and apparatus according to

the invention, a perspective projection matrix in the form of equation (5) is used. In an

additional alternative embodiment, a parallel projection is used by setting d = oo in equation

(5).

[0066] In one embodiment of the method and apparatus according to the

invention, an observer transformation is also applied which is a translation of the

polyhedron in the negative z direction by an observer distance D. The purpose of this

translation is to place the polyhedron entirely behind the projection plane, so the polyhedron

height H must be added to D. This transformation may be provided by equation (6) as

follows: O Equation (6)

[0067] The composite projection matrix, which includes both the observer

transformation and the projection matrices, maybe provided by equation (7) as follows:

O P = Equation (7)

[0068] Since matrix multiplication is not commutative, the observer

transformation must be applied before the projection. Further, it is equally valid to

postmultiply the nutation matrix by the observer transformation matrix as it is to premultiply

the projection matrix by the observer transformation matrix. In both cases, the observer

transformation must occur after the nutation transformation and before the projection.

[0069] In Figure 2A, the projection matrix is established by the projection

matrix initialization unit 155 using the parameters, forward scanning angle, radius at base,

and observer distance. The projection matrix is subsequently used by the three-dimensional

geometric model projection unit 160 to transform the nutated, three-dimensional geometric

model, in this example a polyhedron, into a nutated, projected three-dimensional geometric

model.

[0070] The means of acquiring ultrasound image data are well understood.

For example, U.S. Patent Application No. 11/261,635 by Goodnow, et aL, which is hereby incorporated by reference, describes the acquisition process for an exemplary IVUS imaging

system similar to an exemplary IVUS system which may be utilized with embodiments of the

invention. In particular, U.S. Patent Application No. 11/261,635 describes the generation of

a data table for storing catheter angle and ultrasound image data. Figure 6 depicts an

exemplary data table which may be utilized for embodiments of the invention. The data table

of Figure 6 has three columns. The first is simply an index, which may be omitted if the

indices are contiguous. The second column is the catheter angle, which is derived from data

received from the angle encoder 16 in Figure 1. The third column is the detected or

grayscale ultrasound image data. Note that the data in each of the rows in the third column

is actually an array of grayscale samples but that for notational simplicity these arrays are

treated as ImageData[n, t]. The number of grayscale samples in conjunction with the

sampling rate in samples per second and the speed of sound in water can be used to estimate

the scanning depth. A typical scanning depth is ~ five millimeters for an exemplary IVUS

system, such as that depicted in Figure 1. A complete data table is defined as a data table

with one row for every ultrasound scan line acquired in one complete, ~ 360° rotation of the

catheter.

[0071] Referring again to Figure 2A, the data input unit 180 may control the

ultrasound image acquisition process or receive grayscale data from another process, such as

an acquisition process in the electronics module 18 shown in Figure 1. U.S. Patent

Application No. 11/261,635 describes in detail an exemplary acquisition process. The data

input unit 180 creates with the acquired image data either a complete data table or a subset

of a complete data table which may be merged with previous data. In a manually rotated imaging system, the data table created by the data input unit 180 is likely to be a subset of a

complete data table. A complete data table contains the most recent version of the

ultrasound scan lines from each of the angles in one complete, — 360° rotation of the

catheter.

[0072] Referring again to Figure 2A, in addition to generating a data table, the

data input unit 180 also receives and outputs to the nutation matrix initialization unit 175 the

leading scan line angle. The leading scan line angle is defined as the catheter angle

corresponding to the most recently acquired ultrasound scan line. The leading scan line

angle is used in one embodiment by the nutation matrix initialization unit 175 to compute

the magnitude and direction of nutation.

[0073] In an IVUS system with a manually rotated catheter, the position of the

catheter changes relatively slowly as the operator manipulates the device. One embodiment

of the method and apparatus according to the invention takes advantage of this slow,

manually controlled rotation by setting the orientation of the longitudinal axis of the

geometric model or cone to track the angular position of the catheter. Specifically, the

longitudinal axis of the geometric model or cone is tilted to the opposite side or away from

the angular position or leading scan line angle of the catheter. When projected into two-

dimensional space, this makes the most recently acquired data larger and appear closer to the

observer and easier to view. The diametrically opposing data is foreshortened and therefore

deemphasized. The amount of tilt with respect to the z-axis is a quantity that is independent

of the leading scan line angle and may be a constant. This constant is the magnitude of

nutation and is expressed in angular units such as degrees. The overall tilt can be expressed as a rotation about a direction in the (x, y) plane, with the direction specified as a function of

the leading scan line angle. For example, if the leading scan line angle is 0° or 180°, then the

longitudinal axis of the geometric model or cone is rotated about the y-axis, and if the

leading scan line angle is 90° or 270°, then the longitudinal axis of the geometric model or

cone is rotated about the x-axis. For angles other than these special cases, the axis of rotation

may be given by the normalized direction vector in equation (8) as follows:

u = sin(0)i - cos(0) j Equation (8)

where

θ = Leading scan line angle

From the field of computer graphics, the rotation matrix for a rotation φ about an arbitrary

direction given by the direction vector U = (u_X) Uy_> U₂) is given by Rodrigues' rotation

formula, equation (8) as follows:

u_τ(\ - cosφ) + cos φ u_xu_y(\ - cos φ) - u. sinφ u_xu_z(l -cosφ) + u_y sinφ 0 u_xu_v(\ — cos φ) + u_z sin φ «* (1 - cos φ) + cos φ u_yu_z(y — cosφ) — u_x svn.φ 0

R = U_xU₁ (1 - cos ^) - u_y sin φ u_yu.{l - cosφ) + u_x sυxφ κ*(l -cos φ) + cos φ 0

0 0 0 1

Equation (9)

Substituting using equations (10.1), (10.2), and (10.3) as follows:

U_x= sin θ Equation (10.1)

Uy= -COS θ Equation (10.2)

U₂ = 0 Equation (10.3)

the matrix becomes equation (11) as follows: O

Equation (11)

[0074] Referring to Figure 2A, the nutation matrix . initialization unit 175

computes the nutation matrix according to Equation (11) using the leading scan line angle

(θ) and the magnitude of nutation (<$)

[0075] Referring again to Figure 2A, the axis of the three-dimensional

geometric model, or in one embodiment vertices of the conical polyhedron, are transformed

by the three-dimensional geometric model axis nutation unit 150 and the three-dimensional

geometric model projection unit 160 into an axis or vertices for a nutated, projected, three-

dimensional geometric model. In one embodiment of the method and apparatus according

to the invention, an intermediate vertex stream or nutated, three-dimensional geometric

model is produced. In an alternative embodiment, the nutation and projection

transformation matrices may be concatenated into a single matrix and vertices from the

three-dimensional geometric model are transformed directly into vertices for the nutated,

projected, three-dimensional geometric model.

[0076] The three-dimensional geometric model axis nutation unit 150 and the

three-dimensional geometric model projection unit 160 both transform homogeneous

coordinate streams according to equation (1). These equations expand to equation (12) as

follows: x' = t_ux + t_ny + t_{l∑ + f_Mw y' = f_2Ix + t^y + J₂₃Z +^■ t_uw

Equation (52) z' = t_3ιx + t_ny + t_Ω∑ + t_iiW w' = t_nx + t_Gy + t^z + t^w

where Ui through Ut are the coefficients in the nutation matrix used by the three-

dimensional geometric model axis nutation unit 150 and tn through Ut are the coefficients in

the projection matrix used by the three-dimensional geometric model projection unit 160.

Note that w is set to one by the three-dimensional geometric model axis nutation unit 150,

and to w' by the three-dimensional geometric model projection unit 160.

An example is provided below:

Use the conical polyhedron model defined in Table (1). From this example the polygon

height H is 0.577. Also, for illustrative purposes, let

D = 1.0 (observer distance)

Then, from equation (7), the composite projection matrix is as follows:

1 0 0 0 1 0 0 0

0 1 0 0 0 1 0 0

O P =

0 0 1 - (D ₊ H) 0 0 1 -1.577

0 0 -1 D + H 0 0 -1 1.577

To compute the nutation matrix, for illustrative purposes, let

0 = 30° (leading scan line angle) ^ = 15° (magnitude of nutation)

Then, from equation (11), the nutation matrix is as follows: - sin30^o-cos30°-(l - cosl5°) - COs SO⁰ SiIiIS⁰ 0 os²30°-(l - cosl5°) + cosl5^c -sin30^o-sinl5° 0 sin 30°- sin 15° cos 15° 0

0 0 1

Applying equation (12) with these coefficients yields the following:

x' = 0.974* - 0.015 v - 0.224z y' = -0.015* + 0.991>> - 0.129z z' = 0.224* + 0.129;/ + 0.966z w' = w

Applying equation (12) with the coefficients for the composite projection matrix yields the

following:

x" = x' y = y z" = z' - \.51lW v/ = -z' + 1.577w'

Finally, renormalizing the homogeneous coordinates by dividing through by w" yields the

following:

y = //(-_z" + 1.577O ∑^m = -l w" = l

The projected two-dimensional coordinates are obtained by dropping the fourth coordinate

and evaluating the remaining three-dimensional coordinates in the plane of projection, which

is z"' = -1, yielding (x"\ _Y"). [0077] Referring to Table (2) below, the original non-transformed three-

dimensional conical polyhedron coordinates are listed in the x, y, and z columns and are

identical to the coordinates listed in Table (1). The nutated coordinates are listed in the x',

y, z\ and w' columns. The nutated, projected coordinates are listed in the x", γ", z", and w"

columns. The renormalized two-dimensional coordinates are listed in the x'" and y'"

columns.

Table (2)

[0078] Figure 5 shows a nutated, projected polyhedron. The vertex numbers

are noted in the figure; these correspond to the indices in the first column of Table 1 and

Table 2. A grid is superimposed on the image to allow the (x, y) coordinates to be estimated.

The (x, y) coordinates correspond to the x'" and y'" coordinates listed in the last two

columns of Table 2.

[0079] Referring to Figure 2B, the final step, step S216, in computing the two-

dimensional representation of the nutated, projected three-dimensional geometric model is to scan-convert or rasterize the ultrasound grayscale data according to the transformed

version of the three-dimensional geometric model. In one embodiment, this is accomplished ^"

by the well-understood computer graphics process of texture-mapping the grayscale data

onto the nutated, projected conical polyhedron. Texture mapping was originally taught by E. .

Catmull in "A Subdivision Algorithm for Computer Display of Curved Surfaces," PhD

Thesis, Report UTEC-CS_c-74-133, Computer Science Department, University of Utah, Salt

Lake City, Utah, December 1974 and in "Computer Display of Curved Surfaces," Proc

IEEE, Conference on Computer Graphics, Pattern Recognition, and Data Structures, May

1975, and refined by JJF. Blinn and M.E. Newell in "Texture and Reflection in Computer

Generated Images," Communications of the ACM, 19(10), October 1976, pp. 542-547,

which are all hereby incorporated by reference. P.S. Heckbert in "Survey of Texture

Mapping," IEEE Computer Graphics and Applications, 6(11), November 1986, pp. 56-57,

which is hereby incorporated by reference, provides a thorough survey of texture-mapping

methods.

[0080] In the texture-mapping approach used in one embodiment of the

invention, the vertices of the conical polyhedron may be assigned texture coordinates (u, v)

in addition to their already-assigned (x, y, z) coordinates in three-dimensional space. The u

and v coordinates each range from 0 to 1 with the u-coordinate representing the relative

grayscale sample number and the v-coordinate representing the relative scan line number.

Referring to the exemplary data table shown in Figure 6, sample 0 corresponds to u = 0 and

sample SamplesPerline-l corresponds to u = 1, while line 0 corresponds to v = 0 and line LinesPerFrame-1 corresponds to v = 1. Further, sample 0 with u = 0 corresponds to the apex of the conical polyhedron while sample SamplesPerLine-1 with u = 1 corresponds to

vertices around the base. Note that the v-coordinate must take on multiple values at the

apex, depending on the specific scan line being processed. In order to accommodate this,

the apical vertex is duplicated into LinesPerFrame copies, each with identical (x, y, z) and u-

coordinates but different v-coordinates.

[0081] Step S216 in Figure 2B is accomplished by texture-mapping each of the

transformed isosceles triangles of the conical polyhedron into (x, y) image space. For each

transformed triangle, all the (x, y) image-space pixels are computed using any of a number of

well-understood polygon scan-conversion algorithms. Foley et al. in "Computer Graphic

Principles and Practice," Second Edition, Addison-Wesley Publishing Company, 1990, and

David A. Rogers in "Procedural Elements for Computer Graphics," McGraw-Hill Book

Company, 1985, which are both hereby incorporated by reference, provide descriptions of a

number of polygon scan-conversion algorithms. For each pixel in each transformed triangle,

the (u, v) coordinates of the texture map are computed from the (u, v) coordinates of each

of the three triangle vertices. Note that if a perspective transformation is being used, linear

interpolation of the (u, v) coordinates will result in distortion resulting in texture features

being improperly foreshortened- Mark Feldman in "The PC Game Programminer's

Encyclopedia: Texture Mapping," http://www.geocities.com/SiliconVallev/2151/tmap.htrnl.

which is hereby incorporated by reference, describes several well-understood texture-

mapping algorithms along with their relative accuracy (Le. perspective correctness), speed,

and implementation difficulty in Table 3, as shown below. For example, an approximate

solution can be obtained via the area subdivision algorithm by decomposing each isosceles triangle in the conical polyhedron into several smaller ones, each with its own set of (x, y, z)

and (u, v) vertex coordinates. An exact solution can be obtained via the "perfect" mapping

algorithm by performing the perspective division while interpolating.

Table 3

[0082] Once the (u, v) texture coordinates have been computed for each

image-space pixel (x, y), the (u, v) coordinates are mapped to indices into the data table

created by the data input unit 180. The u coordinate is mapped to a grayscale sample

number and the v coordinate is mapped to a scan line number. Since u and v are nominally

fractional numbers ranging from 0 to 1, a texture coordinate (u, v), when mapped to

fractional grayscale sample and scan line numbers, will be bounded by four grayscale samples

in the data table. In one embodiment, these four samples are linearly weighted according to

their relative distance to the exact fractional sample, then summed to produce the resultant

(x, y) image-space pixel value. This weighting summing technique is often referred to as

bilinear interpolation.

[0083] Certain embodiments of the invention may include adding lighting and

shading to enhance the visual realism of the three-dimensional geometric model. Foley et aL

"Computer Graphic Principles and Practice," Second Edition, Addison-Wesley Publishing

Company, 1990, which is incorporated by reference, describes several well-known lighting and shading models used in the field of computer graphics. One or more light sources is

defined and placed into the object space. In addition, the reflective properties of the three-

dimensional geometric model are specified. This may involve, for example, the calculation

of a normal vector for each of the vertices of the conical polyhedron model or at each point

on the surface of the model, as well as specifying the reflective properties such as, but not

limited to, diffuse or specular reflectivity.

[0084] In one embodiment of the invention, the lighting and shading

calculations are applied to the nutated axis or vertices of the three-dimensional geometric

model in an additional processing step interposed between steps S212 and step S214 in

Figure 2B. These calculations produce a relative intensity value for each of the vertices in

addition to the nutated coordinate values. These relative intensity values are passed through

step S214 and used in step S216 to produce a final image that incorporates the effect of the

lighting and shading.

[0085] In this embodiment, the light source may be placed along the positive

z-axis (note that although the observer is also placed along the positive z-axis, the light

source and observer will not interfere with each other). When the three-dimensional

geometric model is nutated, the portion containing the most recently acquired scan lines

reflects more of the incident light than the diametrically opposed, foreshortened portion

because the normal vectors to the surface around the former portion are more aligned with

the z-axis, Le. the dot product of these normals with the z-axis is relatively higher. The

deemphasized, foreshortened portion reflects light away from the observer because the surface normals are not well-aligned with the z-axis. Hence, it appears dimmer. The overall

effect is to enhance the three-dimensional nature of the model.

[0086] The embodiments described nutate the three-dimensional geometric

model in response to the manual rotation of the catheter. Additional embodiments may-

nutate the model at a constant or variable rate, or in response to manipulation of user

interface controls. For example, in a motor-driven IVUS imaging system, where the image is

being acquired repetitively several to many times per second, the model may be slowly

nutated clockwise or counterclockwise, with the rate of nutation being on the order of one

nutation cycle per second Furthermore, user interface controls such as a computer pointing

device may be used to interactively nutate the model. This technique allows the model to be

nutated or tilted at any orientation desired by the user. The overall nutatation angle with

respect to the z-axis may also be gradually or abruptly increased as new image lines are being

added and gradually or abruptly decreased after a certain time period in which no new image

lines were added. Referring to Figure 2A, in these additional embodiments, the nutation

matrix initialization unit 175 is modified to accept and/or generate nutation angles from

sources other than the leading scan line angle. For example, for slow periodic nutation and

overall nutation angle increases and decreases a timer may be used and for interactive

nutation the coordinates of a computer pointing device may be used to generate and/or

modulate nutation angles.

[0087] Graphics markers or icons may be superimposed on the final rasterized

image. Such markers or icons may indicate, for example, the location and size of one or

more auxiliary diagnostic and therapeutic devices and/or where their effects will occur. Referring to Figure 2A, in one embodiment, the coordinates of these objects are defined in

object space and transformed by the three-dimensional geometric model axis nutation unit

150 and the three-dimensional geometric model projection unit 160. Rasterization is

accomplished by a process similar but not necessarily identical to that performed by the

image data rasterization unit 165.

[0088] An additional embodiment allows the image to persist for a finite

period of time after which it is either gradually or abruptly removed. This technique may be

applied to individual scan lines within the image as well as to the entire image. One

embodiment of this is to gradually fade each of the individual scan lines as more time elapses

without it being updated by a fresh scan line. This has the effect that the least recently

acquired or "oldest" image data becomes progressively dimmer with time while the most

recently acquired or "newest" image is the brightest. This visually suggests that the older

data has decreasing clinical value over time while the newest data has the greatest clinical

value.

[0089] Figures 7A-7F are still frames taken from an exemplary display utilizing

an apparatus and method according to embodiments of the invention. That is, Figures 7A-

7F are still frames taken from a video display. The conical forward-looking image data has

been rasterized onto a nutated, projected three-dimensional geometric model in the form of

a nutated, projected three-dimensional conical polyhedron. The yellow line represents the

leading scan line, and rotates as image data is added. The red marking represents the

position of a therapeutic and/or treatment device, such as an ablation device. [0090] The foregoing embodiments and advantages are merely exemplary and

are not to be construed as limiting the invention. The present teaching can be readily

applied to other types of apparatuses. The description of the invention is intended to be

illustrative, and not to limit the scope of the claims. Many alternatives, modifications, and

variations will be apparent to those skilled in the art. In the claims, means-plus-function

clauses are intended to cover the structures described herein as performing the recited

function and not only structural equivalents but also equivalent structures.

Claims

ΛEΗAT IS CLAIMED IS:

1. A method of rendering for display image data from a forward-looking conical

section of tissue coUected by an image data collecting device, comprising:

receiving the image data;

rasterizing the image data in accordance with a projected three-dimensional

geometric model to produce a rasterized image, thereby giving the appearance of three

dimensions when displayed

2. The method of claim 1, further comprising:

organizing the received image data in the form of a data table, including an

angle and corresponding image data for each scan line of the image data.

3. The method of claim 1, wherein rasterizing the image data in accordance with

the projected three-dimensional geometric model comprises orienting scan lines of the image

data in a correct geometric perspective on a surface of the three-dimensional geometric

model.

4. The method of claim 3, wherein the projected three-dimensional geometric

model comprises a cone.

5. The method of claim 4, wherein the cone is approximated by a plurality of

parametric bicubic surface patches.

6. The method of claim 5, wherein the plurality of parametric bicubic surface

patches comprise at least one of Hermite surfaces, Bezier surfaces, or B-spline surfaces.

7. The method of claim 4, wherein the cone is modeled using implicit surface

equations.

8. The method of claim 4, wherein the cone is modeled using explicit surface

equations.

9. The method of claim 3, wherein the projected three-dimensional geometric

model comprises a polyhedron.

10. The method of claim 9, wherein the projected three-dimensional geometric

model comprises a conical polyhedron.

11. The method of claim 10, wherein the conical polyhedron comprises a number

of triangles equivalent to a number of scan lines of the image data.

12. The method of claim 11, wherein the conical polyhedron comprises a number

of triangles equivalent to a number of scan lines of the image data for a complete 360°

rotation of the image data collection device.

13. The method of claim 11, wherein the number of triangles comprises at least

400 triangles.

14. The method of claim 10, wherein the conical polyhedron comprises a number

of trapezoids equivalent to a number of scan lines of the image data such that an apex of the

conical polyhedron is open.

15. The method of claim 1, further comprising:

nutating an axis of the projected three-dimensional geometric model.

16. The method of claim 15, wherein nutating the axis of the projected three-

dimensional geometric model comprises nutating the axis of the three-dimensional

geometric model at at least one of a constant rate or a variable rate.

17. The method of claim 15, wherein nutating the axis of the projected three-

dimensional geometric model comprises at least one of gradually or abruptly increasing the

magnitude of nutation of the axis of the three-dimensional geometric model as scan lines of

new image data are processed after a predetermined time period in which no new scan lines

are added.

18. The method of claim 15, wherein nutating the axis of the projected three-

dimensional geometric model comprises at least one of gradually or abruptly reducing the magnitude of nutation of the axis of the three-dimensional geometric model after a

predetermined time period in which no new scan lines were added.

19. The method of claim 15, wherein nutating the axis of the projected three-

dimensional geometric model comprises nutating the axis of the three-dimensional

geometric model via manipulation of a user interface.

20. The method of claim 19, wherein the user interface comprises a computer

pointing device.

21. The method of claim 19, wherein the user interface comprises an angle

encoder.

22. The method of claim 15, wherein nutating the axis of the three-dimensional

geometric model comprises nutating the axis of the three-dimensional geometric model away

from scan lines corresponding to most recently acquired image data to emphasize the scan

lines of the most recently acquired image data and de-emphasize via foreshortening lines

diametrically opposed to the scan lines of the most recently acquired image data.

23. The method of claim 1, further comprising:

initialising a three-dimensional geometric model;

initializing a projection matrix; and projecting the mree-dimensional geometric model using the projection matrix

o create the projected three-dimensional geometric modeL

24. The method of claim 23, further comprising:

initializing a nutation matrix; and

nutating an axis of the three-dimensional geometric model using the nutation

matrix prior to projecting the three-dimensional geometric model using the projection matrix

to create the projected three-dimensional geometric model.

25. The method of claim 1, further comprising:

adding lighting, shading an/or texture before, during, or after rasterizing the

image data.

26. The method of claim 1, further comprising:

adding markers and/or icons before, during, or after rasterizing the image

data.

27. The method of claim 26, wherein the respective added markers and/or icons

represent a location of one or more auxiliary diagnostic and therapeutic devices and/or

where their effects will occur.

28. The method of claim 26, wherein the respective added markers and/or icons

represent depth markers.

29. The method of claim 1, further comprising:

displaying the rasterized image.

30. The method of claim 29, further comprising:

displaying the rasterized image for a predetermined period of time.

31. The method of claim 30, further comprising:

at least one of gradually or abruptly removing the rasterized image after the

predetermined period of time.

32. Apparatus for rendering for display image data from a forward-looking conical

section of tissue collected by an image data collection device, comprising:

a data input unit configured to receive the image data; and

an image data processing unit configured to rasterized the image data in

accordance with a projected three-dimensional geometric model to produce a rasterized

image, thereby giving the appearance of three-dimensions when displayed.

33. The apparatus of claim 32, wherein the data input unit organizes the image

data in the form of a data table, including an angle and corresponding image data for each

scan line of the image data.

34. The apparatus of claim 32, wherein the image data processing unit orients scan

lines of the image data in a correct geometric perspective on a surface of the projected three-

dimensional geometric model.

35. The apparatus of claim 34, wherein the projected three-dimensional geometric

model comprises a cone.

36. The apparatus of claim 35, wherein the cone is approximated by a plurality of

parametric bicubic surface patches.

37. The apparatus of claim 36, wherein the plurality of parametric bicubic surface

patches comprise at least one of Hermite surfaces, Bezier surfaces, or B-spline surfaces.

38. The apparatus of claim 35, wherein the cone is modeled using implicit surface

equations.

39. The apparatus of claim 35, wherein the cone is modeled using explicit surface

equations.

40. The apparatus of claim 34, wherein the projected three-dimensional geometric

model comprises a polyhedron.

41. The apparatus of claim 40, wherein the projected three-dimensional model

comprises a conical polyhedron.

42. The apparatus of claim 41, wherein the conical polyhedron comprises a

number of triangles equivalent to a number of scan lines of the image data.

43. The apparatus of claim 42, wherein the conical polyhedron comprises a

number of triangles equivalent to a number of scan lines of the image data for a complete

360° rotation of the image data collection device.

44. The apparatus of claim 42, wherein the number of triangles comprises at least

400 triangles.

45. The apparatus of claim 42, wherein the conical polyhedron comprises a

number of trapezoids equivalent to a number of scan lines of the image data such that an

apex of the conical polyhedron is open.

46. The apparatus of claim 32, wherein the image data processing unit comprises:

a three-dimensional geometric model initialization unit configured to initialize

a three-dimensional geometric model;

a projection matrix initialization unit configured to initialize a projection

matrix; and

a three-dimensional geometric model projection unit configured to project the

three-dimensional geometric model using the projection matrix.

47. The apparatus of claim 46, wherein the image data processing unit further

comprises:

a nutation matrix initialization unit configured to initialize a nutation matrix;

and

a three-dimensional geometric model axis nutation unit configured to nutate

an axis of the three-dimensional geometric model using the nutation matrix prior to

projecting the three-dimensional model using the projection matrix to produce a nutated,

projected three-dimensional geometric modeL

48. The apparatus of claim 47, further comprising:

a user manipulation device configured to control nutation of the axis of the

three-dimensional geometric modeL

49. The apparatus of claim 48, wherein the user manipulation device comprises a

computer pointing device.

50. The apparatus of claim 48, wherein the user manipulation device comprises an

angle encoder.

51. The apparatus of claim 47, wherein the image data processing unit further

comprises:

an image data rasterization unit configured to rasterize the image data in

accordance with the nutated, projected, three-dimensional geometric model.

52. The apparatus of claim 51, wherein the image data rasterization unit is further

configured to add lighting, shading, and/or texture before, during, or after rasterizing the

image data to enhance the appearance of three-dimensions when displayed.

53. The apparatus of claim 51, wherein the image data rasterization unit is further

configured to add markers and/or icons before, during, or after rasterizing the image data.

54. An imaging system comprising the rendering^' apparatus of claim 32.

55. The imaging system of claim 54, wherein the imaging system comprises an

IVUS system.

56. A display apparatus comprising the rendering apparatus of claim 32.

57. An imaging apparatus that scans a conical forward-looking section of tissue,

comprising:

a transducer that is rotated in a forward-looking orientation for transmitting a

signal and receiving image data resulting from interaction of that transmission and tissues it

encounters;

electronic circuitry that processes the received image data; and

a processing unit that renders the image data giving the appearance of having

been displayed on the surface of a three-dimensional geometric model.

58. The apparatus of claim 57, wherein the three-dimensional geometric model

comprises a three-dimensional cone.

59. The imaging apparatus of claim 58, further comprising:

means for adding three-dimensional lighting, shading, and/or texture to the

three-dimensional cone.

60. The imaging apparatus of claim 58, further comprising:

means for nutating an axis of the three-dimensional cone.

61. The imaging apparatus of claim 58, further comprising:

means for superimposing one or more markers and/or icons on the three-

dimensional cone for accurately indicating a location of one or more auxiliary diagnostic and

therapeutic devices and/or where they will have effect.

62. The imaging apparatus of claim 61, further comprising means for

superimposing depth markers.

63. The imaging apparatus of claim 58, further comprising:

an interventional device, wherein the transducer is configured to image a

portion of tissue while the interventional device operates on or near the tissue being imaged.

64. The imaging apparatus of claim 58, further comprising:

means for displaying the image data to persist on the display for a finite period

of time, after which it is either gradually or abruptly removed.

65. A method for displaying image data that has been acquired from a conical

forward-looking section of tissue, comprising:

receiving the image data; and

displaying the image data by orienting scan lines of the image data in their

correct geometric perspective on a surface of a three-dimensional geometric model.

66. The method of claim, 65, wherein the three-dimensional geometric model

comprises a three-dimensional cone.

67. The method of claim 66, further comprising;

nutating an axis of the three-dimensional cone.

68. The method of claim 67, further comprising:

nutating an axis of the three-dimensional cone in real-time in response to new

image lines being added.

69. The method of claim 67, further comprising:

nutating the axis of the three-dimensional cone away from most recently-

acquired lines so as to emphasize the most recently acquired lines and deemphasize via

foreshortening lines diametrically opposed to the most recently acquired lines.

70. The method of claim 67, further comprising:

nutating the axis of the three-dimensional cone in real-time at at least one of a

constant or variable rate independent of new image lines being added.

71. The method of claim 67, further comprising:

nutating the axis of the three-dimensional cone in response to manipulation of

a user interface control device.

72. The method of claim 71, where the user interface control device comprises a

computer pointing device.

73. The method of claim 71, where the user interface control device comprises an

angle encoder.

74. The method of claim 67, further comprising:

gradually or abruptly reducing the magnitude of nutation of the three-

dimensional cone in response to no new image lines being added after a certain time period-

75. The method of claim 67, further comprising:

gradually or abruptly increasing the magnitude of nutation of the three-

dimensional cone in response to new image lines being added after a certain time period in

which no new image lines were added.

76. The method of claim 67, further comprising:

lighting, shading, and/or texturing the image of the three-dimensional cone in

real-time as it nutates.

77. The method of claim 67, further comprising:

superimposing a location of one or more auxiliary diagnostic and therapeutic

devices or a location of one or more actions on the three-dimensional cone.

78. The method of claim 65, further comprising:

retaining the image data on the display for a predetermined period of time and

thereafter either gradually or abruptly removing the image data.