This paper provides
auxiliary results for our general scheme of
computed tomography. In 3D
parallel-beam geometry, we first demonstrate
that the inverse Fourier transform in different
coordinate systems leads to different
reconstruction formulas and explain why the
Radon formula cannot directly work with
truncated projection data. Also, we introduce a
gamma coordinate system, analyze its properties, compute the Jacobian of the coordinate transform, and define weight functions for the inverse Fourier transform assuming a simple scanning model. Then, we generate Orlov's theorem and a weighted Radon formula from the inverse Fourier transform in the new system. Furthermore, we present the motion equation of the frequency plane and the conditions for sharp points of the instantaneous rotation axis. Our analysis on the motion of the frequency plane is related to the Frenet-Serret theorem in the differential geometry.
1. Introduction
In [1], we developed a general scheme for 2D and 3D, parallel- and divergent-beam computed tomography (CT). Different from traditional Radon's or Tuy's formulas starting with the inverse Fourier transform in the spherical coordinate system, our new framework is based on an instantaneous cylindrical coordinate system. However, in [1], the coordinate system was not formally defined, and the Jacobian of the transform was not explicitly calculated. Given the basic role of our general scheme in CT theory and some confusion in the literature [2–5], here, we treat the Γ coordinate system in a mathematically strict way.
The organization of this paper is as follows. In Section 2, we demonstrate how the inverse Fourier transform formula generates different reconstruction formulas, with a special attention on different properties of projection truncation. In Section 3, we define the Γ coordinate system and deduce its Jacobian factor. In Section 4, we study the motion of the frequency plane and present the condition for sharp points of the instantaneous rotation axis. In Section 5, a simplified model for helical CT reconstruction is given. In the appendixes, we present a weighted Λ reconstruction formula, a weighted Radon's formula, a deduction of Orlov's formula, and an alternative way to study the motion of the frequency plane. For convenience, we use the same notations as in [1].
2. Inverse Fourier Transform in Commonly Used Coordinate Systems
In this section, we write the inverse Fourier transform in commonly used coordinate systems, derive corresponding reconstruction formulas, and demonstrate their difference on truncation of projection data.
As shown in Figure 1, in the 3D space R3,Ψ(r⃗) is an object function to be reconstructed, whose Fourier transform is Ψ̂(k⃗). According to Fourier analysis, we have
Ψ(r⃗)=∫-∞∞∫-∞∞∫-∞∞Ψ̂(k⃗)exp(2πik⃗·r⃗)dk1dk2dk3,
where r⃗=(x,y,z) and k⃗=(k1,k2,k3) are 3D vectors in the real space and Fourier domain, respectively. The unit vectors along the three axes are i⃗,j⃗, and k⃗. Note that the frequency vector k⃗ and the unit vector k⃗ are different. This formula states that to reconstruct an 3D object Ψ(r⃗), we need its Fourier transform Ψ̂(k⃗) in the whole 3D space. In X-ray CT, the Fourier transform Ψ̂(k⃗) can be calculated when sufficiently many parallel projections are measured.
3D reconstruction problem in parallel-beam geometry and the cylindrical coordinate system. (a) A parallel-beam scans an object along a semicircle locus on the unit sphere; (b) In the frequency domain, when ϕ0 increases from 0 to π, the frequency plane ϕ=ϕ0 scans every point exactly once except for the points on the k3 axis, and the end point of the normal vector OS⃗ moves along the equator from (0,-1,0) to (0,1,0); (c) For simplicity, we suppose that the detector panel passes through the origin.
Now, let us specify the simplest 3D reconstruction problem. In Figure 1(a), S(ϕ)=S(sinϕ,-cosϕ,0) with ϕ∈[0,π] is a point on a great semicircle of the unit sphere. In the 3D space, we define the unit vectors e⃗=OS⃗=(sinϕ,-cosϕ,0) and e⃗⊥=(cosϕ,sinϕ,0). Then, e⃗,e⃗⊥ and k⃗ are another set of orthogonal unit vectors. As shown in Figure 1(c), the new components of the vector r⃗ can be denoted as u=r⃗·e⃗,ρ=r⃗·e⃗⊥,z=r⃗·k⃗.
The projection at point S is defined by PS(ρ,z)=Pϕ(ρ,z)=∫-∞∞Ψ(ue⃗+ρe⃗⊥+zk⃗)du.
Our purpose is to reconstruct Ψ(r⃗) when Pϕ(ρ,z) is known for all ϕ∈[0,π]. Let us see how the selection of a coordinate system will determine the reconstruction formula, paying an attention to longitudinal data truncation.
2.1. Signed Cylindrical Coordinate System
In a cylindrical coordinate system, the inverse Fourier transform can be rewritten as Ψ(r⃗)=∫0π∫-∞∞∫-∞∞Ψ̂(k⃗)exp(2πik⃗·r⃗)|kρ|dkρdk3dϕ,
where (kρ,k3,ϕ) with -∞<kρ<∞,-∞<k3<∞,0≤ϕ<π is the signed cylindrical coordinates of k⃗. As shown in Figure 1(b), we have k⃗=k⃗(kρ,k3,ϕ)=kρe⃗⊥(ϕ)+k3k⃗.
When ϕ0 increases from 0 to π, the plane ϕ=ϕ0 passes through every frequency point exactly once (except for the points on the axis Ok3 which have a zero measure), and the normal vector OS⃗ of the plane ϕ=ϕ0 moves along a great semicircle on the unit sphere with the starting point A(0,-1,0) and the end point B(0,1,0). Note that the coordinate transform from (k1,k2,k3) to (kρ,k3,ϕ) is one to one (except for the points on the axis Ok3 with zero measure), and the absolute value of Jacobian is |kρ|.
From the parallel projection Pϕ(ρ,z), we can calculate the Fourier transform of the object Ψ̂(kρ,k3,ϕ)=P̂ϕ(kρ,k3)=∫-∞∞∫-∞∞Pϕ(ρ,z)exp[-2πi(kρρ+k3z)]dρdz.
Substituting (6) into (4), we have Ψ(r⃗)=∫0π∫-∞∞∫-∞∞P̂ϕ(kρ,k3)exp(2πik⃗·r⃗)|kρ|dkρdk3dϕ.
The reconstruction scheme behind this formula can be denoted as Pϕ(ρ,z)→Ψ̂(k⃗)→Ψ(r⃗),
which indicates that we first calculate all the Fourier components from projections and then reconstruct the object using the inverse Fourier transform. In this process, the object is reconstructed as a whole body.
In fact, Formula (7) can be written in the filtered backprojection form as Ψ(r⃗)=∫0πP̃ϕ(ρ,z)dϕ,
where the filtered projection is given by P̃ϕ(ρ,z)=∫-∞∞∫-∞∞P̂ϕ(kρ,k3)exp(2πik⃗·r⃗)|kρ|dkρdk3=∫-∞∞∫∞∞P̂ϕ(kρ,k3)exp(2πikρρ)exp(2πik3z)|kρ|dkρdk3=Pϕ(ρ,z)*1-2π2ρ2.
The symbol * denotes the 1D convolution operation about variable ρ. Since the filtration is along the horizontal direction only, the projection can be longitudinally truncated, that is, a plane with a specified z value can be reconstructed independently. The reconstruction scheme behind (9) with (10) can be denoted by Ψ̂(k⃗)→Pϕ(ρ,z)→Ψ(r⃗),
which states that certain frequency components form the projection, and the object can be reconstructed directly from projections.
When projections are longitudinally truncated, the Fourier transform cannot be calculated. How and why can we reconstruct a part of the object?
The answer is as follows. What we want to reconstruct is the object Ψ(r⃗) instead of the Fourier transform Ψ̂(k⃗). Though we cannot calculate every single frequency component, we can calculate the contribution of the frequency components on the whole frequency plane (as well as its neighborhood) to a point to be reconstructed, see (10). In the aforementioned case, projections contain as all frequency components as the object does, but the ratio of the frequency components is not suitable. Then, the only thing we need to do is to perform a 1D filtration to adjust this ratio. An analog is that the Fourier components can be considered as raw material, projections a semifinished product, and the reconstructed object the final product. If we use the semifinished product to make a product, we do not need change it back to the raw material. The role of Fourier analysis is to tell us how far from the semifinished product to the final product. Simply speaking, the essence of image reconstruction from projections is filtration. Therefore, a suitable coordinate system should show the character of the filtration in an easy way.
2.2. Signed Spherical Coordinate System
In the signed spherical coordinate system shown in Figure 2, the inverse Fourier transform can be rewritten as Ψ(r⃗)=∫0π∫-π/2π/2∫-∞∞Ψ̂(k⃗)exp(2πik⃗·r⃗)k2cosϑdkdϑdϕ,
where (k,ϑ,ϕ) with -∞<k<∞,-π/2<ϑ<π/2,0≤ϕ<π is the signed spherical coordinates of k⃗.
The sphere coordinate system and the detector plane. (a) The sphere coordinate system in the Fourier domain, and (b) the relationship between projections and the Radon transform.
For convenience, we denote k⃗=k⃗(k,ϑ,ϕ)=kn⃗, where n⃗=n⃗(ϑ,ϕ)=e⃗⊥(ϕ)cosϑ+k⃗sinϑ is a unit vector.
For a unit vector n⃗=n⃗(ϑ,ϕ), and a real number l∈(-∞,∞), the Radon transform of the object is defined as an integral RΨ(l,n⃗)=RΨ(l,ϑ,ϕ)=∬r⃗·n⃗=lΨ(r⃗)d2r⃗,
on the plane described by r⃗·n⃗=l.
The Fourier transform can be calculated from the Radon transform by Ψ̂(k⃗)=Ψ̂(kn⃗)=∫-∞∞RΨ(l,n⃗)exp(-2πikl)dl.
Thus, the inverse Fourier transform (12) becomes Radon's formulaΨ(r⃗)=-14π2∫0π∫-π/2π/2∂2∂l2|l=r⃗·n⃗RΨ(l,n⃗)cosϑdϑdϕ=-14π2∇2∫0π∫-π/2π/2RΨ(l,n⃗)|l=r⃗·n⃗cosϑdϑdϕ=-18π2∇2∫ΩRΨ(l,n⃗)|l=r⃗·n⃗dn⃗,
where ∇2 is the Laplace operator, and Ω is the unit sphere.
Radon's formula tells us that to reconstruct the object at a point, what we need is the Radon transform of the planes through the point and its neighborhood. However, what we measure by the detector is line integrals along X-rays instead of the Radon transform [6] but it can be calculated from projections by RΨ(l,n⃗)=RΨ(l,ϑ,ϕ)=∫-∞∞Pϕ(lcosϑ-vsinϑ,lsinϑ+vcosϑ)dv.
We call (16) the first relation between the Radon transform and parallel projection.
Formulas (15) and (16) form a scheme to reconstruct the object from its parallel projections. Here, the projection is not allowed to be longitudinally truncated. From Section 2.1, however, we have seen that the object can be reconstructed slice by slice from parallel projections. Why is the truncation not allowed when using the Radon formula?
For comparison, we rewrite Radon’s formula in the filtered backprojection format Ψ(r⃗)=∫0πP̃ϕ(ρ,z)dϕ.
with P̃ϕ(ρ,z)=∫-∞∞∫-π/2π/2Ψ̂(k⃗)exp(2πik⃗·r⃗)k2cosϑdkdϑ=1-4π2∫-π/2π/2∂2∂l2|l=r⃗·n⃗=ρcosϑ+zsinϑRΨ(l,n⃗)cosϑdϑ.
Comparing (18) with (10), we have the following relation between the Radon transform and parallel-beam projections
1-4π2∫-π/2π/2∂2∂l2|l=r⃗·n⃗=ρcosϑ+zsinϑRΨ(l,n⃗(ϑ,ϕ))cosϑdϑ=Pϕ(ρ,z)*1-2π2ρ2.
We call this the second relation between the Radon transform and parallel-beam projections. Based on the relationship between fan-beam and parallel-beam projections [1], the right hand side of (19) can be expressed in terms of cone-beam projections as well.
Substituting (16) into (19), we get an identity of parallel projections 1-4π2∫-π/2π/2∂2∂l2|l=ρcosϑ+zsinϑ×∫-∞∞Pϕ(lcosϑ-vsinϑ,lsinϑ+vcosϑ)dvcosϑdϑ=Pϕ(ρ,z)*1-2π2ρ2.
Now, we see that reconstruction of an object from parallel-beam projections does allow truncation of projection data but to verify the identity (20) needs the information on the whole projection domain. Reconstruction from parallel-beam projections using the Radon formula means numerically verifying this identity. We point out that if the projection Pϕ(ρ,z) is replaced by a general function with two variables, this identity holds true as well; see Appendix A.
Similarly, in cone-beam CT, reconstruction from cone-beam projections using the Radon formula, that is, the Grangeat's framework, means numerically verifying the cone-beam version of this identity. This is the reason why the Grangeat framework cannot deal with the long object problem.
We see that the Radon formula comes from the inverse Fourier transform in the signed spherical coordinate system. In this simplest reconstruction problem, there are two shortcomings with the Radon formula: (1) what we measure with X-ray is line integrals rather than planar integrals, (2) the projection cannot be longitudinally truncated.
2.3. Standard Spherical Coordinate System
In a standard spherical coordinate system, the inverse Fourier transform is written as Ψ(r⃗)=∫02π∫-π/2π/2∫0∞Ψ̂(k⃗)exp(2πik⃗·r⃗)k2cosϑdkdϑdϕ,
where (k,ϑ,ϕ) with 0≤k<∞,-π/2<ϑ<π/2,0≤ϕ<2π is the standard spherical coordinates of k⃗, as shown in Figure 2.
From the well-known generalized function relations [7] δ(l)=∫-∞∞exp(2πikl)dk,iπl=∫-∞∞sgn(k)exp(2πikl)dk,
we have 12δ(l)+i2πl=∫0∞exp(2πikl)dk,12δ′′(l)+iπl3=-4π2∫0∞k2exp(2πikl)dk.
Based on the convolution theorem, we have ∫0∞Ψ̂(k⃗)exp(2πik⃗·r⃗)k2dk=∫0∞Ψ̂(kn⃗)exp(2πikl)k2dk=-18π2∂2∂2lRΨ(l,n⃗)-i4π3l3*RΨ(l,n⃗).
Therefore, the inverse Fourier transform becomes Ψ(r⃗)=-18π2∫02π∫-π/2π/2∂2∂l2RΨ(l,n⃗)|l=r⃗·n⃗cosϑdϑdϕ-i4π3∫02π∫-π/2π/2[RΨ(l,n⃗)*1l3]|l=r⃗·n⃗cosϑdϑdϕ.
Based on the odd and even symmetry, we have the Radon formula again [6] Ψ(r⃗)=-14π2∫0π∫-π/2π/2∂2∂l2RΨ(l,n⃗)|l=r⃗·n⃗cosθdθdϕ.
The difference between the signed and standard spherical coordinate systems is little, and the signed spherical coordinate system leads to Radon's formula more easily.
It is important to underline that the inverse Fourier transform may yield various reconstruct formulas in different coordinate systems and allow different degrees of data truncation. Among the commonly used coordinate systems, the signed cylindrical coordinate system is the most convenient one to solve the simplest 3D reconstruction problem. To handle the general 3D reconstruction problem, in the following we will introduce a variant of the signed cylindrical coordinate system, which is referred to as the Γ coordinate system.
3. Γ Coordinate System and Its Jacobian Factor
What plays an important role in the 3D reconstruction field is a variant of the cylindrical coordinate system, which can be named the Γ coordinate system.
As shown in Figure 3(b),ΓAB is a three-times differentiable curve connecting points A(0,-1,0) and B(0,1,0) on the unit sphere, whose length is θ0≥π.S(θ) is a point on ΓAB parameterized by θ∈[0,θ0] the length of the segment AS. Let us introduce three orthogonal unit vectors e⃗3=OS⃗,e⃗1=ddθe⃗3,e⃗2=e⃗3×e⃗1,
where e⃗1 represents the tangential direction, e⃗2 the instantaneous rotation axis of the frequency plane. Note that they are functions of θ=∫AS|de⃗3| (see Figures 3(a) and 3(b)).
Nonuniqueness of the representation of a point in the Γ coordinates. (a) The Γ coordinate system, (b) a general curve ΓAB on the unit sphere, and (c) the intersections of ΓAB and the great circle orthogonal to a given vector k⃗. When S moves from A to B ΓAB, some region may be scanned more than once.
In the Fourier domain, we call the plane through the origin O and orthogonal to the vector OS⃗ the frequency plane, denoted as ∏(OS⃗). Every point k⃗∈∏(OS⃗) can be expressed as
k⃗=ω1e⃗1+ω2e⃗2,
with ω1=k⃗·e⃗1,ω2=k⃗·e⃗2.
More generally, for every given triplet (ω1,ω2,θ) with ω1∈(-∞,∞),ω2∈(-∞,∞),θ∈[0,θ0], we can define a 3D vector
k⃗=k⃗(ω1,ω2,θ)=ω1e⃗1(θ)+ω2e⃗2(θ),
and the triplet (ω1,ω2,θ) is a set of Γ coordinates of the point k⃗.
However, a given point k⃗∈R3 may be represented by several sets of Γ coordinates, because when S(θ) moves along ΓAB, the rotation axis e⃗2 keeps changing, and hence the frequency plane ∏(OS⃗) may scan the given point k⃗ more than once. The number of times of the point being scanned, denoted as J(k⃗), equals the number of intersections between the curve ΓAB and the great circle orthogonal to the vector k⃗ (Figure 3(c)). To find the all intersections, one can solve the equation
k⃗·e⃗3(θ)=0,
to obtain the solutions θ1,θ2,…,θJ(k⃗). For every solution, such as θ1, one can further have ω11=k⃗·e⃗1(θ1),ω21=k⃗·e⃗2(θ1). Thus, one gets the J(k⃗) groups of coordinates
k⃗=k⃗(ω11,ω21,θ1)=k⃗(ω12,ω22,θ2)=⋯=k⃗(ω1J(k⃗),ω2J(k⃗),θJ(k⃗)).
We call this new system the Γ coordinate system, since the system is based on the curve ΓAB. This coordinate system is a variant of the cylindrical system. For a differential arc SS′, the new system is similar to a cylindrical coordinate system if the scanned volume is considered. When ΓAB is a great semicircle, the Γ coordinate system becomes the cylindrical coordinate system.
In the Γ coordinate system, the inverse Fourier transform can be expressed as
Ψ(r⃗)=∫0θ0∫-∞∞∫-∞∞W(ω1,ω2,θ)exp(2πik⃗·r⃗)×Ψ̂(k⃗)|ω1|dω1dω2dθ,
with k⃗=ω1e⃗1(θ)+ω2e⃗2(θ). In [1], we obtained the absolute value of the Jacobian, |ω1| for the coordinate transform from (k1,k2,k3) to (ω1,ω2,θ), by similarity between the Γ coordinate system and the cylindrical system.
Based on (29) in the Γ coordinate system, the absolute value of the Jacobian factor can be calculated as follows:
|(∂k⃗∂ω1×∂k⃗∂ω2)·∂k⃗∂θ|=|(e⃗1×e⃗2)·(ω1ddθe⃗1+ω2ddθe⃗2)|=|e⃗3·(ω1ddθe⃗1+ω2ddθe⃗2)|=|ω1|,
where we have used the relations discussed in the next sectione⃗3·ddθe⃗1=-1,e⃗3·ddθe⃗2=0.
Then, a weight function W(ω1,ω2,θ) is defined for all the coordinates (ω1,ω2,θ), such that the summation of the weight function on the J(k⃗) groups of coordinates of the same k⃗∈R3 remains one, that is, ∑j=1J(k⃗)W(ω1j,ω2j,θj)=1.
An example of the weight functions is [1, 8]
W(ω1,ω2,θ)=sgn(k2)sgn(ω1),
with k2=k⃗·j⃗=[ω1e⃗1(θ)+ω2e⃗2(θ)]·j⃗.
With this weight function, (32) becomes a parallel-beam reconstruction formula from the Fourier slice theorem, and further Ye-Wang’s cone-beam formula [8] via the relationship between parallel-beam and cone-beam projections [1]. With the weight function, the filtration is no longer along the tangential direction generally.
Clearly, the weight function is not unique. For instance, for frequency points with J(k⃗)=3, in addition to the weight function defined by (36), (1-1+1), we can also choose (1/3+ 1/3+1/3), (0+1+0), or a more complicated one, (1+1-1) for some points and (-1+1+1) for the rest points, leading to different parallel-beam reconstruction formulas. Furthermore, using the approach described in [1], we can arrive at Katsevich’s helical cone-beam formulas [9–11], and so forth.
If one takes the weight function W(ω1,ω2,θ)=1 on the right side of (32), what we will reconstruct is actually another function Φ(r⃗)=∫0θ0∫-∞∞∫-∞∞exp(2πik⃗·r⃗)Ψ̂(k⃗)|ω1|dω1dω2dθ=∫R3exp(2πik⃗·r⃗)Ψ̂(k⃗)J(k⃗)d3k⃗.
From (37), we observe that the reconstructed image Φ(r⃗) equals the true object Ψ(r⃗) if and only if J(k⃗)=1, that is, the trajectory ΓAB is a great semicircle. The condition for Φ(r⃗) to be an approximation of the object function, that is, Φ(r⃗)≈Ψ(r⃗), is that J(k⃗)=1 in a major portion of the Fourier domain and J(k⃗) is close to 1 in the rest region. Based on the relation between J(k⃗) and the curve length of ΓAB, a necessary condition is θ0≈π, or equivalently ε=θ0/π-1≈0. Otherwise, Φ(r⃗) and Ψ(r⃗)will have a great difference. Unfortunately this formula was once thought as exact, which was essentially Theorem 3 in [2] and Theorem 4.3 in [3]. The details can be seen in [1].
Generally speaking, the criterion to identify exact formulas from approximate ones is to evaluate if every frequency component contributes exactly once to the final reconstruction, and the difference between exact formulas comes from different forms of weight functions.
4. Motion of the Frequency Plane and Condition for Sharp Points
In this section, we will study motion of the frequency plane, present the condition for sharp points of the locus of the rotation axis, and introduce the Euler angles. Let us simplify notations e⃗̇3=(d/dθ)e⃗3,e⃗̈3=(d2/dθ2)e⃗3, and so on. In this section, we assume that Γ is a three times differentiable curve on the unit ball, which may be not complete.
Obviously we have the following.
Proposition 1.
For any point S(θ)∈Γ, e⃗1,e⃗2,e⃗3 defined by (27) are orthogonal unit vectors.
Definition 1.
For every point S(θ)∈Γ, let us define
σ=e⃗3·(e⃗̇3×e⃗̈3),
which is the signed volume spanned by the three vectors e⃗3,e⃗̇3,e⃗̈3. If we consider the curve length θ as time, e⃗3,e⃗̇3,e⃗̈3 are the position, velocity, and the acceleration of point S.
Proposition 2.
The motion of e⃗1,e⃗2,e⃗3 can be described by the following differential equations
ddθe⃗3=e⃗1,ddθe⃗1=σe⃗2-e⃗3,ddθe⃗2=-σe⃗1.
Proof.
The first equation is the definition. Now, let us prove the other two. Since e⃗1,e⃗2,e⃗3 are orthogonal unit vectors, we can write
e⃗1·e⃗1=1,e⃗2·e⃗2=1,e⃗3·e⃗3=1,e⃗1·e⃗2=0,e⃗2·e⃗3=0,e⃗3·e⃗1=0.
Applying d/dθ on the above equations, one has
e⃗̇1·e⃗1=0,e⃗̇2·e⃗2=0,e⃗̇3·e⃗3=0,e⃗̇1·e⃗2+e⃗1·e⃗̇2=0,e⃗̇2·e⃗3=0,e⃗3·e⃗̇1=-1.
Using the definition of σ, we have
e⃗̇1·e⃗2=e⃗̇1·(e⃗3×e⃗1)=e⃗3·(e⃗̇3×e⃗̈3)=σ,e⃗1·e⃗̇2=-σ.
Using the orthogonal unit vectors e⃗1,e⃗2,e⃗3, e⃗̇1 and e⃗̇2 can be expressed as
ddθe⃗1=σe⃗2-e⃗3,ddθe⃗2=-σe⃗1,
which completes the proof.
From these differential equations, we recognize that the motion of e⃗1,e⃗2,e⃗3 is a rotation in the 3D space.
Now, we consider the frequency plane Π(OS⃗) fixed with e⃗1,e⃗2, e⃗3 as a rigid body. When S(θ) moves along the curve Γ, the rigid body rotates in the 3D space.
Let us define Ω⃗=e⃗2+σe⃗3.
We will see that Ω⃗ is the angular velocity of the rigid body by the following proposition.
Proposition 3.
The motion of the vectors e⃗1,e⃗2,e⃗3 can be expressed as
ddθe⃗3=Ω⃗×e⃗3,ddθe⃗1=Ω⃗×e⃗1,ddθe⃗2=Ω⃗×e⃗2.
Proof.
It is easy to verify the three equations one by one with Proposition 2 and the definition of Ω⃗.
Proposition 4.
The angular acceleration of the rigid body is
Ω⃗̇=σ̇e⃗3(θ).
Proof.
In fact, we have
Ω⃗̇=ddθ(e⃗2+σe⃗3)=e⃗̇2+σ̇e⃗3+σe⃗̇3=-σe⃗1+σ̇e⃗3+σe⃗1=σ̇e⃗3.
Proposition 5.
For any point S(θ)∈Γ, we have
σ=0⇔e⃗̈3=-e⃗3,σ̇=0⇔e⃗⃛3=-(σ2+1)e⃗̇3.
Proof.
Let us first prove (48).
If σ=0, by Proposition 2 we have e⃗̈3=-e⃗3. On the other hand, if e⃗̈3=-e⃗3, we have σ=e⃗3·[e⃗̇3×e⃗̈3]=0.
Now, let us prove (49).
By Proposition 2, we have
e⃗̇1=σe⃗2-e⃗3.
Applying d/dθ to both sides of (50) and using σ̇=0, we have
e⃗̈1=σ̇e⃗2+σe⃗̇2-e⃗̇3=σ(-σe⃗1)-e⃗1=-(σ2+1)e⃗1.
On the other hand, if e⃗⃛3=-(σ2+1)e⃗̇3, we have
σ̇=ddθ[e⃗3·(e⃗̇3×e⃗̈3)]=e⃗̇3·(e⃗̇3×e⃗̈3)+e⃗3·(e⃗̈3×e⃗̈3)+e⃗3·(e⃗̇3×e⃗⃛3)=e⃗3·(e⃗̇3×e⃗⃛3)=0.
This completes the proof. Physically, e⃗⃛3 is the change rate of the acceleration e⃗̈3.
Proposition 6.
σ=0 for every point S(θ)∈Γ, if and only if Γ is a great circle or a great arc.
Proof.
Since σ=0 for every point S(θ)∈Γ, the motion equation of the three unit vectors, (39), is simplified as
ddθe⃗3=e⃗1,ddθe⃗1=-e⃗3,ddθe⃗2=0.
Since (d/dθ)e⃗2=0 for every point, e⃗2 is a fixed unit vector. Since (d/dθ)e⃗3=e⃗1=e⃗2×e⃗3, e⃗3 rotates about the fixed axis e⃗2. Since e⃗3⊥e⃗2, S, the endpoint of e⃗3, draws a great circle or a great arc; see Figure 4(a).
On the other hand, if Γ is a great circle or a great arc, one can verify that e⃗̈3=-e⃗3. Therefore, we have σ=e⃗3·(e⃗̇3×e⃗̈3)=0. This completes the proof.
The circles on unit sphere and a helix in 3D space. (a) The illustration of σ=0 andσ̇=0, (b) a circular locus on the unit sphere and the motion of the frequency plane, and (c) the helix in 3D space and the motion of its TNB frame. The figure is not drawn to scale.
Proposition 7.
σ̇=0 for every point S(θ)∈Γ, if and only if Γ is a circle or a circular arc.
Proof.
Since σ̇=0, by Proposition 4 we have Ω⃗̇=σ̇e⃗3(θ)=0. We know that Ω⃗ is a fixed vector. By Proposition 3, we have
ddθe⃗3=Ω⃗×e⃗3.
Therefore, e⃗3 rotates about the fixed axis Ω⃗, and S, the endpoint of e⃗3, draws a circular arc. Furthermore, the radius of the circle is 1/1+σ2; see Figure 4(b).
On the other hand, if Γ is a circular arc, the line connecting the circle center and the origin O is the rotation axis. We know that e⃗̈3 is pointing to the axis. By circular symmetry, σ=e⃗3·(e⃗̇3×e⃗̈3) is constant at every point of the circular arc. Therefore, σ̇=0 on the circular arc. This completes the proof.
Clearly, Proposition 6 is a special case of Proposition 7; see Figure 4(a).
One important case is that σ=e⃗3·(e⃗̇3×e⃗̈3)=0, for a certain point S(θ)∈Γ. At such a moment, we have (d/dθ)e⃗2=-σe⃗1=0, the rotation axis of the frequency plane does not move, the point S moves along a great arc, and the Γ coordinate system is more similar to a cylindrical coordinate system. We say that θ is a stationary point of e⃗2 if σ(θ)=0. If σ has different signs before and after σ=0, the stationary point will become a sharp point, since in this case e⃗2 goes forward, stops and goes back according to (d/dθ)e⃗2=-σe⃗1. Therefore, we have the following straightforward proposition.
Proposition 8.
A necessary condition for the curve e⃗2(θ) to have a sharp point at θ1∈(0,θ0) is σ(θ1)=0; a sufficient condition for the curve e⃗2(θ) to have a sharp point at θ1∈(0,θ0) is that σ(θ) has different signs for θ<θ1 and for θ>θ1.
Based on the definition of σ and the equation of the osculating plane, σ(θ1)=0 means that the osculating plane at S(θ1)∈Γ passes through the origin since (e⃗3-0)·(e⃗̇3×e⃗̈3)=0.σ(θ) changes its sign before and after θ1 means that the osculating plane sweeps the origin, or relatively speaking, the origin goes from one side of the osculating plane to the other. An example can be seen in the next section. Sharp points serve as the landmarks for defining weight functions.
Proposition 9.
For any given point on the frequency plane k⃗=ω1e⃗1+ω2e⃗2, the motion equation is
ddθk⃗=Ω⃗×k⃗=-σω2e⃗1(θ)+σω1e⃗2(θ)-ω1e⃗3(θ).
Proof.
It is straightforward from Propositions 2 or 3.
We call (55) the motion equation of the frequency plane.
There are two applications of Proposition 9. First, it can be used to compute the Jacobian of the Γ coordinate system
|(∂k⃗∂ω1×∂k⃗∂ω2)·∂k⃗∂θ|=|e⃗3·∂k⃗∂θ|=|-ω1|=|ω1|.
Second, since
e⃗1·∂∂θk⃗=-σω2,e⃗2·∂∂θk⃗=σω1,
we know that among the three vectors ∂k⃗/∂ω1,∂k⃗/∂ω2,∂k⃗/∂θ, the last one is not orthogonal to the first two. Hence, a general Γ coordinate system is not orthogonal. Notice that ∂k⃗/∂θ=∂k⃗/∂θ|ω1,ω2=dk⃗/dθ
Note that the frequency plane rotates about the axis e⃗2 with the unit angular velocity while the vectors e⃗1 and e⃗2 themselves rotate about the axis e⃗3 with the angular velocity σ; see Figure 5(a). We see that the Γ coordinate system is different from the cylindrical coordinate system in the motion process. However, since the spin about the axis e⃗3 of the frequency plane does not have an effect on the scanned volume, their Jacobians have the same expression. Figure 5(b) shows the shape of a differential volume of the Γ coordinate system.
Differential motion of the frequency plane. (a) Two decomposed rotations and (b) one combined rotation.
Proposition 10.
When S moves along a curve Γ on the unit sphere, the scanned volume in the unit ball by the frequency plane is
V=4θ03=∫|k⃗|<1J(k⃗)d3k⃗,
where θ0 is the length of Γ, and J(k⃗) is the number of times point k⃗ is scanned.
Proof.
In fact, the scanned volume can be calculated in both the Γ and the Cartesian coordinate systems as follows:
V=∫0θ0∫k2=ω12+ω22<1|ω1|dω1dω2dθ=∫0θ0∫k=01∫ϕ=02πk2|cosϕ|dkdϕdθ=43∫0θ0dθ=43θ0=∫k12+k22+k32<1J(k⃗)d3k⃗.
This completes the proof.
This concise formula provides a basic relation between the curve length, the scanned volume by the frequency plane, and the J function.
When we study the rotation problem, Euler angles (ϑ,ϕ,δ) are helpful. As shown in Figure 6(a), (ϑ,ϕ) with ϑ∈(-π/2,π/2),ϕ∈(-∞,∞) are the spherical coordinate of S=S(ϑ,ϕ),δ is the angle between e⃗ϑ and e⃗2. e⃗ϕ, and e⃗ϑ are the latitude and longitude unit vectors at point S in the spherical coordinate system.
Euler angles. (a) The Euler angles of the frequency plane and (b) the rotation angle of frequency plane relative to the detector plane.
The Euler angles are functions of the curve length ϑ=ϑ(θ),ϕ=ϕ(θ),δ=δ(θ).
The three unit vectors can be represented as follows: e⃗3=e⃗3(ϑ,ϕ),e⃗1=sinδe⃗ϑ+cosδe⃗ϕ=ϑ̇e⃗ϑ+ϕ̇cosϑe⃗ϕ,e⃗2=-cosδe⃗ϑ+sinδe⃗ϕ=-ϕ̇cosϑe⃗ϑ+ϑ̇e⃗ϕ.
Furthermore, one can verify the motion equations of the three unit vectors in Proposition 2.
Based on the motion theory of rigid body, the angular velocity of a rigid body can be written as Ω⃗=-ϑ̇e⃗ϕ+ϕ̇k⃗+δ̇e⃗3=-ϑ̇e⃗ϕ+ϕ̇cosϑe⃗ϑ+(ϕ̇sinϑ+δ̇)e⃗3=e⃗2+σe⃗3.
Therefore, we have the following relationships:e⃗2=-ϑ̇e⃗ϕ+ϕ̇cosϑe⃗ϑ=Ω⃗0-(Ω⃗0·e⃗3)e⃗3,σ=ϕ̇sinϑ+δ̇=Ω⃗0·e⃗3-Ω⃗̇0·e⃗1,δ̇=-Ω⃗̇0·e⃗1=Ω⃗0·e⃗̇1,
where we have used a new notation Ω⃗0=-ϑ̇e⃗ϕ+ϕ̇k⃗,
which is the first two terms of the total angular velocity Ω⃗ of (62).
In fact, Ω⃗0 is the angular velocity of the detector with the unit vectors e⃗ϕ,e⃗ϑ when S moves along Γ. To be specific, we consider the detector for parallel-beam projection as a rectangle of finite width and length through the origin orthogonal to direction e⃗3=OS⃗. During any motion, one side of the detector always keeps horizontal, that is, parallel to the direction e⃗ϕ. The pose of the detector is completely determined by the point S=S(ϑ,ϕ), or two Euler's angles (ϑ,ϕ). The relative motion of the frequency plane to the detector is described by the angle δ; see Figure 6(b).
In summary, there are three sets of orthogonal vectors (beside i⃗,j⃗,k⃗, the unit vectors of the Cartesian coordinate system) and three important planes. The orthogonal unit vectors e⃗ϕ,e⃗ϑ, e⃗3 are related to the movement of the detector and e⃗3, e⃗̇3=e⃗1, e⃗2=e⃗3×e⃗1 are related to the movement of the frequency plane, see Figure 6(b). It is easy to verify that Ω⃗=e⃗2+σe⃗3=e⃗1×e⃗̇1 and hence the vectors e⃗1, e⃗̇1, Ω⃗ are the third set of orthogonal vectors. The direction of Ω⃗ is the normal of the osculating plane and its magnitude, 1+σ2, is the reciprocal of the radius of the osculating circle. Here the osculating plane is spanned by e⃗1, e⃗̇1 [15] and the osculating circle is referred to as the intersection between the osculating plane and the unit ball. When S moves along a curve Γ on the unit ball, it can be viewed that the point S moves along an osculating circle, whose orientation and radius are described by Ω⃗. From Ω⃗=e⃗2+σe⃗3, we can see that σ is an angular velocity of the spin of the frequency plane. The locus lengths of the endpoints of e⃗3,e⃗1, e⃗2 are θ0, ∫0θ01+σ2dθ, ∫0θ0|σ|dθ, respectively, according to Proposition 2.
5. Example of a Curve with J(k⃗)=1,3
Here, we discuss the reconstruction based on the simplified model of the 3D helical cone beam reconstruction, that is, 3D helical parallel-beam reconstruction.
The curve C in R3 defined by C={S′∈R3:OS′⃗=y⃗(φ)=(rcosφ,rsinφ,φh2π),φ∈[-π2,π2]}
is a segment of a helix with radius r and pitch h, which is shown in Figure 7 for r=1,h=3.
Motion process of the frequency plane. The red curve is a segment of a helical scanning trajectory. The blue curve is its projection on the unit sphere. The black curve is the great circle through S1,S3, and S5. Five positions of the frequency plane are color coded. The points inside the triangle are scanned 3 times while the rest points are scanned only once.
In fact, it is formed by a combination of a uniform circular motion in the XOY plane and a uniform straight motion along the Z direction.
The projection of the curve C on the unit sphere Ω is Γ={S∈Ω:OS⃗=e⃗3(φ)=y⃗(φ)|y⃗(φ)|,φ∈[-π2,π2]},
or Γ={S∈Ω:OS⃗=e⃗3(ϑ,ϕ),ϑ=tg-1(hϕ2π),ϕ∈[-π2,π2]}.
The curve length θ can be given by dθ=|de⃗3(φ)|=|d(y⃗(φ)|y⃗(φ)|)|,θ(φ)=∫-π/2φdθdφ′dφ′=∫-π/2φ|d(y⃗(φ′)|y⃗(φ′)|)|.
For example, we can calculate θ0=θ(π/2)≈3.177 and ε=0.011, when r=1 and h=3, indicating that the overlapping degree in the Fourier domain is small.
For every φ∈[-π/2,π/2], the three orthogonal unit vectors aree⃗3=OS⃗=y⃗(φ)|y⃗(φ)|,e⃗1=(d/dφ)e⃗3|(d/dφ)e⃗3|,e⃗2=e⃗3×e⃗1.
The inverse Fourier transform can be expressed with either the Γ coordinates (ω1,ω2,θ) or the new coordinates (ω1,ω2,φ)Ψ(r⃗)=∫0θ0∫-∞∞∫-∞∞W(ω1,ω2,θ)exp(2πik⃗·r⃗)Ψ̂(k⃗)×|ω1|dω1dω2dθ=∫-π/2π/2∫-∞∞∫-∞∞W(ω1,ω2,θ(φ))exp(2πik⃗·r⃗)×Ψ̂(k⃗)|ω1|dθdφdω1dω2dφ,
with k⃗=k⃗(ω1,ω2,θ(φ))=ω1e⃗1(θ(φ))+ω2e⃗2(θ(φ)).
To visualize the motion of the frequency plane ∏(OS⃗), its five positions are marked in Figure 7 at φ1=-π/2, φ2=-π/4, φ3=0,φ4=π/4 and φ5=π/2, respectively. Note that the initial position Π(OS1⃗) and the final position Π(OS5⃗) coincide with each other.
In Figure 7, the two curved sides of the red triangle on the top of the unit sphere is the locus of e⃗3, the instantaneous rotation axis of the frequency plane. Similarly, -e⃗2 and Π(OS1⃗) form the counterpart triangle on the bottom of the unit sphere. The locus of e⃗2 is the division line of the regions with different J(k⃗) values. Based on (58), the total area of the two triangles is only 0.55% of the surface area of the unit sphere.
Note that a sharp point on the locus of e⃗2 appears at φ=φ3=0. The reason is as follows.
Let us defineσ0(φ)=y⃗(φ)·(ddφy⃗(φ)×d2dφ2y⃗(φ)).
It can be verified that σ0(0)=0, σ0(φ)<0 for φ<0, and σ0(φ)>0 for φ>0.
It is not difficult to prove
σ=e⃗3·(e⃗1×e⃗̇1)=y⃗(φ)·(ddφy⃗(φ)×d2dφ2y⃗(φ))1|y⃗(φ)|×1|(d/dφ)y⃗(φ)|2(dφdθ)3=σ0(φ)K(φ),
where K(φ) is positive. Therefore, σ and σ0 have the same zeros and sign.
Based on the motion process of Π(OS⃗), we have J(k⃗)={1,k⃗outsidethetwotriangles,3,k⃗insidethetwotriangles.
The overlapped region in the frequency space is due to the motion of the instantaneous rotation axis e⃗2 of the frequency plane, or equivalently the discrepancy between the curve Γ and half a great circle.
As shown in Figure 8, the weight factor can be (1-1+1), (1/3+ 1/3+1/3), or (0+1+0), and so forth, for the overlapped region. By dividing the overlapped region into 2 symmetric parts, one can define the weight function as (-1+1+1) on the left side and (1+1-1) on the right side, or (1-1+1) on the left and (1+1-1) on the right, and so forth. Hence, one can design various parallel-beam reconstruction formulas and derive the associated cone-beam formulas using the approach described in [1].
Examples of possible weight factors. (a) 1-1+1, (b) 1/3+1/3+1/3, (c) 0+1+0, and (d) -1+1+1 for the left half, 1+1-1 for the right half, (e) 1-1+1 for the left half, and 1+1-1 for the right half. The straight line represents a great circle on the unit sphere. The weight factor in (a) for ∏(OS2⃗) is further illustrated in (f) as an example.
AppendicesA. Weighted Λ Reconstruction Formula
f(x,y) is a function in the 2D space with the Fourier transform f̂(k1,k2)=∫-∞∞∫-∞∞f(x,y)exp(-i2πk1x)exp(-i2πk2y)dxdy.
Its parallel-beam projection is denoted as Pϑ(l)=∫-∞∞f(lcosϑ-vsinϑ,lsinϑ+vcosϑ)dv
with l∈(-∞,∞),ϑ∈[-π/2,π/2].
Let f(x,y) pass through the filter |k1|, we get a new function Φ(x,y)=∫-∞∞∫-∞∞|k1|f̂(k1,k2)exp(i2πk1x)exp(i2πk2y)dk1dk2=f(x,y)*1-2π2x2.
Defining the signed polar coordinates (kρ,ϑ) with kρ∈(-∞,∞),ϑ∈(-π/2,π/2), which satisfy k1=kρcosϑ,k2=kρsinϑ, we haveΦ(x,y)=∫-π/2π/2∫-∞∞|k1|f̂(kρ,ϑ)×exp[-i2πkρ(xcosϑ+ysinϑ)]|kρ|dkρdϑ=∫-π/2π/2∫-∞∞f̂(kρ,ϑ)exp[-i2πkρ(xcosϑ+ysinϑ)]×kρ2cosϑdkρdϑ=1-4π2∫-π/2π/2∂2∂l2|l=xcosϑ+ysinϑ∫-∞∞f̂(kρ,ϑ)×exp(-i2πkρl)dkρcosϑdϑ=1-4π2∫-π/2π/2∂2∂l2|l=xcosϑ+ysinϑPϑ(l)cosϑdϑ.
Comparing (A.3) and (A.4), we havef(x,y)*1-2π2x2=1-4π2∫-π/2π/2∂2∂l2|l=xcosϑ+ysinϑPϑ(l)cosϑdϑ.
That is, we have rediscovered the identity (20) in a general way. In contrast to the classical Λ reconstruction formula [1], a weight factor cosϑ is involved in (A.5). Hence, we can call (A.5) a weighted Λ reconstruction formula.
B. A Weighted Radon Formula and the Orlov Formula
Let us define u=r⃗·e⃗3,t=r⃗·e⃗1,s=r⃗·e⃗2,
and the parallel-beam projection is Pθ(t,s)=∫-∞∞Ψ(ue⃗3+te⃗1+se⃗2)du,
as shown in Figure 6(b). In the frequency plane Π(OS⃗), we define a signed polar coordinate system (ω,α) with ω∈(-∞,∞),α∈[0,π] for a point (ω1,ω2) with ω1=ωcosα,ω2=ωcosα. Similarly, let us define the signed polar coordinates (X,β) with X∈(-∞,∞),β∈[-π/2,π/2] for a point (t,s) with t=Xcosβ,s=Xsinβ.
Let the weight function in (32) be W(ω1,ω2,θ)=W(α,θ), we have Ψ(r⃗)=∫0θ0∫-∞∞∫-∞∞W(ω1,ω2,α)exp(2πik⃗·r⃗)Ψ̂(k⃗)×|ω1|dω1dω2dθ=∫0θ0∫0π∫-∞∞W(α,θ)|cosα|exp(2πiω1t+2πiω2s)×Ψ̂(ω1,ω2,θ)ω2dωdαdθ=∫0θ0∫0πW(α,θ)|cosα|∫-∞∞ω2exp(2πiω(tcosα+ssinα))×Ψ̂(ω,α,θ)dωdαdθ=-14π2∫0θ0∫0πW(α,θ)∂2∂l2|l=r⃗·n⃗=tcosα+ssinαRΨ(l,n⃗)×|cosα|dαdθ=-14π2∇2∫0θ0∫0πW(α,θ)RΨ(tcosα+ssinα,n⃗)×|cosα|dαdθ,
with n⃗=e⃗1cosα+e⃗2sinα.
We call it the weighted Radon formula. When ΓAB is a great semicircle, we have W(α,θ)=1 and θ0=π, and the above formula becomes the classic Radon formula [12, 13].
Moreover, it is easy to rederive Orlov's formula [14] from (32) Ψ(r⃗)=∫0θ0∫-∞∞∫-∞∞W(ω1,ω2,θ)exp(2πik⃗·r⃗)Ψ̂(k⃗)×|ω1|dω1dω2dθ=∫0θ0∫-∞∞∫-∞∞W(α,θ)exp(2πiω1t+2πiω2s)×Ψ̂(ω1,ω2,θ)|ω1|dω1dω2dθ=∫0θ0∫-∞∞∫-∞∞-14π2(∂2∂t2+∂2∂s2)W(α,θ)×exp(2πiω1t+2πiω2s)Ψ̂(ω1,ω2,θ)|ω1|ω2dω1dω2dθ=-14π2∇2∫0θ0∫-∞∞∫-∞∞|cosα||ω|W(α,θ)·Ψ̂(ω1,ω2,θ)×exp(2πiω1t+2πiω2s)dω1dω2dθ=-14π2∇2∫0θ0pθ(t,s)**gθ(t,s)dθ,
where gθ(t,s)=∫-∞∞∫-∞∞|cosα||ω|W(α,θ)exp(2πiω1t+2πiω2s)dω1dω2=|cosα||X|W(α,θ)|α=π/2+β.
The sign ** denotes the 2D convolution operation.
We recall that a frequency point has J(k⃗) sets of Γ coordinates k⃗=k⃗(ω1,1ω2,1θ1)=k⃗(ω1,2ω22,θ2)=⋯=k⃗(ω1,J(k⃗)ω2,J(k⃗)θJ(k⃗)).
If a given triplet (ω1,ω2,θ) is the jth set of Γ coordinates of k⃗,
(ω1,ω2,θ)=(ω1,jω2j,θj),
a new normalized weight function [1] for the triplet can be defined by W(ω1,ω2,θ)=W(α,θ)=1/|k⃗·e⃗1(θj)|∑j=1J(k⃗)1/|k⃗·e⃗1(θj)|=1/|cosα|∑j=1J(k⃗0)1/|k⃗0·e⃗1(θj)|,
and in this case we have gθ(t,s)=1|X|1∑j=1J(k⃗0)1/|k⃗0·e⃗1(θj)||k⃗0=e⃗3×X⃗0.
Here, k⃗0=k⃗/|k⃗|, and X⃗0=(te⃗1+se⃗2)/t2+s2=e⃗1cosβ+e⃗2sinβ are unit vectors.
Equation (B.4) with (B.9) is the Orlov formula for a general curve on the unit sphere ((2a) with (2c) in [14]). Note that in [14] different letters were used such as m↔k⃗0,τ↔e⃗3,τ̇↔e⃗1,x′↔(te⃗1+se⃗2),|x′|↔|X|,∫GdGτ↔∫0θ0dθ, and so forth, and τ̇k′ should be printed as τ̇k.
C. Link between the Motion of the Frequency Plane and Differential Geometry
To study the property of a curve Γ on the unit sphere and the associated motion of the frequency plane, one can introduce another curve, Γ̃, defined as r⃗(θ)=∫0θe⃗3(θ′)dθ′+r⃗(0),
where r⃗(0) is an arbitrary vector in the 3D space, and θ (as well as θ′) is an arc-length parameter of the curve Γ. Clearly,
r⃗̇=ddθr⃗(θ)=e⃗3,r⃗̈=e⃗1,r⃗⃛=e⃗̇1.
Suppose that the arc-length parameter of the curve Γ̃ is s. The tangential direction of the curve Γ̃ is T⃗=ddsr⃗=dθdsddθr⃗(θ)=dθdse⃗3(θ).
Furthermore, we have |ddsr⃗|=|dθds||e⃗3(θ)|,1=|dθds|,
where |(d/ds)r⃗|=|dr⃗/ds|=|dr⃗|/ds=ds/ds=1 has been used.
When the positive directions are the same, we have dθds=1,θ=s+s0,T⃗=e⃗3(θ).
Picking up the constant s0=0, we have θ=s. Therefore, the symbol θ represents the arc length parameters of both Γ̃ and Γ. Based on the tangential direction T⃗, the norm and binormal unit vectors [15] can be defined as N⃗=T⃗̇|T⃗̇|=e⃗1|e⃗1|=e⃗1,B⃗=T⃗×N⃗=e⃗3×e⃗1=e⃗2.
The curvature κ and the torsion τ of the curve Γ̃ are κ=|r⃗̈|=|e⃗1|=1,τ=r⃗̇·(r⃗̈×r⃗⃛)=e⃗3·(e⃗1×e⃗̇1)=σ,
and the Darboux vector is Ω⃗=τT⃗+κB⃗=e⃗2+σe⃗3.
The Frenet-Serret Equations of the curve Γ̃T⃗̇=N⃗,N⃗̇=-T⃗+σB⃗,B⃗̇=-σN⃗,
and
T⃗̇=Ω⃗×T⃗,N⃗̇=Ω⃗×N⃗,B⃗̇=Ω⃗×B⃗,
are (39) and (45). In this frame, r⃗(θ) is the original curve (function), and e⃗3(θ) becomes its derived curve (function). In practice, it is difficult to find the physical meaning of the 3D curve r⃗(θ).
Example C.
A circle on the unit sphere and its corresponding curve in the 3D space.
As shown in Figure 4(b), a circle on the unit sphere, marked as C, can be described as e⃗3=sinϑ0k⃗+cosϑ0cosϕi⃗+cosϑ0sinϕj⃗
with a fixed parameter ϑ0∈(-π/2,+π/2) and a variable ϕ∈[0,2π].
Using the arc-length parameter θ∈[0,2πcosϑ0], we have e⃗3(θ)=sinϑ0k⃗+cosϑ0cosθcosϑ0i⃗+cosϑ0sinθcosϑ0j⃗.
Furthermore, we have e⃗1(θ)=-sinθcosϑ0i⃗+cosθcosϑ0j⃗,e⃗̇1(θ)=1cosϑ0cosθcosϑ0i⃗-1cosϑ0sinθcosϑ0j⃗,σ=e⃗3·(e⃗1×e⃗̇1)=tgϑ0,Ω⃗=e⃗2+σe⃗3=1cosϑ0k⃗.
Based on the definition (C.1), the corresponding 3D curve C̃ is
r⃗(θ)=θsinϑ0k⃗+cos2ϑ0sinθcosϑ0i⃗-cos2ϑ0cosθcosϑ0j⃗,
with the choice r⃗(0)=-j⃗cos2ϑ0.
Now, we see that C̃ is a helix, whose standard form is r⃗(ϕ)=bϕk⃗+acos(ϕ-π2)i⃗+asin(ϕ-π2)j⃗,
with a=cos2ϑ0,b=sinϑ0cosϑ0,ϕ∈[0,2π] (see Figure 4(c)). Its curvature and torsion are κ=1,τ=tgϑ0.
The length of the helix C̃ for ϕ∈[0,2π] is 2πa2+b2=2πcosϑ0, which is equal to the circumference of circle C on the unit sphere. The radius of the helix C̃(a=cos2ϑ0) is smaller than that of the circle C(cosϑ0). The figure is not drawn to scale. The TNB frame rotates around the z direction at a constant velocity Ω⃗=2π2πcosϑ0k⃗=1cosϑ0k⃗.
Acknowledgment
The works of H. Yu and G. Wang are partially supported by the NSF/MRI Program (no. CMMI-0923297) and NIH/NIBIB Grant (no. EB011785). The authors also thank a reviewer for a valuable suggestion.
WeiY.YuH.HsiehJ.CongW.WangG.Scheme of computed tomography20071542352702-s2.0-37549028055PalamodovV. P.Reconstruction from ray integrals with sources on a curve20042012392422-s2.0-1342287233PalamodovV. P.2004Boston, Mass, USABirhauserYuH.YeY.ZhaoS.WangG.Studies on Palamodov's algorithm for cone-beam CT along a general curve20062224474602-s2.0-3364527519110.1088/0266-5611/22/2/004YuH.YeY.ZhaoS.WangG.Reply to the comment on 'studies on Palamodov's algorithm for cone-beam CT along a general curve'2006224150515062-s2.0-3374631731610.1088/0266-5611/22/4/N02RammA. G.KatsevichA. I.1996New York, NY, USACRC PressGel'fandI. M.ShilovG. E.19771New York, NY, USAAcademic PressYangboY.WangG.Filtered backprojection formula for exact image reconstruction from cone-beam data along a general scanning curve200532142482-s2.0-1284425631210.1118/1.1828673KatsevichA.A general scheme for constructing inversion algorithms for cone beam CT2003200321130513212-s2.0-004290371010.1155/S0161171203209315KatsevichA.Theoretically exact filtered backprojection-type inversion algorithm for spiral CT2002626201220262-s2.0-003666522410.1137/S0036139901387186KatsevichA.Improved exact FBP algorithm for spiral CT2004324681697RadonJ.Über die Bestimmung von Funktionen durch ihre Integralwerte längs gewisser Mannigfaltigkeiten1917692672772-s2.0-0022880440ParksP. C.On the determination of functions from their integral values along certain manifolds1986541701762-s2.0-0022880440OrlovS.Theory of three dimensional reconstruction: II. The recovery operator19752044294332-s2.0-0022880440PressleyA. N.2001New York, NY, USASpringerSpringer Undergraduate Mathematics Series