Tomodensitometry image acquisition and reconstruction process and a tomodensitometer using this process

Claim 2

A process for image reconstruction by tomodensitometry of the type including the steps of rotating a radiation source and detectors assembly in the plane of a cross-section the image of which is to be displayed, said radiation source emitting a beam of penetrating radiation in the form of a fan, and taking and storing a succession of image views for different angles of rotation of the source and detectors assembly with respect to a fixed reference axis, of said plane, the detectors of said assembly being arranged to converge towards a focal point of said radiation source and each view being sampled for angles γ representative of the detectors and indexed from a straight line joining the center of rotation of said radiation source and detectors assembly to said focal point, the angular shift between two adjacent detectors being constant Δγ and the angular shift between views being constant; with the angular shift between said straight line and the detector the nearest thereto adjusted to {Δγ/4| and the following conditions being provided: π/Δγe=N, with N a positive integer, assuming γe=(1+α )γ, where α is a coefficient dependent on the speed of rotation of the source and detectors assembly and on the time for acquisition of a view, said process further comprising the steps of: taking and storing two series of views (Vi, Wi) with an angular shift between views of hΔβ for the first series (Vi) and Δβ for the second series (Wi), at least some of the views of the two series being shifted with respect to each other by |Δγe/2| with Δβ=2Δγe, h being a positive interger, completing the views of the first series by sampled values of the second series whose coordinates (θ',γ') located in the space (θ,γe) with θ=β+γe, satisfy the relationship: (θx,γx(=(θ'+(2k+1)π, -γ') with k a relative integer, (θx, γ'x) being the coordinates in space (θ,γe) of desired sampled values of the first series; and applying a reconstruction algorithm, comprising a convolution and back projection, to the completed views of the first series for reconstituting an image.

Claim 3

The process as claimed in claim 1, wherein the views (V i ) of a series are taken and stored during a revolution of the source-detectors assembly and the views (W i ) of the other series during the next revolution and a shift of |Δγe/2| is caused in the reading of the two series, between the two revolutions.

Claim 4

The process as claimed in claim 1, wherein the views (V i , W i ) of the two series are taken and stored alternately during the same revolution.

Claim 5

The process as claimed in claim 2 or 3, wherein, with h equal to 1, N views of the first series are completed by N views of the second series.

Claim 6

The process as claimed in claim 2 or 3, wherein, with h greater than 1, only one sampled value out of h is taken and/or processed in each views of the second series.

Claim 7

The process as claimed in claim 1 wherein γ e =γ.

Claim 8

The process as claimed in claim 1 for continuous mode tomodensitometry, wherein α=|v.τ/Δγ|, v being the speed of rotation of the source-detector assembly and τ being the time interval separating the acquisition of the sampled values corresponding to two adjacent detectors.

Claim 9

A tomodensitometer of the type comprising a radiation source and detectors assembly mounted for rotation in the plane of a cross-section the image of which is to be formed, said radiation source being a source of penetrating radiation in the form of a fan, said detectors being arranged so as to converge towards the focal point of said source, said detectors being seen from said focal point at respective angles γ indexed from a straight line joining said focal point to the center of rotation of the source and detectors assembly, two adjacent detectors being angularly shifted by Δγ, control means for determining a certain number of view taking angles β of the source and detector assembly with respect to a reference axis system and for generating signals for controlling said taking of views, means for storing information representative of said sampled views for said angles β, computing means for applying a reconstruction algorithm, comprising a convolution and a back projection to a series of views, wherein: with the following condition satisfied: π/Δγe=N with N a positive integer, assuming γ e =(1+α)γ, where α is a coefficient dependent on the speed of rotation of the source-detectors assembly and on the time for acquisition of a view, an angular shift between said straight line (OF) and the detector the nearest thereto is predetermined at |Δγ/4| said control means comprises means for taking and storing two series (Vi, Wi) of sampled views, said series being such that at least some adjacent views of different series are shifted therebetween by |Δγe/2|; the views of the first series being sifted angularly at the pitch of hΔβ and those of the second series being shifted angularly at the pitch of Δβ, with Δβ=2Δγ and h a positive integer, and rearrangement means for completing the views of the first series by sampled values taken from the second series whose coordinates (θ', γ') in the space (θ, γ e ) with θ=β+γ e satisfy the relationship (θx, γx)=(θ+(2k+1)π, -γ') with k a relative integer, (θx, γx) being the coordinates in the space (θ,γ) of desired sampled values of the first series.

Claim 10

A tomodensitometer of the type comprising a radiation source and detectors assembly mounted for rotation in the plane of a cross-section the image of which is to be formed, said radiation source being a source of penetrating radiation in the form of a fan, said detectors being arranged so as to converge towards the focal point of said source, said detectors being seen from said focal point at respective angles γ indexed from a straight line joining said focal point to the center of rotation of the source and detectors assembly, two adjacent detectors being angularly shifted by Δγ, control means for determining a certain number of view taking angles β of the source and detector assembly with respect to a reference axis system and for generating signals for controlling said taking of views, means for storing information representative of said sampled views for said angles β, computing means for applying a reconstruction algorithm, comprising a convolution and a back projection to a series of views, wherein: with the following condition satisfied: π/Δγe=N with N a positive integer, assuming γ e =(1+α)γ, where α is a coefficient dependent on the speed of rotation of the source-detectors assembly and on the time for acquisition of a view, an angular shift between said straight line (OF) and the detector the nearest thereto is predetermined at |Δγ/4| said control means comprises means for taking and storing two series (Vi, Wi) of sampled views, said series being such that at least some adjacent views of different series are shifted therebetween by |Δγe/2|; the views of the first series being sifted angularly at the pitch of hΔβ and those of the second series being shifted angularly at the pitch of Δβ, with Δβ=2Δγ and h a positive integer, and rearrangement means for completing the views of the first series by sampled values taken from the second series whose coordinates (θ', γ') in the space (θ, γ e ) with θ=β+γ e satisfy the relationship (θx, γx)=(θ+(2k+1)π, -γ') with k a relative integer, (θx, γx) being the coordinates in the space (θ,γ) of desired sampled values of the first series.

Claim 11

The tomodensitometer as claimed in claim 8, wherein said control means comprise an incremental coder.

Claim 12

The tomodensitometer as claimed in claim 9, wherein the elementary counting pitch of said incremental coder is Δγe/2, or a submultiple of this value.

Claim 13

The tomodensitometer as claimed in one of claims 8 to 10, of the type with pulsed mode operation, wherein γ e =γ.

Claim 14

The tomodensitometer as claimed in claim 8, wherein α=Vτ/Δγ, v being the speed of rotation of the source-detector assembly and τ being the time interval separating the acquisition of the sampled values corresponding to two adjacent detectors.