Patent US5265012A successfully refreshed from Google Patents.
US5265012A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method to determine a space from a known discrete space for the reconstruction of bidimensional or tridimensional images, as well as a device to implement and apply the method
Assignee: Commissariat à l'Énergie Atomique et Aux Énergies Alternatives
•
Filing: 1990-06-07
•
Granted: 1993-11-23
•
Expires: 2010-06-07
Abstract
Abstract The invention concerns a method and a device to determine a space from a known discrete space for the reconstruction of bidimensional or tridimensional images. In order to achieve this, the method comprises an iterative elementary phase whose aim is to determine the contribution to the value of an element of the unknown space on the basis of the value of the element (and possibly from its vicinity) of the known space by controlling the following successive stages: for predetermining parameters (V1 to V4); for the index determining one point of the known space by a generalized relation including the main acquisition devices V1+eV2/V3+eV4; for determining contributions by interpolation on adjacent points; for weighting contributions, and for adding the weighted contribution to the value of the element of the unknown space during processing. Application for medical optical image formation and for non-destructive industrial controls.
Inventors
- 👤 Jean-Louis Amans
- 👤 Christophe Caquineau
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-24 21:18
CPC Classifications
G; G06; G06T; G06T12/00; G06T12/20; G06T2211/00; G06T2211/40; G06T2211/424
Claims Analysis
Total Claims: 20
Claim 1
Independent Claim
500 words
Method to determine data of an unknown discrete space from data of a known discrete space by an operation of reprojection or retroprojection, this operation being one of the steps allowing the reconstruction of a bidimensional or trimensional image of an analyzed body, the data of the known space coming from data of projections of the body obtained by means of a defined geometry acquisition device comprising a radiation source and a set of detectors, eventually processed by other steps of the process of the reconstruction, the data of each of the space being defined respectively by indices corresponding to the coordinates along three axes of the space, this determination of the unknown space being obtained by calculating the numerical value of each point (or data) of the unknown space on the basis of a determination of the contributions brought by the numerical values of the corresponding points (or data) of the known space, wherein said method consists of: decomposing each space into one monodimensional set of points along given axes; placing in correspondence successively for each monodimensional set of one of the spaces, the monodimensional sets of the other space; realizing for each correspondence between one monodimensional set of points of the unknown space borne by an axis E of this space and one monodimensional set of points of the known space borne by an axis F of this space, one elementary iterative phase for each point of the monodimensional set of the unknown space, this elementary phase iteratively comprising for each point the following stages: determining the index associated with the coordinate on the axis F of the corresponding point of the known space by taking the entire part ADR from the generalized expression a 0 (e) with: ##EQU18## e being the iteration index corresponding to the coordinate on the axis E of the relevant point of the unknown space and V1, V2, V3 and V4 being constant parameters in the elementary phase, all these parameters being determined prior to this phase according to the selected acquisition geometry, the type of operation (reprojection or retroprojection) and the monodimensional sets placed in correspondence; determining the contribution C made to the numerical value of the relevant point of the unknown space by interpolation on the numerical values of the corresponding point of the index ADR and possibly of adjacent points; weighting this contribution by a factor f determined by the following equation: ##EQU19## V5, V6 and V7 being constant parameters in the elementary phase, all these parameters being determined prior to this phase according to the selected acquisition geometry, the type of operation and the monodimensional units in question; adding this weighted contribution to the numerical value of the relevant point of the unknown space determined during the preceding elementary phases having placed in correspondence the point of the unknown space with points of other known monodimensional spaces, all the contributions obtained at the end of the method to determine the unknown space constituting the final value of this point.
Claim 2
Dependent Claim
500 words
Method to determine data of an unknown discrete space from data of a known discrete space by an operation of reprojection or retroprojection, this operation being one of the steps allowing the reconstruction of a bidimensional or trimensional image of an analyzed body, the data of the known space coming from data of projections of the body obtained by means of a defined geometry acquisition device comprising a radiation source and a set of detectors, eventually processed by other steps of the process of the reconstruction, the data of each of the space being defined respectively by indices corresponding to the coordinates along three axes of the space, this determination of the unknown space being obtained by calculating the numerical value of each point (or data) of the unknown space on the basis of a determination of the contributions brought by the numerical values of the corresponding points (or data) of the known space, wherein said method consists of: decomposing each space into one monodimensional set of points along given axes; placing in correspondence successively for each monodimensional set of one of the spaces, the monodimensional sets of the other space; realizing for each correspondence between one monodimensional set of points of the unknown space borne by an axis E of this space and one monodimensional set of points of the known space borne by an axis F of this space, one elementary iterative phase for each point of the monodimensional set of the unknown space, this elementary phase iteratively comprising for each point the following stages: determining the index associated with the coordinate on the axis F of the corresponding point of the known space by taking the entire part ADR from the generalized expression a 0 (e) with: ##EQU18## e being the iteration index corresponding to the coordinate on the axis E of the relevant point of the unknown space and V1, V2, V3 and V4 being constant parameters in the elementary phase, all these parameters being determined prior to this phase according to the selected acquisition geometry, the type of operation (reprojection or retroprojection) and the monodimensional sets placed in correspondence; determining the contribution C made to the numerical value of the relevant point of the unknown space by interpolation on the numerical values of the corresponding point of the index ADR and possibly of adjacent points; weighting this contribution by a factor f determined by the following equation: ##EQU19## V5, V6 and V7 being constant parameters in the elementary phase, all these parameters being determined prior to this phase according to the selected acquisition geometry, the type of operation and the monodimensional units in question; adding this weighted contribution to the numerical value of the relevant point of the unknown space determined during the preceding elementary phases having placed in correspondence the point of the unknown space with points of other known monodimensional spaces, all the contributions obtained at the end of the method to determine the unknown space constituting the final value of this point.
Claim 3
Dependent Claim
197 words
Method to determine a space from a known discrete space according to claim 1, wherein the elementary phase represents a parallel geometry retroprojection operation, the known space being a set of projection data indexed by l, m and n, m and n being fixed, the unknown space being the image space indexed by i, j and k (j and k) or (i and k) being fixed; the sought-after index a 0 (e) is either the index l 0 (j) when i and k and fixed, the axis E being indexed by j and the axis F by l, or the index l 0 (i) when j and k are fixed, the axis E being indexed by i and the axis F by l, the following then being selected: V3=1 V4=0 and for i fixed at i 0 , k=0, m=0, n=n 0 , l 0 (j) being sought V1=L.sub.0 +i.sub.0 ΔI V2=ΔJ e=j or for j fixed at j 0 , k=0, m=0, n=n 0 , l 0 (i) being sought V1=L.sub.0 +j.sub.0 ΔJ V2=ΔI e=i the terms L 0 , ΔI, ΔJ being predetermined, l 0 being a real whole number of the entire part 1.
Claim 4
Dependent Claim
61 words
Method to determine data of an unknown discrete space from data of a known discrete space according to claim 2, wherein a conical and parallel geometry in the case where the elementary phase represents a retroprojection and for a parallel geometry when this phase represents a reprojection, the weighting factor f is determined by taking: V7=0 V6=1 and V5 is predetermined.
Claim 5
Dependent Claim
195 words
Method to determine a space from a known discrete space according to claim 1, wherein the elementary phase represents a parallel geometry reprojection phase, the known space being the image space indexed by l, m and n, (m and n) or (l and m) being fixed, the unknown space being all the projection data indexed i, j and k, j and k being fixed; the index a 0 (e) is the index l 0 (i) or n 0 (i), the axis E is then indexed by i and the axis F is indexed by l or n and the following is selected: m fixed at 0, n at n 0 , j at j 0 , k at 0, l 0 (i) is sought V3=1 V4=0 V1=h.sub.1 (n.sub.0,j.sub.0) V2=ΔL.sub.1 (j.sub.0) e=i or m fixed at 0, l at l 0 , j at j 0 , k at 0 and n 0 (i) is sought V3=1 V4=0 V1=h.sub.2 (l.sub.0 j.sub.0) V2=ΔL.sub.2 (j.sub.0) e=i the terms h 1 , h 2 , ΔL 1 and ΔL 2 being predetermined and n 0 and l 0 are respectively real numbers of the entire number n and l.
Claim 6
Dependent Claim
61 words
Method to determine data of an unknown discrete space from data of a known discrete space according to claim 4, wherein a conical and parallel geometry in the case where the elementary phase represents a retroprojection and for a parallel geometry when this phase represents a reprojection, the weighting factor f is determined by taking: V7=0 V6=V1 and V5 is predetermined.
Claim 7
Dependent Claim
188 words
Method to determine a space from a known discrete space according to claim 1, wherein the elementary phase represents a fan-shaped geometry retroprojection operation with constant linear pitches, the known space being all the projection data indexed by l, m, n; m and n are fixed, the unknown space being the image space indexed by i, j, k (i and k) or (j and k) being fixed, the sought-after index a 0 (e) being either the index l 0 (i) or l 0 (j), the axis E being indexed by i or by j and the axis F by l, the following being selected: k=0, m=0, n=n 0 , j=j 0 , l 0 (i) being sought, V1=h.sub.3 (j.sub.0,n.sub.0) V2=h.sub.4 (n.sub.0) V3=h.sub.5 (j.sub.0,n.sub.0) V4=h.sub.6 (n.sub.0) e=i or: k=0, m=0, n=n 0 , i=i 0 , l 0 (j) being sought V1=h'.sub.3 (i.sub.0,n.sub.0) V2=h'.sub.4 (n.sub.0) V3=h'.sub.5 (i.sub.0,n.sub.0) V4=h'.sub.6 (n.sub.0) e=j. the terms h 3 , h 4 , h 5 , h 6 , h' 3 , h' 4 , h' 5 , h' 6 being predetermined and l 0 is a real number of the entire part l.
Claim 8
Dependent Claim
57 words
Method to determine data of an unknown discrete space from data of a known discrete space according to claim 6, wherein in the weighting stage for a fan-shaped geometry in the case of a retroprojection, the weighting factor f is determined by taking: V5 predetermined on the basis of the parameters of the acquisition device V6=V3 V7=V4.
Claim 9
Dependent Claim
217 words
Method to determine a space from a known discrete space according to claim 1, wherein the elementary phase represents a constant linear pitch fan-shaped geometry reprojection operation, the known space being the image space indexed by l , m, n; (m and n) or (l and m) being fixed, the unknown space being all the projection data indexed by i, j, k, j and k being fixed, the sought-after index a 0 (e) being the index l 0 (i) or n 0 (i), the axis E being indexed by i and the axis F by l or n and the following is then selected: m fixed at 0, n at n 0 , j at j 0 , k at 0, l 0 (i) being sought, V1=h.sub.7 (n.sub.0, j.sub.0) V2=h.sub.8 (n.sub.0, j.sub.0) V3=h.sub.9 (j.sub.0) V4=h.sub.10 (j.sub.0) e=i or =m fixed at 0, l at l 0 , j at j 0 , k at 0, n 0 (i) being sought, V1=h'.sub.7 (l.sub.0, j.sub.0) V2=h'.sub.8 (l.sub.0, j.sub.0) V3=h'.sub.9 (j.sub.0) V3: h'.sub.10 (j.sub.0) e=i the terms h 7 , h 8 , h 9 , h 10 , h' 7 , h' 8 , h' 9 , h' 10 , being predetermined and n 0 , l 0 respectively being real numbers of the entire part n and l.
Claim 10
Dependent Claim
47 words
Method to determine data of an unknown discrete space from data of a known discrete space according to claim 8, wherein in the weighting stage for a fan-shaped and conical geometry in the case of reprojection, the weighting factor f is determined by taking: V5=1 V6=1 V7=0.
Claim 11
Dependent Claim
123 words
Method to determine a space from a known discrete space according to claim 1, wherein the elementary phase represents a conical geometry retroprojection operation, the unknown space being an image space indexed by i, j, k (j and k being fixed), the known space being the projection space indexed by l, m, n (n being fixed and m fixed since depending on j and k), the sought-after index a 0 (e) is the index l 0 (i), the axis E is indexed by i and the axis F by l, and the following is then selected: ##EQU20## the terms Z 0 , C, Q 0 , ΔZ and ΔQ being predetermined and l 0 being a real number whose entire part is l.
Claim 12
Dependent Claim
61 words
Method to determine data of an unknown discrete space from data of a known discrete space according to claim 10, wherein a conical and parallel geometry in the case where the elementary phase represents a retroprojection and for a parallel geometry when this phase represents a reprojection, the weighting factor f is determined by taking: V7=0 V6=V1 and V5 is predetermined.
Claim 13
Dependent Claim
128 words
Method to determine data of an unknown discrete space from data of a known discrete space according to claim 1, wherein the elementary phase represents a conical geometry reprojection, the unknown space being the projection space indexed by i, j, k (j and k being fixed) and the known space being the image space indexed by l, m, n (m being fixed, n being fixed since dependent on j and k); the sought-after index a 0 (e) is the index l (i), the axis E is indexed by i and the axis F by l and the following is then selected: ##EQU21## the terms Z 0 , Q 0 , c, ΔZ and ΔQ being predetermined and l 0 being a real number whose entire part is l.
Claim 14
Dependent Claim
47 words
Method to determine data of an unknown discrete space from data of a known discrete space according to claim 12, wherein in the weighting stage for a fan-shaped and conical geometry in the case of reprojection, the weighting factor f is determined by taking: V5=1 V6=1 V7=0.
Claim 15
Dependent Claim
21 words
Application of the method according to claim 1 for the reconstruction of images by synthesis and inversion of the RADON transform.
Claim 16
Dependent Claim
69 words
Application of the method according to claim 14, wherein the reconstruction consists of three stages: a first stage consisting of one parallel geometry bidimensional reprojection making it possible to obtain a RADON transform; a second stage consisting of a parallel geometry bidimensional retroprojection making it possible to obtain rearranged projections; a third stage consisting of a parallel geometry bidimensional retroprojection making it possible to obtain the final sought-after volume.
Claim 17
Dependent Claim
294 words
Device for obtaining data of an unknown discrete space from data of a known discrete space comprising: a defined geometry acquisition device having a radiation source and a set of detectors, the data of each of the spaces being defined respectively by indices corresponding to the coordinates along three axes of the space, the coordinate on an axis F of a corresponding point of the known space being determined by taking the entire part of the address ADR from the generalized expression where: ##EQU22## e being the iteration index corresponding to the coordinate on the axis E of the relevant point of the unknown space and V1, V2, V3 and V4 being constant parameters in the elementary phase, address generation means receiving the parameters V1 to V4 delivering the address ADR which is the entire part of a 0 (e); means to store the numerical values of the points of the known space addressed by the generation means delivering all the (α+1) numerical values of the points of the known space of V(ADR)-β) to V(ADR+δ) (β and δ being positive integers; β+δ=α corresponding to the index ADR being the order of interpolation; interpolation means receiving the numerical values V(ADR-β) to V(ADR+δ) contained in the memory and addressable by the index ADR and determining at the end of interpolation a contribution non-weighted to the numerical value of each corresponding point of the unknown space; weighting means receiving each non-weighted contribution and the parameters V5, V6 and V7 to deliver each weighted contribution, V5, V6 and V7 being constant parameters in the elementary phase; adding means to add to the numerical value of the relevant point of the unknown space the weighted contribution relating to this point in question; and storage means containing the current value.
Claim 18
Dependent Claim
294 words
Device for obtaining data of an unknown discrete space from data of a known discrete space comprising: a defined geometry acquisition device having a radiation source and a set of detectors, the data of each of the spaces being defined respectively by indices corresponding to the coordinates along three axes of the space, the coordinate on an axis F of a corresponding point of the known space being determined by taking the entire part of the address ADR from the generalized expression where: ##EQU22## e being the iteration index corresponding to the coordinate on the axis E of the relevant point of the unknown space and V1, V2, V3 and V4 being constant parameters in the elementary phase, address generation means receiving the parameters V1 to V4 delivering the address ADR which is the entire part of a 0 (e); means to store the numerical values of the points of the known space addressed by the generation means delivering all the (α+1) numerical values of the points of the known space of V(ADR)-β) to V(ADR+δ) (β and δ being positive integers; β+δ=α corresponding to the index ADR being the order of interpolation; interpolation means receiving the numerical values V(ADR-β) to V(ADR+δ) contained in the memory and addressable by the index ADR and determining at the end of interpolation a contribution non-weighted to the numerical value of each corresponding point of the unknown space; weighting means receiving each non-weighted contribution and the parameters V5, V6 and V7 to deliver each weighted contribution, V5, V6 and V7 being constant parameters in the elementary phase; adding means to add to the numerical value of the relevant point of the unknown space the weighted contribution relating to this point in question; and storage means containing the current value.
Claim 19
Dependent Claim
253 words
Device for obtaining data of an unknown discrete space from data of a known discrete space comprising: a defined geometry acquisition device having a radiation source and a set of detectors, the data of each of the spaces being defined respectively by indices corresponding to the coordinates along three axes of the space, the coordinate on an axis F of a corresponding point of the known space being determined by taking the entire part of the address ADR from the generalized expression where: ##EQU23## e being the iteration index corresponding to the coordinate on the axis E of the relevant point of the unknown space and V1, V2, V3 and V4 being constant parameters in the elementary phase, address generation means receiving the parameters V1 to V4 delivering the address ADR which is the entire part of a o (e); means to store the numerical values of the points of the known space addressed by the address generation means delivering all the data V(ADR-β) to V(ADR+δ) (with β+δ=α, β and δ being positive integers) corresponding to the index ADR, being the order of interpolation and delivering the numerical values of the data of the unknown space addressed by the address generation means and receiving the numerical values of the data of the unknown space after processing; an operative interpolation, weighting and adding unit determining the non-weighted contribution, the weighted contribution and determining at the end the numerical value of the current point by adding the weighted contribution to the value preceding this point.
Claim 20
Dependent Claim
253 words
Device for obtaining data of an unknown discrete space from data of a known discrete space comprising: a defined geometry acquisition device having a radiation source and a set of detectors, the data of each of the spaces being defined respectively by indices corresponding to the coordinates along three axes of the space, the coordinate on an axis F of a corresponding point of the known space being determined by taking the entire part of the address ADR from the generalized expression where: ##EQU23## e being the iteration index corresponding to the coordinate on the axis E of the relevant point of the unknown space and V1, V2, V3 and V4 being constant parameters in the elementary phase, address generation means receiving the parameters V1 to V4 delivering the address ADR which is the entire part of a o (e); means to store the numerical values of the points of the known space addressed by the address generation means delivering all the data V(ADR-β) to V(ADR+δ) (with β+δ=α, β and δ being positive integers) corresponding to the index ADR, being the order of interpolation and delivering the numerical values of the data of the unknown space addressed by the address generation means and receiving the numerical values of the data of the unknown space after processing; an operative interpolation, weighting and adding unit determining the non-weighted contribution, the weighted contribution and determining at the end the numerical value of the current point by adding the weighted contribution to the value preceding this point.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1996-11-22 | 1997-05-24 | 1997-11-23 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2000-11-22 | 2001-05-24 | 2001-11-23 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2004-11-22 | 2005-05-24 | 2005-11-23 | $7700.00 | Window Closed / Lapsed |