IMAGE PROCESSING DEVICE FOR MATCHING IMAGES OF A SAME PORTION OF A BODY OBTAINED BY MAGNETIC RESONANCE AND ULTRASOUNDS

• US 20100119135A1
• Filed: 03/25/2008
• Published: 05/13/2010
• Est. Priority Date: 04/02/2007
• Status: Active Grant

• Current Assignee: Inria Institut National de Recherche EN Informatique et EN Automatique
• Original Assignee: Inria Institut National de Recherche EN Informatique et EN Automatique
• Inventors: Morandi, Xavier, Barillot, Christian, Coupe, Pierrick, Hellier, Pierre
• Granted Patent: US 8,755,581 B2
• Time in Patent Office: Days
• US Class Current: 382/131
• CPC Class Codes: G06T 2207/10088 Magnetic resonance imaging [MRI]
• CPC Class Codes: G06T 2207/10136 3D ultrasound image
• CPC Class Codes: G06T 2207/20076 Probabilistic image processing
• CPC Class Codes: G06T 2207/30016 Brain
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction
• CPC Class Codes: G06T 7/32 using correlation-based methods
• CPC Class Codes: G06T 7/35 using statistical methods
• CPC Class Codes: G06T 7/38 Registration of image sequences
• CPC Class Codes: G06V 10/24 Aligning, centring, orientation detection or correction of the image

Abstract

According to the invention, a device (D) is dedicated to the processing of data consisting of images of a portion or portions of a body, the device (D) includes processing means (MT) capable of, in the presence of first data representative of the elementary portions of an image obtained by magnetic resonance in a region of the body and of second data representative of the elementary portions of images obtained by ultrasounds in a portion of that region: i) generating first and second maps of the probability that the elementary portions belong to structures of interest of the portion of the region based at least on the intensities associated thereto and respectively defined by the first and second data; ii) estimating a conversion for shifting from one of the first and second maps to the other while maximising for each of their elementary portion the joined probability that it belongs to a same structure of interest; and iii) resetting the structures of interest of one of the first and second maps relative to those of the other by means of the conversion.

Claims

1. Device (D) for processing data that constitute images of a portion or portions of a body, characterised in that it comprises processing means (MT) which are arranged, in the presence of first data representative of elementary parts of an image or images obtained by magnetic resonance in a region of a body and second data representative of elementary parts of an image or images obtained by ultrasound in a portion of this region, so as to i) construct first and second maps of the probability of the elementary parts belonging to structures of interest in said portion of the region, as a function of at least the intensities associated therewith and defined by the first and second data, respectively, ii) estimate a transformation which makes it possible to move from one of said first and second maps to the other, while maximising, for each of their elementary parts, the joint probability that it belongs to the same structure of interest, and iii) register the structures of interest of one of the first and second maps relative to those of the other by means of this transformation.

2. Device according to claim 1, characterised in that the said processing means (MT) are arranged so as to maximise, for each elementary part, a joint probability which is a function of the product of the probability that it belongs to a structure of interest of the magnetic resonance image(s) by the product of the probability that it belongs to a structure of interest of the corresponding ultrasound image(s).

3. Device according to claim 1, characterised in that in the presence of two-dimensional ultrasound images, the said processing means (MT) are arranged so as to construct the said second map by constituting first data which define a three-dimensional ultrasound image from the first data defining the said two-dimensional ultrasound images, then transforming the intensity of each elementary part of the said three-dimensional ultrasound image into a probability of being part of a structure of interest of between 0 and 1.

4. Device according to claim 2, characterised in that in the presence of two-dimensional ultrasound images, the said processing means (MT) are arranged so as to construct the said second map by constituting first data which define a three-dimensional ultrasound image from the first data defining the said two-dimensional ultrasound images, then transforming the intensity of each elementary part of the said three-dimensional ultrasound image into a probability of being part of a structure of interest of between 0 and 1, and, characterised in that the said processing means (MT) are arranged so as to determine, for each elementary part of said second map, the probability of belonging to a structure of interest equal to the ratio between its intensity and 2ⁿ, where n is equal to the number of bits of the ultrasound image(s).

5. Device according to claim 1, characterised in that in the presence of a three-dimensional magnetic resonance image representing a region including the portion represented by said second data, said processing means (MT) are arranged so as to construct a first intermediate map, while determining for each elementary part of said three-dimensional magnetic resonance image a local mean curvature of the intensity value, then constructing said first map from the local mean curvatures of said first intermediate map and auxiliary data relating to said region of the body.

6. Device according to claim 5, characterised in that the said processing means are arranged so as to extract from said first intermediate map the part that corresponds to the portion of region that is represented by said second map, as a function of positional data obtained by means of navigation equipment (EN).

7. Device according to claim 5, characterised in that the said probability of belonging to a structure of interest is equal either to the ratio between the local mean curvature of an elementary part of said first intermediate map and 2ⁿ, if it belongs to a healthy zone and if the local mean curvature is positive, n being equal to the number of bits in the three-dimensional magnetic resonance image, or to a parameter constituting auxiliary data attributed to the corresponding elementary part of the three-dimensional magnetic resonance image, subsequent to the determination of the positional data, if it belongs to a pathological zone, or to zero in every other case.

8. Device according to claim 1, characterised in that the said processing means (MT) are arranged so as to maximise the joint probability by means of an optimising function that does not involve calculation of the partial derivatives of the objective matching function.

9. Device according to claim 1, characterised in that the said processing means (MT) are arranged so as to estimate a rigid registration transformation.

10. Device according to claim 1, characterised in that the said processing means (MT) are arranged so as to estimate a non-rigid registration transformation.

11. Device according to claim 1, characterised in that the said processing means (MT) are arranged so as to construct first and second maps of the probability of belonging to so-called hyper-echo genic structures of interest.

12. Device according to claim 1, characterised in that the said processing means (MT) are arranged so as to construct first and second maps of the probability of belonging to so-called hypo-echo genic structures of interest.

13. Use of the processing device (D) according to claim 1, for processing images of body portiones) suitable for being subjected to imageassisted neurosurgical operations