Computer implemented methods for identifying channels in a 3D volume and computer program product implementing the methods

Patent metadata

Field	Value
Publication number	US-9875537-B2
Application number	US-201414760479-A
Country	US
Kind code	B2
Filing date	Jan 13, 2014
Priority date	Jan 14, 2013
Publication date	Jan 23, 2018
Grant date	Jan 23, 2018

Abstract

The methods comprise:
a) obtaining a 3D volume of the object containing two different sub-volumes identified as: well-defined zone (S) and not-well-defined zone (BZ) sub-volumes; b) generating well-defined zone (S) and not-well-defined zone (BZ) patches from the two sub-volumes; c) automatically identifying the possible channels by means of automatically obtaining candidate channels regions (CCR), dilating the perimeters of the well-defined zone (S) patches.

The method includes embodiments for a layered approach, an EAM polygonal mesh approach and a volume approach.
The computer program product is adapted to implement part or all of the steps of the method of the invention.
The EAM system comprises computing navigation means implementing the method of the invention.

First claim

The invention claimed is:

1. Computer implemented method for identifying channels in a 3D data volume, said channels being passage-like structures of a 3D object within said 3D data volume, which comprises:
a) obtaining a 3D volume of the object containing, directly or on a sub-volume isolated therefrom, at least two different data sub-volumes identified based on physical parameter values representative of physical properties of said object, and termed as: zone S and zone BZ;
b) generating zone S and zone BZ patches from, respectively, said at least two sub-volumes; and
c) automatically identifying possible channels, in said zone BZ patches, by means of automatically obtaining candidate channels regions, as follows:
c.1) dilating at least the perimeters of the zone S patches;
c.2) considering as candidate channel points (CCP) the perimeter points that intersect the perimeter of adjacent zone S patches, and/or the perimeter points that intersect with the same perimeter of the same zone S patch, before reaching a maximum dilation and lie within a zone BZ patch, and
c.3) determining that adjacent candidate channel points (CCP) form a candidate channel region (CCR),

wherein the method being applicable to a medical or veterinary field in particular for the automatic detection of channels in internal organs.

2. The computer implemented method of claim 1 , comprising performing said channels identification using a layered approach, as follows:
performing said patches generation of step b) by means of:
b.1) defining a series of layers representing sections of at least part of the 3D volume or of said sub-volume isolated therefrom, said layers being polygonal meshes; and
b.2) generating patches regarding zone S and zone BZ, from the intersection of the at least two sub-volumes with the layers, defined at b.1), interpolated therein; and

performing said automatic identification of step c) bi-dimensionally on at least one of the defined layers, where the dilation of said sub-step c.1) is performed regarding the zone S patches included in said at least one of the defined layers, and said sub-step c.2) comprises considering as candidate channel points (CCP) the perimeter points that intersect the perimeter of adjacent zone S patches.

3. The computer implemented method of claim 2 , wherein:
said at least two different sub-volumes are at least three sub-volumes identified, based on voxel intensity values and/or colour values, as: zone H, zone S and zone BZ;
said step b) comprises generating zone H, zone S and zone BZ patches from, respectively, said at least three sub-volumes; and
said sub-step b.2) comprises generating patches regarding zone H, zone S and zone BZ, from the intersection of the at least three sub-volumes with the layers, defined at b.1), interpolated therein.

4. The computer implemented method of claim 1 , wherein said physical parameter is associated to at least one of absorption or reflection of light, magnetic or electromagnetic radiation, temperature, electricity, signal intensity, signal phase, time, frequency and colour, or a combination thereof.

5. The computer implemented method of claim 1 , wherein:
at step a):
said obtaining of said 3D volume comprises obtaining an Electro Anatomical Mapping, EAM, 3D volume from memory means; and
said zone S and zone BZ sub-volumes are identified based on values of an electrical parameter and/or of a parameter associated thereto;

the method comprising:
performing said patches generation of step b) by means of:
b.1) retrieving at least one EAM 3D polygonal mesh from said EAM 3D volume or said sub-volume isolated therefrom; and
b.2) generating patches regarding zone S and zone BZ on respective zones of said at least one EAM 3D polygonal mesh coincident with or constituted by the at least two identified sub-volumes;

and performing said automatic identification of step c) bi-dimensionally on said at least one EAM 3D polygonal mesh, where the dilation of said sub-step c.1) is performed regarding the zone S patches included in said at least one EAM 3D polygonal mesh, and said sub-step c.2) comprises considering as candidate channel points (CCP) the perimeter points that intersect the perimeter of adjacent well-defined zone (S) patches.

6. The method of claim 3 , wherein said step c) comprises automatically obtaining candidate channel regions bi-dimensionally on at least two of the defined layers, by means of said steps c.1) to c.3).

7. The method of claim 5 , wherein said step c) comprises automatically obtaining candidate channel regions bi-dimensionally on at least two EAM 3D polygonal meshes, by means of said steps c.1) to c.3).

8. The method of claim 3 , wherein said dilation of step c.1) is performed radially and uniformly.

9. The method of claim 3 , further comprising a step d) for identifying the possible channels by means of automatically obtaining candidate channel regions three-dimensionally across at least two layers of said layers defined at b.1), for finding channels running through different layers.

10. The method of claim 9 , comprising performing said step d) as follows:
d.1) for each zone BZ patch in a layer whose perimeter is completely surrounded by a zone S patch, a representative number of points of the zone BZ patch are classified as candidate channel points (CCP);
d.2) for each zone BZ patch containing candidate channel points (CCP), these are projected towards at least two other of said layers to check for intersections, and when they fall on a zone H patch or when they fall on a zone BZ patch which perimeter is at least in part in contact with a zone H patch they are classified as candidate exit points (CEP); and
d.3) candidate channel regions (CCR) are defined as those regarding a path running across layers through links connecting two groups of candidate exit points (CEP), and containing one or more candidate channel points (CCP).

11. The method of claim 5 , further comprising a step d) for identifying the possible channels by means of automatically obtaining candidate channel regions three-dimensionally across at least two EAM 3D polygonal meshes retrieved at b.1), for finding channels running through different EAM 3D polygonal meshes.

12. The method of claim 11 , comprising performing said step d) as follows:
d.1) for each zone BZ patch in a EAM 3D polygonal mesh whose perimeter is completely surrounded by a zone S patch, a representative number of points of the zone BZ patch are classified as candidate channel points (CCP);
d.2) for each zone BZ patch containing candidate channel points (CCP), these are projected towards at least two other of said EAM 3D polygonal meshes to check for intersections, and when they fall on a zone H patch or when they fall on a zone BZ patch which perimeter is at least in part in contact with a zone H patch they are classified as candidate exit points (CEP); and
d.3) candidate channel regions (CCR) are defined as those regarding a path running across EAM 3D polygonal meshes through links connecting two groups of candidate exit points (CEP), and containing one or more candidate channel points (CCP).

13. The method of claim 1 , wherein said patches are volume patches and said candidate channel regions are candidate channel volumes, the method comprising performing said dilation of sub-step c.1) three-dimensionally on the zone S volume patches.

14. The method of claim 13 , wherein said dilation is performed uniformly and perpendicularly to every point of the faces of each well-defined zone (S) volume patch:
for the faces of the external perimeter of the zone S volume patch, in order to consider, at su

Assignees

• Univ Pompeu Fabra
• Hospital Clínic De Barcelona
• Univ Barcelona
• Hospital Clinic Barcelona

Inventors

• Barbarito Valeria
• Carotenuto Luigi
• Serra Del Molino Luis
• Frangi Alejandro
• Brugada Josep
• Berruezo Antonio

Classifications

• G06T7/0012 Primary

  Biomedical image inspection · CPC title

• G06T7/11

  Region-based segmentation · CPC title

• G06T2207/10088

  Magnetic resonance imaging [MRI] · CPC title

• G06T7/187

  involving region growing; involving region merging; involving connected component labelling · CPC title