High resolution alignment of 3D imaging with 2D imaging
Alignment of a 2D image to a corresponding 3D image is provided by writing a pattern into a 3D sample. The pattern is at known positions in the 3D image, and provides visible reference features in the 2D image. This permits accurate determination of the plane in the 3D image that corresponds to the 2D image.
1. A method of aligning a 2D image to a corresponding 3D image, the method comprising:
providing a 3D sample;
writing a predetermined pattern within the 3D sample;
cutting a 2D slice from the 3D sample after the predetermined pattern is present;
imaging the 3D sample to provide a 3D image with a first imaging modality;
imaging the 2D slice to provide a 2D image with a second imaging modality, wherein the predetermined pattern provides reference features in the 2D image; and
automatically determining a 2D section of the 3D image that corresponds to the 2D image using the reference features in the 2D image;
wherein the writing a predetermined pattern within the 3D sample comprises performing a sensitization method on the 3D sample to provide a fluorescence-sensitized sample, followed by optically writing the predetermined pattern via bleaching of fluorescence in the fluorescence-sensitized sample;
wherein the sensitization method is selected from the group consisting of: providing a genetic fluorescence marker to a biological sample, encasing the sample in a fluorescent gel, encasing the sample in a fluorescent glue, and staining the sample with a fluorescent stain.
2. The method of claim 1 , wherein the predetermined pattern is formed by writing a top surface of the 3D sample with a pattern of intersecting lines, whereby the predetermined pattern is a corresponding pattern of intersecting planes.
3. The method of claim 2 , wherein the pattern of intersecting lines includes a first set of lines perpendicular to a second set of lines.
4. The method of claim 3 , wherein the first set of lines has a first spacing pattern between the lines that permits identification of a line from its distance ratios to adjacent lines in the 2D image.
5. The method of claim 3 , wherein the second set of lines has a second spacing pattern between the lines that permits identification of a line from its distance ratios to adjacent lines in the 2D image.
6. The method of claim 3 , wherein the 2D section is cut substantially perpendicular to the top surface of the 3D sample, wherein a plane of the 2D slice intersects the first set of lines and wherein the plane of the 2D slice intersects the second set of lines.
7. A method of aligning a 2D image to a corresponding 3D image, the method comprising:
providing a 3D sample;
writing a predetermined pattern within the 3D sample;
cutting a 2D slice from the 3D sample after the predetermined pattern is present;
imaging the 3D sample to provide a 3D image with a first imaging modality;
imaging the 2D slice to provide a 2D image with a second imaging modality, wherein the predetermined pattern provides reference features in the 2D image; and
automatically determining a 2D section of the 3D image that corresponds to the 2D image using the reference features in the 2D image;
wherein the predetermined pattern is formed by writing a top surface of the 3D sample with a pattern of intersecting lines, whereby the predetermined pattern is a corresponding pattern of intersecting planes;
wherein the 2D section is parametrized by a vector origin ({right arrow over (h)}) and two vector directions ({right arrow over (u)} and {right arrow over (v)}), wherein a z direction is perpendicular to the top surface of the sample, wherein x and y components of {right arrow over (u)}, {right arrow over (v)}, and {right arrow over (h)} are determined by least squares fitting, wherein z components of {right arrow over (u)} and {right arrow over (v)} are determined by assuming uniform shrinkage and no shear, and wherein a z component of {right arrow over (h)} is determined from an independent measurement.
8. The method of claim 1 , further comprising adding particles to the 3D sample, wherein the particles are visible in the first imaging modality, wherein the particles are visible in the second imaging modality, whereby particle features in the 2D and 3D images can be used to improve and/or quantify alignment accuracy between the 2D image and the 3D image.
9. A method of aligning a first 3D image to a second 3D image, the method comprising:
performing the method of claim 1 to align two or more planes in the second 3D image to the first 3D image.
10. The method of claim 1 , wherein the first imaging modality is selected from the group consisting of: intravital microscopy, lipid-cleared 3D imaging, 3D optical imaging, optical coherence tomography, photoacoustic imaging, two photon microscopy, confocal microscopy, diffusive optical imaging and magnetic resonance imaging.
11. The method of claim 1 , wherein the second imaging modality is selected from the group consisting of: histological imaging, non-histological imaging, optical microscopy, electron microscopy, and mass spectroscopy imaging.
12. The method of claim 1 , wherein the sample is selected from the group consisting of: ex vivo biological sample, in vivo biological sample, and non-biological sample.
13. A method comprising:
performing the method of claim 1 multiple times on different tissue samples to provide a training data set, wherein the second imaging modality includes histology, whereby the training data set relates the first imaging modality to corresponding histology;
training a machine learning system using the training data set, whereby the machine learning system can provide synthesized histology images using the first image modality.
14. The method of claim 1 , further comprising characterizing sample distortion from cutting the 2D slice and/or from performing the second imaging modality.
15. The method of claim 14 , wherein the characterizing sample distortion is a time-dependent characterization of dynamic distortion.