Automated sample alignment for microscopy
Systems and methods for automated sample alignment for microscopy are described herein. In one aspect a method can include: rotating the sample along a first axis by each of a plurality of rotation angles; imaging, with a charged particle beam, the sample for each rotation angle; and determining a first rotation angle based on the image for each rotation angle, wherein the first rotation angle aligns the sample to the charged particle beam in relation to the first axis.
1. A method for aligning a sample in a microscopy system, comprising:
rotating the sample along a first axis by each of a plurality of rotation angles,
imaging, with a charged particle beam, the sample for each rotation angle;
determining a first rotation angle based on the image for each rotation angle, wherein the first rotation angle aligns the sample to the charged particle beam in relation to the first axis; and
determining a second rotation angle based on the first rotation angle, wherein the second rotation angle aligns the sample to the charged particle beam in relation to a second axis, and wherein the second axis is orthogonal to the first axis.
2. The method according to claim 1 , wherein determining the first rotation angle comprises:
calculating a power spectrum of the image for each rotation angle; and
determining an orientation of the power spectrum for each rotation angle.
3. The method according to claim 2 , wherein the determining the orientation of the power spectrum for each rotation angle is further based on a principal component decomposition of the power spectrum.
4. The method according to claim 1 , wherein determining the first rotation angle comprises:
calculating a power spectrum of the image for each rotation angle; and
determining an anisotropy of the power spectrum for each rotation angle.
5. The method according to claim 1 , wherein the determining the first rotation angle is according to:
θ
maxSlope
=
atan
(
sin
(
β
-
β
stage
)
cos
(
α
-
α
stage
)
sin
(
α
-
α
stage
)
)
,
wherein β−β stage comprises the second rotation angle, α−α stage comprises a rotation angle of the plurality of rotation angles, and θ maxSlope comprises a maximum slope of a power spectrum for a corresponding image of the sample.
6. The method according to claim 1 , wherein the sample comprises a base and at least one cell arranged on the base, and wherein the method further comprises segmenting the image of the sample to extract images of the at least one cell.
7. The method according to claim 6 , wherein the segmentation of the image is, at least in part, carried out by a machine learning algorithm.
8. The method according to claim 6 , wherein determining the first rotation angle comprises:
calculating a power spectrum of the image for each rotation angle; and
determining an orientation of the power spectrum for each rotation angle, wherein the power spectrum of the image for each rotation angle comprises a power spectrum over a region comprising the at least one cell.
9. The method according to claim 1 , wherein determining the first rotation angle comprises:
calculating a power spectrum of the image for each rotation angle; and
determining an orientation of the power spectrum for each rotation angle, and wherein the second rotation angle is determined based on the orientation of the power spectrum.
10. The method according to claim 9 , wherein the method further comprises:
selecting a plurality of principal components of the power spectrum to determine an orientation of the power spectrum, the plurality of principal components comprising a component having a largest eigenvalue and a component having a second largest eigenvalue, wherein the orientation of the power spectrum is determined based on a ratio of the largest eigenvalue to the second largest eigenvalue.
11. The method of claim 1 , wherein the first axis corresponds to a length or a width of the sample.
12. The method of claim 1 , further comprising positioning the sample according to the first rotation angle.
13. The method of claim 1 , wherein rotating the sample along the first axis comprises a sweep along the first axis.
14. The method of claim 1 , wherein the first rotation angle corresponds to a second axis of the sample that is orthogonal to the first axis.
15. A microscopy system comprising:
a charged particle beam;
a sample; and
a controller configured to:
rotate the sample along a first axis by each of a plurality of rotation angles;
image, with the charged particle beam, the sample for the each of the plurality of rotation angles;
determine a first rotation angle based on the image for the each of the plurality of rotation angles, wherein the first rotation angle aligns the sample to the charged particle beam in relation to the first axis; and
determine a second rotation angle based on the first rotation angle, wherein the second rotation angle aligns the sample to the charged particle beam in relation to a second axis, and wherein the second axis is orthogonal to the first axis.
16. The microscopy system according to claim 15 , wherein determining the first rotation angle comprises:
calculating a power spectrum of the image for each rotation angle; and
determining an anisotropy of the power spectrum for each rotation angle.
17. The microscopy system according to claim 15 , wherein determining the first rotation angle comprises:
calculating a power spectrum of the image for each rotation angle; and
determining an orientation of the power spectrum for each rotation angle.
18. The microscopy system according to claim 15 , wherein the controller is further configured to control a tilt of the sample.
19. The microscopy system according to claim 18 , further comprising a data processing unit configured to at least send data to the controller, the data comprising a plurality of rotation angles corresponding to a rotation about the first axis.
20. A non-transitory computer program product comprising instructions that, when executed cause a processor to perform the method according to claim 1 .