Granted Patent B2

US 12,614,265 · App. 18/173,593 · Granted Apr 28, 2026

Inventors: Alexander Maier (Maxdorf, DE); Jens Deppermann (Schriesheim, DE); Kai Reppin (Speyer, DE); Joost Van Duuren (Mannheim, DE)

Assignee: Roche Molecular Systems, Inc.

G06T7/0002G06T5/20G06T15/00G06V10/25G06V10/56G06V10/764H04N23/67G06T2207/20084G06T2207/30024G06T2207/30168

Abstract

Claims (43)

1 . A computer-implemented method comprising:

accessing a z-stack comprising a plurality of digital images of a sample, wherein each digital image of the plurality of digital images is a two-dimensional image depicts a sample mounted on a slide and corresponds to a different focal distance between the slide and an at imaging device of a slide imaging apparatus;

determining, for each digital image in the z-stack, sharpness values of sub-regions of at least one region of interest of the digital image by using at least one edge detection image filter;

comparing the sharpness values for each sub-regions of the at least one region of interest across the plurality of digital images;

selecting in-focus regions of each digital images in the z-stack based on the comparison; and

generating a three-dimensional digital image based on the selected in-focus regions.

2 . The method of claim 1 , further comprising generating a graphical indication comprising a heating map, wherein the heating map is generated by mapping the sharpness values of the sub-regions as a function of their image coordinates.

3 . The method of claim 2 , wherein the generating of the heating map comprises assigning a color value of a color space to each of the sharpness values.

4 . The method of claim 3 , wherein the color space is an RGB color space, wherein the color space ranges from white 255 for high quality to black 0 for out-of-focus.

5 . The method of claim 2 , wherein the sub-regions correspond to pixels of the digital image, wherein the generating of the heating map comprises selecting groups of pixels of the region of interest, determining an average sharpness value for each of the groups of pixels of the region of interest and mapping the average sharpness value as a function of their image coordinates.

6 . The method of claim 1 , wherein the edge detection image filter is a Laplace filter.

7 . The method of claim 1 , wherein at least one of the sharpness values is a color gradient or intensity gradient.

8 . The method of claim 1 , wherein the region of interest is the whole digital image or a section of the digital image.

9 . The method of claim 1 , further comprising:

i) determining the z-stack, wherein the imaging device comprises at least one transfer device having a focal length, wherein the plurality of digital images is determined at at least three different distances between the transfer device and the slide;

determining information about the sharpness values of the sub-regions; and

generating a graphical indication of the information about the sharpness values as a function of distance for each of the sub-regions.

10 . The method of claim 9 , further comprising: adjusting a focus setting of the transfer device for different distances between the transfer device and the slide depending on the graphical indication.

11 . The computer-implemented method of claim 1 , further comprising determining a seed plane within the z-stack, wherein the seed plane corresponds to a suggested optimal focal plane for the sample as determined by the slide imaging apparatus, and wherein the selection of the in-focus regions is performed with reference to the seed plane.

12 . The method of claim 11 , further comprising coloring the three-dimensional digital image by applying color information to pixels of the three-dimensional digital image.

13 . A computer-implemented method for training a machine and deep learning model for analyzing at least one digital image of a sample mounted on a slide,

wherein the method comprises:

accessing at least one training data set generated by determining a z-stack of digital images of a known sample mounted on the slide using at least one imaging device of a slide imaging apparatus, wherein the known sample has at least one pre-determined or pre-defined feature, wherein the imaging device comprises at least one transfer device having a focal length, wherein the z-stack comprises a plurality of digital images determined at at least three different distances between the transfer device and the slide, wherein said distances are defined based on sharpness values determined by using the computer-implemented method of claim 1 ; and

applying the machine and deep learning model to the z-stack of the digital images and adjusting the machine and deep learning model.

14 . The method of claim 13 , wherein the machine and deep learning model is based on a convolutional neural network and/or a regular neural network.

15 . A system comprising:

one or more data processors; and

a non-transitory computer readable storage medium containing instructions which, when executed on the one or more data processors, cause the one or more data processors to perform a set of actions including:

accessing a z-stack comprising a plurality of digital images of a sample, wherein each digital image of the plurality of digital images is a two-dimensional image depicts a sample mounted on a slide and corresponds to a different focal distance between the slide and an imaging device of a slide imaging apparatus;

automatically determining, for each digital image in the z-stack, sharpness values of sub-regions of at least one region of interest of the digital image by using at least one edge detection image filters;

comparing the sharpness values for each sub-regions of the at least one region of interest across the plurality of digital images;

selecting in-focus regions of each digital images in the z-stack based on the comparison; and

automatically generating a three-dimensional digital image based on the selected in-focus regions.

16 . The system of claim 15 , further comprising the imaging device.

17 . The system of claim 15 , wherein the slide imaging apparatus includes least one controlling and evaluation device.

18 . The system of claim 15 , wherein the set of actions further includes generating a graphical indication comprising a heating map, wherein the heating map is generated by mapping the sharpness values of the sub-regions as a function of their image coordinates.

19 . The system of claim 18 , wherein the generating of the heating map comprises assigning a color value of a color space to each of the sharpness values.

20 . A computer-program product tangibly embodied in a non-transitory machine-readable storage medium, including instructions configured to cause one or more data processors to perform a set of actions including:

accessing a z-stack comprising a plurality of digital images of a sample, wherein each digital image of the plurality of digital images is a two-dimensional image depicts a sample mounted on a slide and corresponds to a different focal distance between the slide and an imaging device of a slide imaging apparatus;

automatically determining, for each digital image in the z-stack, sharpness values of sub-regions of at least one region of interest of the digital image by using at least one edge detection image filter;

comparing the sharpness values for each sub-regions of the at least one region of interest across the plurality of digital images;

selecting in-focus regions of each digital images in the z-stack based on the comparison; and

automatically generating a three-dimensional digital image based on the selected in-focus regions.

Assignments (4)

Priority Claims (1)

Continuity (2)

References Cited (18)