Some content of this application is unavailable at the moment.
If this situation persist, please contact us atFeedback&Contact
1. (WO2019025533) AUTOMATIC ASSAY ASSESSMENT AND NORMALIZATION FOR IMAGE PROCESSING
Note: Text based on automatic Optical Character Recognition processes. Please use the PDF version for legal matters

Claims

1. A method of normalizing a titer of a first stain within a query image to a titer of the first stain in a template image, the query image being of a biological sample stained with at least the first stain, comprising:

(i) deriving chromatic and density distribution coordinates in the query image within a color model which incorporates density information (603);

(ii) aligning the derived chromatic distributions coordinates in the query image with template image chromatic distribution coordinates to provide transformed chromatic distribution coordinates (604);

(iii) scaling the derived density distribution coordinates in the query image with template image density distribution coordinates to provide transformed density distribution coordinates (605); and

(iv) reconstructing an RGB image by inversely transforming the query image within the color model incorporating the density information (608) using weighted transformed chromatic and density distribution coordinates (607);

wherein the aligning and scaling utilize predetermined parameter values specific for an estimated titer level of the query image.

2. The method of claim 1, wherein the predetermined parameter values are derived mean, angle, and scaling parameters determined at a known first stain titer level.

3. The method of any of claims 1 and 2, wherein the estimated titer level of the query image is determined by computing a weighted average titer score for the query image based on derived first stain color and intensity features (254).

4. The method of any of the preceding claims, wherein the estimated titer level is determined prior to normalization.

5. The method of any of the preceding claims, wherein the estimated titer level is determined during normalization.

6. The method of any of the preceding claims, wherein the weighted average titer score is computed by (a) deriving a plurality of first stain image features from each of a series of patches in the query image (252), and (b) classifying the plurality of derived image features from each of the image patches using a trained titer-identification classifier (253).

7. The method of claim 6, wherein the series of patches are derived by (a) extracting a predefined number of FOVs from the query image (250); (b) computing a set of patches for each of the extracted FOVs (251); and (c) retaining those patches from the set of patches for each extracted FOV that meet threshold patch criteria.

8. The method of claim 6, wherein the titer-identification classifier is a multi-class classifier trained on first stain color and intensity features derived from standardized samples using first stain titer levels as class labels.

9. The method of any of the preceding claims, wherein the weighted transformed chromatic and density distribution coordinates are derived by (i) computing probabilities that pixels are first stain pixels (606); and (ii) weighting the transformed chromatic and density distribution coordinates with the computed probabilities (607).

10. The method of claim 9, wherein chromatic and density distribution coordinates are derived for each pixel in a series of patches in the query image, wherein the series of patches are derived by (a) extracting a predefined number of FOVs from the query image (601); (b) computing a set of patches for each of the extracted FOVs (602); and (c) retaining those patches from the set of patches for each extracted FOV that meet threshold patch criteria.

11. The method of any of the preceding claims, wherein the alignment comprises shifting and rotating the derived chromatic distribution coordinates in the query image to have a same mean and orientation as template chromatic distribution coordinates (604).

12. The method of any of the preceding claims, wherein the scaling comprises transforming the derived density distribution coordinates to have the same weighted mean and weighted standard deviation as the template density distribution coordinates (605).

13. The method of any of the preceding claims, wherein the color module that incorporates density information is an HSD color module.

14. An imaging system (200) comprising: (i) one or more processors (203), and (ii) one or more memories (201) coupled to the one or more processors (203), the one or more memories (201) to store computer-executable instructions that, when executed by the one or more processors (203), cause the imaging system (200) to perform the method of any one of claims 1-13.

15. A non-transitory computer-readable medium (201) storing instructions which, when executed by one or more processors (203) of an imaging system (200), cause the imaging system (200) to perform the method of any one of claims 1-13.