Systems and methods for processing electronic images to determine testing for unstained specimens

1. A computer-implemented method, comprising:

receiving a collection of unstained digital histopathology slide images at a storage device;

running a trained machine learning model on one or more slide images of the collection to infer a presence or an absence of a salient feature,

wherein the trained machine learning model has been trained by processing a second collection of unstained or stained digital histopathology slide images and at least one synoptic annotation for one or more unstained or stained digital histopathology slide images of the second collection;

determining at least one map based on output of the trained machine learning model, the determining of the at least one map further comprising:

determining the at least one map using an introspection technique if the output comprises a presence or absence of the salient feature,

determining the at least one map using the output if the output comprises instance/semantic segmentations, and

determining the at least one map by indicating a location of the salient feature if the output comprises detection regions; and

providing an output from the trained machine learning model to the storage device.

2. The computer-implemented method of claim 1 , further comprising:

running an artificial intelligence (AI) system to virtually stain the one or more slide images of the collection of unstained digital histopathology slide images;

partitioning the one or more slide images of the collection into a collection of tiles;

detecting or segmenting at least one tissue region of the one or more slide images of the collection from a non-tissue background of the one or more slide images of the collection to create a tissue mask; and

removing one or more tiles of the collection of tiles detected to be the non-tissue background.

3. The computer-implemented method of claim 1 , further comprising:

indicating, on the one or more slide images of the collection of unstained digital histopathology slide images, where a salient region is located.

4. The computer-implemented method of claim 1 , further comprising:

determining that no slide images in the collection of unstained digital histopathology slide images contain salient tissue; and

selecting another collection of unstained digital histopathology slide images for testing.

5. The computer-implemented method of claim 1 , wherein the determining of the at least one map further comprises:

determining the at least one map using the introspection technique to produce at least one heat map as the at least one map.

6. The computer-implemented method of claim 1 , further comprising:

receiving the second collection of unstained or stained digital histopathology slide images at the storage device;

receiving the at least one synoptic annotation;

training a machine learning model to take as input one or more locations on a slide image to infer a presence of a salient label, wherein the training forms the trained machine learning model; and

saving a set of parameters of the trained machine learning model to the storage device.

7. The computer-implemented method of claim 6 , further comprising:

virtually staining one or more unstained digital histopathology slide images of the second collection to a stain or using an image processing technique to un-stain one or more stained digital histopathology slide images of the second collection;

partitioning the one or more slide images of the second collection into a collection of tiles;

detecting or segmenting at least one tissue region of the one or more slide images of the second collection from a non-tissue background of the one or more slide images of the second collection to create a tissue mask;

removing one or more tiles detected to be the non-tissue background; and

wherein the training of the machine learning model further comprises:

training the machine learning model to take as input one or more locations on a slide image, except for the one or more removed tiles.

8. The computer-implemented method of claim 6 , wherein the at least one synoptic annotation comprises one or more labels for one or more slide images of the second collection.

9. The computer-implemented method of claim 8 , wherein the one or more labels are at a pixel-level, a tile-level, a slide-level, or a part specimen-level.

10. The computer-implemented method of claim 8 , wherein the one or more labels are binary, multi-label binary, categorical, ordinal, or real-valued.

11. A system for using a trained machine learning model for tissue analysis includes memory storing instructions, and at least one processor executing the instructions to perform a process including:

receiving a collection of unstained digital histopathology slide images at a storage device;

running the trained machine learning model on one or more slide images of the collection to infer a presence or an absence of a salient feature,

wherein the trained machine learning model has been trained by processing a second collection of unstained or stained digital histopathology slide images and at least one synoptic annotation for one or more unstained or stained digital histopathology slide images of the second collection;

determining at least one map based on output of the trained machine learning model, the determining of the at least one map further comprising:

determining the at least one map using an introspection technique if the output comprises a presence or absence of the salient feature,

determining the at least one map using the output if the output comprises instance/semantic segmentations, and

determining the at least one map by indicating a location of the salient feature if the output comprises detection regions; and

providing an output from the trained machine learning model to the storage device.

12. The system of claim 11 , wherein the process further comprises:

running an artificial intelligence (AI) system to virtually stain the one or more slide images of the collection of unstained digital histopathology slide images;

partitioning the one or more slide images of the collection into a collection of tiles; detecting or segmenting at least one tissue region of the one or more slide images of the collection from a non-tissue background of the one or more slide images of the collection to create a tissue mask; and

removing one or more tiles of the collection of tiles detected to be the non-tissue background.

13. The system of claim 11 , wherein the process further comprises:

indicating, on the one or more slide images of the collection of unstained digital histopathology slide images, where a salient region is located.

14. The system of claim 11 , wherein the process further comprises:

determining that no slide images in the collection of unstained digital histopathology slide images contain salient tissue; and

selecting another collection of unstained digital histopathology slide images for testing.

15. The system of claim 11 , wherein the determining of the at least one map further comprises:

determining the at least one map using the introspection technique to produce at least one heat map as the at least one map.

16. A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to perform a method for using a trained machine learning model for tissue analysis, the method including:

receiving a collection of unstained digital histopathology slide images at a storage device;

running a trained machine learning model on one or more digital histopathology slide images to infer a presence or an absence of a salient feature,

wherein the trained machine learning model has been trained by processing a second collection of unstained or stained digital histopathology slide images and at least one synoptic annotation for one or more unstained or stained digital histopathology slide images of the second collection;

determining at least one map based on output of the trained machine learning model, the determining of the at least one map further comprising:

determining the at least one map using an introspection technique if the output comprises a presence or absence of the salient feature,

determining the at least one map using the output if the output comprises instance/semantic segmentations, and

determining the at least one map by indicating a location of the salient feature if the output comprises detection regions; and

providing an output from the trained machine learning model to the storage device.

17. The non-transitory computer-readable medium of claim 16 , wherein the method further comprises:

receiving the second collection of unstained or stained digital histopathology slide images at the storage device;

receiving the at least one synoptic annotation;

training a machine learning model to take as input one or more locations on a slide image to infer a presence of a salient label, wherein the training forms the trained machine learning model; and

saving a set of parameters of the trained machine learning model to the storage device.

18. The non-transitory computer-readable medium of claim 16 , further comprising:

virtually staining one or more unstained digital histopathology slide images of the second collection to a stain or using an image processing technique to un-stain one or more stained digital histopathology slide images of the second collection;

partitioning one or more slide images of the second collection into a collection of tiles;

detecting or segmenting at least one tissue region of the one or more slide images of the second collection from a non-tissue background of the one or more slide images of the collection to create a tissue mask;

removing one or more tiles detected to be the non-tissue background; and

wherein the training of the machine learning model further comprises:

training the machine learning model to take as input one or more locations on a slide image, except for the one or more removed tiles.

19. The non-transitory computer-readable medium of claim 16 , wherein the at least one synoptic annotation comprises one or more labels for one or more slide images of the second collection.

20. The non-transitory computer-readable medium of claim 19 , wherein the one or more labels are at a pixel-level, a tile-level, a slide-level, or a part specimen-level.