SYSTEMS AND METHODS FOR PROCESSING ELECTRONIC IMAGES FOR BIOMARKER LOCALIZATION

1 . A computer-implemented method for analyzing an image corresponding to a specimen, the method comprising:

receiving one or more digital images of a pathology specimen from a patient, the pathology specimen comprising tumor tissue, the one or more digital images being associated with data about a plurality of biomarkers in the tumor tissue;

identifying the tumor tissue and surrounding invasive margin region for each of the one or more digital images;

generating, using a first machine learning model on the one or more digital images, an inference of a plurality of immune markers in the tumor tissue and surrounding invasive margin region;

determining, by a second machine learning model, a spatial relationship of each of the plurality of immune markers identified in the tumor tissue and the surrounding invasive margin region to themselves and to other cell types; and

determining, based on the spatial relationship of each of the plurality of immune markers to themselves and to other cell types, a prediction for a treatment outcome and/or at least one treatment recommendation for the patient.

2 . The computer-implemented method of claim 1 , wherein generating the inference of the plurality of immune markers further comprises using a computer vision model.

3 . The computer-implemented method of claim 1 , wherein the pathology specimen comprises a histology and/or cytology specimen.

4 . The computer-implemented method of claim 1 , further including receiving data about a surrounding invasive margin around the tumor tissues, wherein the data about the plurality of biomarkers is identified from genetic testing, flow cytometry, and/or immunohistochemistry.

5 . The computer-implemented method of claim 1 , wherein identifying the tumor tissue and the surrounding invasive margin region uses a third machine learning model, and wherein training the third machine learning model further comprises:

receiving one or more training digital images associated with a training pathology specimen and an associated indication of a presence or an absence of a tumor region;

dividing the one or more training digital images into at least one sub-region to determine if the tumor is present in the at least one sub-region; and

training the third machine learning model that takes, as an input, one of the one or more training digital images associated with the pathology specimen and predicts whether the tumor is present or not.

6 . The computer-implemented method of claim 5 , wherein training the third machine learning model further comprises:

receiving one or more digital images associated with a target pathology specimen and an associated indication of a presence or an absence of a tumor region;

dividing the one or more digital images into at least one sub-region to analyze to determine if the tumor is present in the at least one sub-region;

applying the third machine learning model to one of the one or more digital images to predict which regions of the digital image show a tumor tissue or an invasive margin and could exhibit a biomarker of interest; and

indicating and flagging a location of at least one tumor region.

7 . The computer-implemented method of claim 1 , wherein identifying the tumor tissue and surrounding invasive margin region further includes:

applying a third machine learning model to the one or more digital images; and

identifying respective locations of any identified tumor tissues and the surrounding invasive margin region.

8 . The computer-implemented method of claim 1 , wherein generating the inference comprises:

determining at least one region of interest in the tumor tissue and the surrounding invasive margin region; and

applying the first machine learning model to determine a prediction of a biomarker expression level in the at least one region of interest.

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

training the second machine learning model to compare the at least one biomarker and the spatial relationship by:

receiving at least one spatially structured training input;

receiving metadata corresponding to each spatially structured training input; and

training the second machine learning model to predict a treatment outcome or a resistance prediction from a localized biomarker.

10 . The computer-implemented method of claim 1 , wherein generating the inference further comprises:

receiving a spatially structured input;

receiving metadata corresponding to the spatially structured input; and

applying the second machine learning model to predict the at least one treatment recommendation or a resistance prediction from a localized biomarker.

11 . A system for analyzing an image corresponding to a specimen, the system comprising:

at least one memory storing instructions; and

at least one processor configured to execute the instructions to perform operations comprising:

receiving one or more digital images of a pathology specimen from a patient, the pathology specimen comprising tumor tissue, the one or more digital images being associated with data about a plurality of biomarkers in the tumor tissue;

identifying the tumor tissue and surrounding invasive margin region for each of the one or more digital images;

generating, using a first machine learning model on the one or more digital images, an inference of a plurality of immune markers in the tumor tissue and surrounding invasive margin region;

determining, by a second machine learning model, a spatial relationship of each of the plurality of immune markers identified in the tumor tissue and the surrounding invasive margin region to themselves and to other cell types; and

determining, based on the spatial relationship of each of the plurality of immune markers to themselves and to other cell types, a prediction for a treatment outcome and/or a at least one treatment recommendation for the patient.

12 . The system of claim 11 , wherein generating the inference of the plurality of biomarkers further comprises using a computer vision model.

13 . The system of claim 11 , wherein the pathology specimen comprises a histology and/or cytology specimen.

14 . The system of claim 11 , further including receiving data about a surrounding invasive margin around the tumor tissues, wherein the data about the plurality of biomarkers is identified from genetic testing, flow cytometry, and/or immunohistochemistry.

15 . The system of claim 11 , wherein identifying the tumor tissue and the surrounding invasive margin region uses a third machine learning model, and wherein training the third machine learning model further comprises:

receiving one or more training digital images associated with a training pathology specimen and an associated indication of a presence or an absence of a tumor region;

dividing the one or more training digital images into at least one sub-region to determine if the tumor is present in the at least one sub-region; and

training the third machine learning model that takes, as an input, one of the one or more training digital images associated with the pathology specimen and predicts whether the tumor is present or not.

16 . The system of claim 15 , wherein training the third machine learning model further comprises:

receiving one or more digital images associated with a target pathology specimen and an associated indication of a presence or an absence of a tumor region;

dividing the one or more digital images into at least one sub-region to analyze to determine if the tumor is present in the at least one sub-region;

applying the third machine learning model to one of the one or more digital images to predict which regions of the digital image show a tumor tissue or an invasive margin and could exhibit a biomarker of interest; and

indicating and flagging a location of at least one tumor region.

17 . The system of claim 11 , wherein identifying the tumor tissue and surrounding invasive margin region further includes:

applying a third machine learning model to the one or more digital images; and

identifying respective locations of any identified tumor tissues and the surrounding invasive margin region.

18 . The system of claim 11 , wherein generating the inference comprises:

determining at least one region of interest in the tumor tissue and the surrounding invasive margin region; and

applying the first machine learning model to determine a prediction of a biomarker expression level in the at least one region of interest.

19 . The system of claim 11 , further comprising:

training the second machine learning model to compare the at least one biomarker and the spatial relationship by:

receiving at least one spatially structured training input;

receiving metadata corresponding to each spatially structured training input; and

training the second machine learning model to predict a treatment outcome or a resistance prediction from a localized biomarker.

20 . A non-transitory computer readable medium storing instructions that, when executed by a processor, cause the processor to perform a method for analyzing an image corresponding to a specimen, the method comprising:

receiving one or more digital images of a pathology specimen from a patient, the pathology specimen comprising tumor tissue, the one or more digital images being associated with data about a plurality of biomarkers in the tumor tissue;

identifying the tumor tissue and surrounding invasive margin region for each of the one or more digital images;

generating, using a first machine learning model on the one or more digital images, an inference of a plurality of immune markers in the tumor tissue and surrounding invasive margin region;

determining, by a second machine learning model, a spatial relationship of each of the plurality of immune markers identified in the tumor tissue and the surrounding invasive margin region to themselves and to other cell types; and

determining, based on the spatial relationship of each of the plurality of immune markers to themselves and to other cell types, a prediction for a treatment outcome and/or a at least one treatment recommendation for the patient.