Systems and methods for image-based disease characterization - Patents.com

Abstract

Systems and methods are presented herein for automated cytological evaluation of images of in vitro biological samples. More specifically, in certain embodiments, the present disclosure encompasses systems, methods, and devices for diagnostic evaluation of two-dimensional images of stained biological samples using cell cycle transformation functions.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of the filing dates of U.S. Provisional Patent Application No. 63 / 305,540 and U.S. Provisional Patent Application No. 63 / 410,588, both of which are incorporated herein by reference in their entireties. Additionally, U.S. Patent Application No. 15 / 965,664 is also incorporated herein by reference in its entirety.

[0002] Technical Field The present disclosure relates to automated cytological evaluation of images of biological samples. [Background technology]

[0003] Images of biological samples may characterize the various stages of the cell cycle, e.g., M, G0, G1, S, and G2. To divide, a cell must complete several tasks: growing, copying its genetic material (DNA), and physically dividing into two daughter cells. Cells perform these tasks in an organized, predictable series of steps that constitute the cell cycle. The cell cycle is not a linear path, but a cycle, because at the end of each turn, the two daughter cells may begin the exact same process all over again.

[0004] Generally, cell division (sometimes called the "cell cycle") occurs in two distinct phases: interphase and mitosis (sometimes called the "M phase," "mitotic phase," or simply "M"). Interphase is composed of the G1, S, and G2 phases. In the G1 phase (sometimes called the "first gap phase"), the cell is actively growing and is receptive to signals to initiate DNA synthesis. In the S phase, the cell synthesizes a complete copy of its DNA in the nucleus and replicates microtubule-organizing structures called centrosomes. The centrosomes help separate the DNA for the M phase. In the G2 phase (sometimes called the "second gap phase"), the cell is actively preparing for the M phase and contains twice the normal amount of DNA in the nucleus. Summary of the Invention

[0005] In one aspect, a computer-implemented method is described herein, the method including: receiving, by a processor of a computing device, a digital image of a biological sample, the digital image including an area divided into a plurality of subsections; calculating, by the processor, a diagnostic score for a first subsection of the plurality of subsections, the diagnostic score including a plurality of values ​​(e.g., vector values) collectively related to nuclear heat and nuclear area of ​​one or more cells imaged in the first subsection of the biological sample; performing, by the processor, a cell cycle variant function to generate a diagnostic index corresponding to the first subsection, the cell cycle variant function identifying one or more cell cycle variants based on a subset of the plurality of values ​​of the diagnostic score for the first subsection; mapping, by the processor, the diagnostic index of the first subsection to a reference scale to determine whether the diagnostic index of the first subsection exceeds a threshold value of the cell cycle variant reference scale; and determining, by the processor, the presence of one or more cell cycle variants associated with the first subsection based on the diagnostic index exceeding the threshold value.

[0006] In certain embodiments, the method further includes graphically indicating (e.g., highlighting) the first subsection in the region of the digital image based on the diagnostic indicator exceeding the threshold. In certain embodiments, the graphically indicating includes generating a two-dimensional shape (e.g., a polygon) to be rendered on the digital image, the two-dimensional shape including a boundary (e.g., a continuous boundary) surrounding the first subsection based on the diagnostic indicator exceeding the threshold.

[0007] In certain embodiments, the method further includes generating one or more pathology annotations associated with the boundary (e.g., annotating the boundary for subsequent review by a pathologist). In certain embodiments, the boundary further encloses one or more additional subsections of the plurality of subsections based on the respective diagnostic index exceeding a threshold. In certain embodiments, the boundary excludes one or more subsections whose corresponding diagnostic index does not exceed a threshold. In certain embodiments, the method further includes generating a heat map within the two-dimensional shape, the heat map indicating the degree(s) to which the diagnostic index of the corresponding first subsection and each of the corresponding diagnostic indexes of the one or more additional subsections exceed a threshold. In certain embodiments, the two-dimensional shape includes a convex hull polygon.

[0008] In certain embodiments, the execution includes mapping a subset of the plurality of values ​​of the diagnostic score to a corresponding variable in the cell cycle transformation function.

[0009] In certain embodiments, the identified degree of presence of cell cycle alteration in the first subsection is proportional to the degree to which the diagnostic indicator exceeds a threshold value.

[0010] In certain embodiments, cell cycle transformation functions are used to identify regions of high grade cancer (e.g., to identify regions containing mitotic figures) [e.g., shape ( <c>) by quantifying junction curvature variance] to calculate cell cycle S-phase dysregulation.

[0011] In certain embodiments, a cell cycle transformation function is calculated for active cell cycles to identify regions of high Ki67 expression (e.g., to identify regions containing mitotic figures) [e.g., by quantifying (C_L initial skewness) x (C_L maximum skewness)].

[0012] In certain embodiments, the cell cycle transformation function calculates nuclear size skewness and chromosomal instability to identify regions of high DNA ploidy [e.g., by quantifying the percentage of tissue regions with high (C_L) maximum skewness].

[0013] In certain embodiments, the cell cycle transformation function is [e.g., shape ( <c>) to identify regions of high stromal TILs] by quantifying the deviations in stromal architectural distortion and cell cycle arrest.

[0014] In certain embodiments, the cell cycle transformation function is [e.g., shape (<C_L> ) by quantifying divergence] to calculate cell cycle G1 / S dysregulation and arrest signatures to identify HER2 positive / amplified regions.

[0015] In certain embodiments, the cell cycle transformation function is [e.g., shape ( <c>) by quantifying divergence] to calculate cell cycle G1 / S dysregulation and arrest signatures to identify HR positive / overexpressing regions.

[0016] In certain embodiments, the cell cycle transformation function is [e.g., ( <c>maximum- <c>Minimum)×( <c>Cell cycle G1 entry and G0 arrest signatures are calculated to identify regions harboring high quiescent population loads (QPLs) [by quantifying the maximal skewness].

[0017] In certain embodiments, the cell cycle transformation function is [e.g., shape ( <c>) Calculate chromosomal instability to identify BRCA-positive and HRD-carrying regions by quantifying boundary jumps.

[0018] In certain embodiments, the cell cycle transformation function is [e.g., shape ( <c>) by quantifying cross-sectional gradient variation to calculate cell cycle arrest and chromosomal instability to identify regions with MMR function deficiency.

[0019] In certain embodiments, the graphically indicating includes rendering an overlay on the digital image, the overlay graphically identifying one or more regions of the digital image that correspond to the cellular abnormality.

[0020] In certain embodiments, the method includes automatically identifying, by a processor, one or more edges of the biological specimen depicted in the digital image and constraining the two-dimensional shape to a region within the one or more edges of the biological specimen.

[0021] In certain embodiments, the image of the biological sample is a stained tissue image (e.g., a two-dimensional image, e.g., a hematoxylin and eosin stained formalin-fixed paraffin-embedded tissue image (H&E stained FFPE image)).

[0022] In certain embodiments, the method includes executing, by a processor, one or more cell cycle transformation functions each to generate a diagnostic indicator corresponding to each of a plurality of divided subsections of the digital image, wherein the one or more cell cycle transformation functions (collectively) determine one or more diagnostic entities selected from the group consisting of: histological grade of the cancer, mitotic figures, Ki67 status, DNA ploidy, stromal TIL status, HER2 status, HR (combined ER, PR) status, quiescent population burden (QPL), BRCA mutation or HRD status, MMR function deficiency status, or any combination of the above.

[0023] In certain embodiments, the method includes executing, by a processor, one or more cell cycle transformation functions each to generate a corresponding diagnostic indicator for each of a plurality of divided subsections of the digital image, wherein the one or more cell cycle transformation functions (collectively) calculate one or more surrogate biological signatures selected from the group consisting of: cell cycle DNA synthesis phase (S phase) dysregulation, length of the active phase of the cell cycle, M phase activity, nuclear size skewness and chromosomal instability, global measures of interstitial architectural distortion and cell cycle arrest, cell cycle G1 / S dysregulation and arrest signature, cell cycle G1 entry and G0 arrest signature, chromosomal instability, cell cycle arrest, and / or combinations of the above.

[0024] In another aspect, a method is described herein that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically assessing, by the processor, a measure of cell cycle S-phase dysregulation for each of at least one subsection of the digital image to identify regions of high-grade cancer (e.g., the digital image is divided into multiple subsections corresponding to different regions of the imaged biological sample).

[0025] In certain embodiments, the automatically evaluating includes determining a shape ( <c>) includes calculating the curvature variance of the joint.

[0026] In certain embodiments, the method comprises using a measure of cell cycle S-phase dysregulation to identify one or more regions of the digital image that contain mitotic figures.

[0027] In another aspect, a method is described herein that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically evaluating, by the processor, a measure of cell cycle activity in each of at least one subsection of the digital image to identify regions of high Ki67 expression (e.g., the digital image is divided into multiple subsections corresponding to different regions of the imaged biological sample).

[0028] In certain embodiments, the automatically evaluating includes calculating (C_L-initial skewness) x (C_L-maximum skewness) for each of the at least one subsection.

[0029] In certain embodiments, the methods involve using measurements of cell cycle activity to identify one or more regions of the digital image that contain mitotic figures.

[0030] In another aspect, a method is described herein that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically assessing, by the processor, measures of nuclear size skewness and chromosomal instability for each of at least one subsection of the digital image to identify regions of high DNA ploidy (e.g., the digital image is divided into multiple subsections corresponding to different regions of the imaged biological sample).

[0031] In certain embodiments, the automatically evaluating includes calculating a percentage of tissue area with high skewness of the (C_L) maximum for each of the at least one subsection.

[0032] In another aspect, a method is described herein that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically assessing, by the processor, measures of stromal architectural distortion and cell cycle arrest in each of at least one subsection of the digital image to identify regions of stromal rich TILs (e.g., the digital image is divided into multiple subsections corresponding to different regions of the biological sample imaged).

[0033] In certain embodiments, the automatically evaluating includes determining a shape ( <c>) deviation.

[0034] In another aspect, a method is described herein that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically assessing, by the processor, measures of cell cycle G1 / S dysregulation and arrest signatures for each of at least one subsection of the digital image to identify HER2 positive / amplified regions (e.g., the digital image is divided into multiple subsections corresponding to different regions of the imaged biological sample).

[0035] In certain embodiments, the automatically evaluating includes determining a shape (<C_L> ) divergence.

[0036] In another aspect, a method is described herein that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically assessing, by the processor, measures of cell cycle G1 / S dysregulation and arrest signatures for each of at least one subsection of the digital image to identify HR positive / overexpression regions (e.g., the digital image is divided into multiple subsections corresponding to different regions of the imaged biological sample).

[0037] In certain embodiments, the automatically evaluating includes determining a shape ( <c>) divergence.

[0038] In another aspect, a method is described herein that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically evaluating, by the processor, measures of cell cycle G1 entry and G0 arrest signatures for each of at least one subsection of the digital image to identify regions possessing a high quiescent population load (QPL). (e.g., the digital image is divided into multiple subsections corresponding to different regions of the imaged biological sample). In certain embodiments, the automatically evaluating includes automatically evaluating, for each of the at least one subsection, ( <c>maximum- <c>Minimum)×( <c>This involves calculating the maximum skewness.

[0039] In another aspect, a method is described herein that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically assessing, by the processor, a measure of chromosomal instability for each of at least one subsection of the digital image to identify BRCA positive and HRD carrying regions (e.g., the digital image is divided into multiple subsections corresponding to different regions of the imaged biological sample).

[0040] In certain embodiments, the automatically evaluating includes determining a shape ( <c>) This involves calculating the boundary jumps.

[0041] In another aspect, a method is described herein that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically assessing, by the processor, measures of cell cycle arrest and chromosomal instability for each of at least one subsection of the digital image to identify regions of MMR function deficiency (e.g., the digital image is divided into multiple subsections corresponding to different regions of the biological sample imaged).

[0042] In certain embodiments, the automatically evaluating includes determining a shape ( <c>) including calculating the cross-sectional slope variation.

[0043] In another aspect, a system is described herein that includes a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to perform one or more steps of the methods described herein.

[0044] In another aspect, a computer-implemented method is described herein, the method including determining, by a processor of a computing device, the presence of one or more cell cycle variants associated with a first subsection of a digital image of a biological sample, the digital image including an area divided into a plurality of subsections, the determining being based on a diagnostic index exceeding a threshold, the diagnostic score for the first subsection including a plurality of values ​​(e.g., vector values) collectively related to nuclear heat and nuclear area of ​​one or more cells imaged in the first subsection of the biological sample, and the diagnostic index for the first subsection being determined using a cell cycle variant function that identifies the one or more cell cycle variants based on a subset of the plurality of values ​​of the diagnostic score for the first subsection.

[0045] In certain embodiments, the cell cycle transformation function calculates a surrogate biological signature selected from the group consisting of: cell cycle DNA synthesis phase (S phase) dysregulation, length of the active phase of the cell cycle, M phase activity, nuclear size skewness and chromosomal instability, global measures of interstitial architectural distortion and cell cycle arrest, cell cycle G1 / S dysregulation and arrest signature, cell cycle G1 entry and G0 arrest signature, chromosomal instability, cell cycle arrest, or any combination of the above.

[0046] In certain embodiments, the cell cycle transformation function determines a diagnostic entity selected from the group consisting of: histological grade of the cancer, mitotic figures, Ki67 status, DNA ploidy, stromal TIL status, HER2 status, HR (ER, PR combined) status, quiescent population load (QPL), BRCA mutation or HRD status, MMR function deficiency status, or any combination of the above.

[0047] In certain embodiments, the method further includes automatically identifying and graphically rendering, by the processor, a two-dimensional shape (e.g., a polygon) including a boundary (e.g., a continuous boundary) enclosing the first subsection and one or more additional subsections of the digital image based on the diagnostic indicators of each of the first subsections and each of the additional subsections that exceed the threshold.

[0048] In certain embodiments, the cell cycle transformation function calculates cell cycle G1 / S dysregulation and cell cycle G1 phase entry dysregulation to identify regions containing one or more of BRAF and NRAS gene mutations.

[0049] According to another aspect, a method is provided that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically assessing, by the processor, measures of cell cycle G1 / S dysregulation and cell cycle G1 phase entry dysregulation for each of at least one subsection of the digital image to identify regions containing one or more of a BRAF and NRAS mutation.

[0050] In certain embodiments, the automatically evaluating includes determining a shape ( <c>maximum, <c>Minimum, <c>) divergence.

[0051] In certain embodiments, the cell cycle transformation function calculates dysregulation of the cell cycle G1 / S to identify areas of dysplasia.

[0052] According to another aspect, a method is provided that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically assessing, by the processor, a measure of cell cycle G1 / S dysregulation in each of at least one subsection of the digital image to identify areas of dysplasia.

[0053] In certain embodiments, the automatically evaluating includes determining a shape (<C_L> , <c>) divergence.

[0054] In certain embodiments, the cell cycle transformation function calculates cell cycle G2 / M dysregulation to identify the degree of dysplasia within a region.

[0055] According to another aspect, a method is provided that includes receiving, by a processor of a computing device, a digital image of a biological sample, and automatically assessing, by the processor, a measure of cell cycle G2 / M dysregulation in each of at least one subsection of the digital image to identify a degree of dysplasia in an area of ​​the digital image.

[0056] In certain embodiments, the automatically evaluating includes determining a shape (<C_L> , <c>Maximum) divergence and ( <c>maximum- <c>This involves calculating the minimum (minimum) shift.

[0057] The details of one or more embodiments of the subject matter described herein are set forth in the accompanying drawings and the description below. Other features, aspects, and advantages of the subject matter will become apparent from the description, drawings, and claims.

[0058] definition In order that this disclosure may be more readily understood, certain terms are first defined below. Additional definitions for the following terms, as well as other terms, are set forth throughout the specification.

[0059] In this application, the use of "or" means "and / or" unless otherwise specified. As used in this application, the term "comprises" and variations of the term such as "comprising" and "comprises" are not intended to exclude other additives, ingredients, integers, or steps. As used in this application, the terms "about" and "approximately" are used synonymously. Any numbers used in this application, with or without "about" are intended to cover normal variations that would be recognized by one of ordinary skill in the relevant art. In certain embodiments, the term "about" or "about" refers to a range of values ​​that falls within 25%, 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or less in either direction (greater or less) of the stated reference value, unless otherwise stated (except where such numbers exceed 100% of the possible values) or are otherwise clear from the context.

[0060] As used herein, the term "cancer" refers to a disease, disorder, or condition in which cells exhibit relatively abnormal, uncontrolled, or autonomous growth, resulting in an abnormal growth phenotype characterized by an abnormally high proliferation rate or a significant loss of control of cell proliferation. In some embodiments, cancer is characterized by one or more tumors. Those skilled in the art are aware of various types of cancer, including, for example, adrenocortical carcinoma, astrocytoma, basal cell carcinoma, carcinoid, cardiac cancer, cholangiocarcinoma, chordoma, chronic myeloproliferative neoplasms, craniopharyngioma, ductal carcinoma in situ, ependymoma, intraocular melanoma, gastrointestinal carcinoid tumor, gastrointestinal stromal tumor (GIST), gestational trophoblastic disease, glioma, histiocytosis, leukemia (e.g., acute lymphocytic leukemia (ALL)), acute bone marrow tumor, leukemia (e.g., acute myeloblastic leukemia (ALL)), acute myeloblastic leukemia (AL ... Myeloid leukemia (AML, chronic lymphocytic leukemia (CLL, chronic myelogenous leukemia (CML), hairy cell leukemia, myeloid leukemia, myelogenous leukemia), lymphoma (e.g., Burkitt lymphoma [non-Hodgkin's lymphoma]), cutaneous T-cell lymphoma, Hodgkin's lymphoma, mycosis fungoides, Sézary syndrome, AIDS-related lymphoma, follicular lymphoma, diffuse large B-cell lymphoma), melanoma, Merkel cell carcinoma, Mesothelioma, myeloma (e.g., multiple myeloma), myelodysplastic syndrome, papillomatosis, paraganglioma, pheochromocytoma, pleuropulmonary blastoma, retinoblastoma, sarcoma (e.g., Ewing's sarcoma, Kaposi's sarcoma, osteosarcoma, rhabdomyosarcoma, uterine sarcoma, angiosarcoma), Wilms' tumor, or adrenal cortical carcinoma, anal cancer, appendix cancer, bile duct cancer, bladder cancer, bone cancer, brain cancer, breast cancer, bronchial cancer, central nervous system cancer, cervical cancer, colon cancer, endometrial cancer, esophageal cancer, eye cancer Cancer, fallopian tube cancer, gallbladder cancer, gastrointestinal cancer, germ cell cancer, head and neck cancer, heart cancer, intestine cancer, kidney cancer (e.g., Wilms' tumor), laryngeal cancer, liver cancer, lung cancer (e.g., non-small cell lung cancer, small cell lung cancer), oral cavity cancer, nasal cavity cancer, oral cancer, ovarian cancer, pancreatic cancer, rectal cancer, skin cancer, stomach cancer, testicular cancer, throat cancer, thyroid cancer, penis cancer, pharynx cancer, peritoneal cancer, pituitary cancer, prostate cancer, rectal cancer, salivary gland cancer, ureter cancer, urethra cancer, uterine cancer, vaginal cancer, or vulvar cancer. For example, tissue images shown in the figures include endometrial cancer (FIG. 15) and breast cancer (all others).

[0061] As used herein, the term "detector" includes any detector of electromagnetic radiation, including, but not limited to, CCD cameras, photomultiplier tubes, photodiodes, and avalanche photodiodes.

[0062] Many of the methods described herein include a "determining" step. Upon reading this specification, those skilled in the art will understand that such "determining" can be utilized or performed using any of a variety of techniques available to those skilled in the art, including, for example, the specific techniques explicitly mentioned herein. In some embodiments, determining includes physical manipulation of the sample. In some embodiments, determining includes considering or manipulating data or information, for example, using a computer or other processing unit adapted to perform a relevant analysis. In some embodiments, determining includes receiving relevant information or material from a source. In some embodiments, determining includes comparing one or more characteristics of the sample or entity to a comparable reference. Similarly, the "calculating", "performing" or "mapping" steps can utilize or be performed using the methods discussed above with respect to the "determining" step.

[0063] As used herein, "diagnosis" refers to providing any type of diagnostic information (including, but not limited to, whether a subject has or is likely to develop a disease, disorder, or condition, the status, staging, or characteristics of a disease, disorder, or condition present in the subject; information regarding the nature or classification of a tumor; information regarding prognosis, or information that aids in selecting an appropriate treatment). Selection of treatment may include selection of a particular therapeutic agent or other treatment modality (e.g., surgery, radiation, etc.), selection of whether to withhold or administer treatment, selection of a dosing regimen (e.g., frequency or level of administration of one or more doses of a particular therapeutic agent or combination of therapeutic agents), and the like.

[0064] As used herein, the term "electromagnetic radiation" or "radiation" is understood to mean self-propagating waves in space of electric and magnetic components that vibrate at right angles to each other and to the direction of propagation and are in phase with each other. Electromagnetic radiation includes radio waves, microwaves, red light, infrared and near infrared, visible light, ultraviolet light, x-rays, and gamma rays.

[0065] As used herein, the term "genotype" refers to the diploid combination of alleles at a given locus, or set of related loci, in a given cell or organism. A homozygous subject has two copies of the same allele, and a heterozygous subject has two different alleles. In the simplest case, a locus with two alleles, "A" and "a", can form three genotypes (A / A, A / a, and a / a).

[0066] As used herein, an "image" (e.g., a two-dimensional or three-dimensional image of an in vitro biological sample such as tissue) includes any visual representation (e.g., a photograph, a video frame, streaming video) and any electronic, digital, or mathematical analogue of a photograph, video frame, or streaming video. Any of the devices described herein, in certain embodiments, includes a display for displaying an image or any other result generated by the processor. Any of the methods described herein, in certain embodiments, includes a step of displaying an image or any other result generated by the method.

[0067] As used herein, the term "in vitro" refers to events that take place in an artificial environment, such as in cell culture, in a test tube or reaction vessel, rather than in a multicellular organism.

[0068] As used herein, the term "in vivo" refers to events that occur within multicellular organisms, such as humans and non-human animals. In the context of cell-based systems, the term may be used to refer to events that occur within living cells (e.g., as opposed to in vitro systems).

[0069] As used herein, the term "map" is understood to mean a visual display containing spatially correlated information, or any data representation that can be interpreted for visual display. For example, a three-dimensional map of a given volume may include a data set of values ​​of a given quantity that varies in three spatial dimensions throughout the volume. A three-dimensional map may be displayed in two dimensions (e.g., a two-dimensional screen or a two-dimensional printout).

[0070] As used herein, the term "phenotype" refers to a trait, or a class or set of traits, exhibited by a cell or organism. In some embodiments, a particular phenotype may be correlated with a particular allele or genotype. In some embodiments, a phenotype may be discrete, and in some embodiments, a phenotype may be continuous.

[0071] As used herein, the term "reference" describes a standard or control against which a comparison is made. For example, in some embodiments, a drug, animal, individual, population, sample, sequence, or value of interest is compared to a drug, animal, individual, population, sample, sequence, or value of a reference or control. In some embodiments, the reference or control is tested or determined substantially simultaneously with the test or determination of interest. In some embodiments, the reference or control is a historical reference or control, optionally embodied in a tangible medium. Typically, as will be understood by those skilled in the art, the reference or control is determined or characterized under similar conditions or circumstances as those of the evaluation. Those skilled in the art will understand when there is sufficient similarity to justify reliance or comparison on a particular possible reference or control.

[0072] As used herein, the term "sample" generally refers to an aliquot of material obtained or generated from a source of interest, as described herein. In some embodiments, the source of interest is a biological or environmental source. In some embodiments, the source of interest may be or include a cell or organism, such as a microorganism, a plant, or an animal (e.g., a human). In some embodiments, the source of interest is or includes a biological tissue or bodily fluid. In some embodiments, the biological tissue is an in vitro sample prepared for two-dimensional or three-dimensional imaging or other cytological examination. For example, the sample may be stained to better reveal the cellular structure of the sample (e.g., nuclei, cytoplasm, and cell granules). Various staining methods or other sample processing, such as fixation, dehydration, clearing, and slide mounting, may be used. For example, the sample may be a hematoxylin and eosin (H&E) stained sample, or the sample may include formalin-fixed paraffin-embedded (FFPE) tissue. In some embodiments, the biological tissue or fluid may be or include amniotic fluid, aqueous humor, peritoneal fluid, bile, bone marrow, blood, breast milk, cerebrospinal fluid, earwax, chyle, chime, semen, endolymphatic fluid, exudate, feces, gastric acid, gastric juice, lymph, mucus, pericardial fluid, perilymphatic fluid, peritoneal fluid, pleural fluid, pus, catarrhal secretions, saliva, sebum, semen, serum, smegma, phlegm, synovial fluid, sweat, tears, urine, vaginal secretions, vitreous humor, vomit, or a combination or component(s) thereof. In some embodiments, the biological fluid may be or include intracellular fluid, extracellular fluid, intravascular fluid (plasma), interstitial fluid, lymphatic fluid, or transcellular fluid. In some embodiments, the biological fluid may be or include plant exudates. In some embodiments, the biological tissue or sample may be obtained, for example, by aspiration, biopsy (e.g., fine needle or tissue biopsy), swab (e.g., oral, nasal, skin, or vaginal swab), scraping, surgery, lavage or washing (e.g., bronchoalveolar, ductal, nasal, ocular, oral, uterine, vaginal, or other lavage or washing). In some embodiments, the biological sample is or comprises cells obtained from an individual.In some embodiments, a sample is a "primary sample" obtained directly from a source of interest by any suitable means. In some embodiments, as the context will allow, the term "sample" refers to a preparation obtained by processing the primary sample (e.g., by removing one or more components of the primary sample or by adding one or more agents thereto), e.g., filtration using a semi-permeable membrane. Such a "processed sample" may, for example, include nucleic acids or proteins extracted from the sample or obtained by subjecting the primary sample to one or more techniques (e.g., amplification or reverse transcription of nucleic acids, separation or purification of specific components, etc.).

[0073] As used herein, the term "substantially" refers to a qualitative state that indicates a total or nearly total degree or degree of a desired property or characteristic. Those skilled in the art of biology will understand that few, if any, biological and chemical phenomena proceed to completion or completeness, or achieve or avoid absolute results. Thus, the term "substantially" is used herein to capture the potential lack of completeness inherent in many biological and chemical phenomena.

[0074] The above and other objects, aspects, features, and advantages of the present disclosure will become more apparent and be better understood by referring to the following description taken in conjunction with the accompanying drawings. [Brief description of the drawings]

[0075] [Figure 1] FIG. 1 is a schematic diagram of a diagnostic procedure involving digital analysis of a two-dimensional image of a stained biological sample, according to an exemplary embodiment. [Diagram 2] FIG. 1 is a schematic diagram of the cell cycle presented for contextual purposes. [Figure 3-1] FIG. 1 is a block flow diagram of a diagnostic procedure for cytological analysis of images of an in vitro biological sample, according to an exemplary embodiment. [Figure 3-2] FIG. 1 is a block flow diagram of a diagnostic procedure for cytological analysis of images of an in vitro biological sample, according to an exemplary embodiment. [Figure 4] AF are images of biological samples and corresponding plots showing quantification of cell cycle variations to identify regions with clinically significant morphometric features such as grade, according to an exemplary embodiment. [Diagram 5] 5A-5D are images of a biological sample including computed and rendered graphical overlays showing user interfaces such as diagnostic boundary backprojection, pathology annotations, and graphical heat maps, according to an exemplary embodiment. [Figure 6] A-E are images of biological samples and corresponding plots showing the determination of a cell cycle transformation function (Function 1) to calculate cell cycle S-phase dysregulation and identify areas of high-grade cancer, according to an exemplary embodiment. [Figure 7] A-E are images of biological samples and corresponding plots showing the determination of a cell cycle transformation function (Function 2) to calculate cell cycle activity and identify regions of high Ki67 expression, according to an exemplary embodiment. [Figure 8] 1A-C are images of a biological sample and corresponding overlays illustrating the use of cell cycle transformation functions 1 and 2 described above to identify regions containing mitotic figures, according to an exemplary embodiment. [Figure 9] A-E are images of biological samples and corresponding plots showing determination of a cell cycle transformation function (Function 3) to calculate nuclear size skewness and chromosomal instability and identify regions of high DNA ploidy, according to an exemplary embodiment. [Figure 10] A-E are images of biological samples and corresponding plots showing the determination of a cell cycle transformation function (Function 4) to calculate architectural distortion and cell cycle arrest within the stroma to identify regions of stromal-rich TILs, according to an exemplary embodiment. [Figure 11] A-E are images of biological samples and corresponding plots showing the determination of a cell cycle transformation function (Function 5) to calculate cell cycle G1 / S dysregulation and arrest signatures and identify HER2 positive / amplified regions according to an exemplary embodiment. [Figure 12] A-E are images of biological samples and corresponding plots showing the determination of a cell cycle transformation function (Function 6) to calculate cell cycle G1 / S dysregulation and arrest signatures and identify HR positive / overexpression regions according to an exemplary embodiment. [Figure 13] A-E are images of biological samples and corresponding plots showing the determination of a cell cycle transformation function (Function 7) to calculate cell cycle G1 entry and G0 arrest signatures and identify regions harboring high quiescent population load (QPL) according to an exemplary embodiment. [Figure 14] A-F are images of biological samples and corresponding plots and heatmap overlays showing the determination of a cell cycle transformation function (Function 8) to calculate chromosomal instability and identify BRCA positive and HRD carrying regions, according to an exemplary embodiment. [Figure 15] A-E are images of biological samples and corresponding plots showing the determination of a cell cycle transformation function (function 9) to calculate cell cycle arrest and chromosomal instability to find regions of MMR function deficiency, according to an exemplary embodiment. [Figure 16] FIG. 1 is a block flow diagram of a method for determining the presence of one or more cell cycle variants associated with each of one or more subsections of a digital image of a biological sample, according to an exemplary embodiment. [Figure 17] FIG. 1 is a block flow diagram of a method for automated assessment of a measure of cell cycle S-phase dysregulation for each of at least one subsection of a digital image of a biological sample, according to an exemplary embodiment. [Figure 18] FIG. 1 is a block flow diagram of a method for automatically assessing cell cycle measures of activity for each of at least one subsection of a digital image of a biological sample, according to an exemplary embodiment. [Figure 19] FIG. 1 is a block flow diagram of a method for automatically assessing measures of nuclear size skewness and chromosomal instability for each of at least one subsection of a digital image of a biological sample, according to an exemplary embodiment. [Figure 20] FIG. 1 is a block flow diagram of a method for automatically assessing measures of interstitial architectural distortion and cell cycle arrest for each of at least one subsection of a digital image of a biological sample, according to an exemplary embodiment. [Figure 21] FIG. 1 is a block flow diagram of a method for automated assessment of cell cycle G1 / S dysregulation and arrest signature measurements for each of at least one subsection of a digital image of a biological sample to identify HER2 positive / amplified regions, according to an exemplary embodiment. [Figure 22] FIG. 1 is a block flow diagram of a method for automated assessment of cell cycle G1 / S dysregulation and arrest signature measurements for each of at least one subsection of a digital image of a biological sample to identify HR positive / overexpression regions, according to an exemplary embodiment. [Diagram 23] FIG. 1 is a block flow diagram of a method for automatically assessing measurements of cell cycle G1 entry and G0 arrest signatures for each of at least one subsection of a digital image of a biological sample, according to an exemplary embodiment. [Figure 24] FIG. 1 is a block flow diagram of a method for automatically assessing a measure of chromosomal instability for each of at least one subsection of a digital image of a biological sample, according to an exemplary embodiment. [Diagram 25] FIG. 1 is a block flow diagram of a method for automatically assessing measures of cell cycle arrest and chromosomal instability for each of at least one subsection of a digital image of a biological sample, according to an exemplary embodiment. [Figure 26] FIG. 1 is a block flow diagram of a method for determining the presence of one or more cell cycle variants associated with each of at least one subsection of a digital image of a biological sample, according to an exemplary embodiment. [Figure 27-1] FIG. 1 is a schematic diagram of a system and associated method for acquiring and analyzing digital images of a biological sample, according to an exemplary embodiment. [Figure 27-2] FIG. 1 is a schematic diagram of a system and associated method for acquiring and analyzing digital images of a biological sample, according to an exemplary embodiment. [Figure 28] A-E are images of biological samples and corresponding plots showing the determination of a cell cycle transformation function (function 10) to calculate cell cycle G1 entry and G1 / S transformation to find regions harboring BRAF / NRAS mutations, according to an exemplary embodiment. [Figure 29] Heatmap of BRAF / NRAS mutation status in melanoma. [Diagram 30] FIG. 1 is a block flow diagram of a method for automated assessment of measures of cell cycle G1 / S dysregulation and cell cycle G1 phase entry dysregulation for each of at least one subsection of a digital image of a biological sample, according to an exemplary embodiment. [Diagram 31] A-F are images of biological samples and corresponding plots showing the determination of cell cycle transformation functions (Function 11 and Function 12) to calculate cell cycle G1 / S dysregulation signature and cell cycle G2 / M dysregulation signature to identify dysplasia and degree of dysplasia in Barrett's esophagus. [Diagram 32] FIG. 1 is a block flow diagram of a method for automated assessment of measurements of cell cycle G1 / S dysregulation in each of at least one subsection of a digital image of a biological sample to identify areas of dysplasia, according to an exemplary embodiment. [Diagram 33] FIG. 1 is a block flow diagram of a method for automatically assessing measurements of cell cycle G2 / M dysregulation in each of at least one subsection of a digital image to identify the degree of dysplasia in an area of ​​the digital image, according to an exemplary embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0076] Throughout the description, where a composition is described as having, including, or comprising particular components, or where a method is described as having, including, or comprising particular steps, it is further contemplated that there are compositions herein that consist essentially of or consist of the recited components, and that there are methods herein that consist essentially of or consist of the recited processing steps.

[0077] The order of steps or order for performing certain actions is not important so long as the results of the associated processes or their overall functions are operable. Further, two or more steps or actions may be performed simultaneously.

[0078] The reference to any publication herein, for example in the Background section, is not an admission that it serves as prior art with respect to any of the claims presented herein. The Background section is presented for clarity and is not intended as a description of prior art with respect to any claim.

[0079] All references cited herein, including patent applications and publications, are incorporated by reference in their entirety.

[0080] Systems and methods are presented herein for automated cytological evaluation of images of in vitro biological samples. More specifically, in certain embodiments, the present disclosure encompasses systems, methods, and devices for diagnostic evaluation of two-dimensional images of stained biological samples using cell cycle transformation functions.

[0081] FIG. 2 is a schematic diagram of the cell cycle presented for contextual purposes. The diagram shows the various phases of the cell cycle (M, G0, G1, S, and G2), which are described in more detail herein. To divide, a cell must complete several tasks: growing, copying its genetic material (DNA), and physically dividing into two daughter cells. The cell performs these tasks in an organized and predictable series of steps that constitute the cell cycle. The cell cycle is a cycle, not a linear path, because at the end of each revolution, the two daughter cells can start the exact same process all over again. In general, cell division (sometimes called the "cell cycle") occurs in two distinct phases: interphase and mitosis (often called the "M phase," the "mitotic phase," or simply "M"). Interphase is composed of the G1, S, and G2 phases. In the G1 phase (sometimes called the "first gap phase"), the cell is actively growing and is receptive to signals to initiate DNA synthesis. During S phase, the cell synthesizes a complete copy of its DNA in the nucleus and replicates microtubule-organizing structures called centrosomes, which help separate the DNA for M phase. During G2 phase (sometimes called "second gap phase"), the cell actively prepares for M phase and contains twice the normal amount of DNA in the nucleus.

[0082] M phase occurs in a four-stage process: (1) prophase, (2) metaphase, (3) anaphase, and (4) telophase. In prophase (the first stage of cell division), the chromosomes become visible as paired chromatids and the nuclear membrane disappears. The first prophase of meiosis involves reduction division. In metaphase (the second stage of cell division), the chromosomes become attached to the spindle fibers. In anaphase (the third stage of cell division), the chromosomes move to opposite poles of the spindle so that they move away from each other. Finally, in telophase (the final stage of cell division), the chromatids or chromosomes move to opposite ends of the cell such that two nuclei are formed, which allows for the final separation of the parent cell to form two daughter cells.

[0083] After the formation of two daughter cells, some types of cells divide rapidly, and in such cases, the daughter cells may immediately undergo the next series of cell divisions. For example, many cell types in early embryos divide rapidly, as do cells in tumors. Other types of cells divide slowly or not at all. These cells may exit the G1 phase and enter a resting state called G0. Cells in G0 have exited mitosis and are quiescent. That is, in G0, the cell may not be actively preparing to divide, but may simply be performing its function. For example, cells in G0 may conduct signals as neurons or store carbohydrates as hepatocytes. In some cells, G0 is a permanent state, while others may resume division if they receive the right signals from the environment.

[0084] The present specification includes techniques for performing diagnostic tests that calculate biological signatures of various molecular profiles of biological samples, for example, for detecting cellular abnormalities (or lack thereof).In some examples, the disclosure describes methods for using measured cellular features (e.g., nuclear contrast features or nuclear area features, and methods for calculating and quantifying the same, such as those described in U.S. Application Serial No. 15 / 965,664) to characterize or determine the presence, absence, or degree of: (1) to determine the degree of (histological) grade of cancer, thereby distinguishing between areas of high cancer grade (grade 3 or grade 2 cancer depending on clinical findings) and areas with mitotic figures; (2) duration of active cell cycle to determine the degree of Ki67 expression in cancer tissue, thereby identifying areas of high Ki67 expression and areas with mitotic figures; (3) nuclear size skewness and chromosomal instability to determine the degree of DNA ploidy, thereby identifying areas of high DNA ploidy; (4) stromal architectural distortion and cell cycle arrest to determine the degree of stromal tumor-infiltrating lymphocytes (TILs), thereby identifying areas of high stromal TILs. (5) determining the degree of HER2 amplification or expression status, thereby identifying HER2 positive (amplified or overexpressed) regions, specific cell cycle growth 1 (G1) to S phase transition (G1 / S transition) dysregulation and cell cycle arrest signatures; (6) determining the degree of hormone receptor (HR) expression status (combining estrogen receptor (ER) and progesterone receptor (PR) expression status), thereby identifying HR positive (ER or PR overexpressed) regions, specific cell cycle G1 / S transition dysregulation and cell cycle arrest signatures. (7) cell cycle G1 entry and G0 arrest signatures, which determine the degree of "quiescent cell population load" (QPL) and thereby identify regions carrying high QPL; (8) chromosomal instability, which determines the degree of BRCA mutation status and homologous recombination repair deficiency (HRD) status (also referred to as "BRCA-ness") and thereby identify regions carrying BRCA mutations and HRD; and (9) cell cycle arrest and chromosomal instability, which determines the degree of mismatch repair (MMR) functional deficiency and identify regions carrying MMR functional deficiency. In each and all of the above cases, the region of interest is identified using one or more thresholds.

[0085] Step 1 of the diagnostic procedure 100 shown in Figure 1 is receiving a digital image of a segmented biological specimen (or segmenting the image). Segments of the slide image 102 that correspond to regions of the biological specimen imaged (also referred to herein as "subsections" or "fields of view", 104a-104x) are identified by (x,y) coordinate reference points 106, which in this example are the coordinates of the lower left corner of each field of view / subsection.

[0086] Next, in step 2 of procedure 100, a diagnostic score for each visual field is calculated using a nuclear heat / nuclear area algorithm, an example of which is described throughout this specification. The diagnostic score for each visual field is a vector of values, each representing a different measurement or characteristic. In this example, 15 values ​​are shown, e.g., diagnostic score d24.

[0087] In certain embodiments, the thermal and thermodynamic diagnostic parameters (e.g., included in the diagnostic score) are calculated from the ensemble (e.g., ensemble average) thermodynamics of total heat generated in different cellular compartments, their contrast, and their respective area or other size parameters. In certain embodiments, the thermal and thermodynamic diagnostic parameters are calculated as a function of all wavelengths (e.g., a function averaged over the entire range of IR wavelengths naturally emitted by the cell or the cell). An exemplary method described herein includes identifying one or more cells in an image (e.g., an emitted IR image, an H&E image) of a target sample (e.g., a tissue or a body fluid). Each of the cells in the field of view is divided into a cellular region and a nuclear region. The nuclear region feature (a) relates to the ratio of the nuclear area to the nuclear volume projection. For example, the nuclear region feature (a) may be a function of the number of pixels in the nuclear region. The nuclear contrast feature (δh) relates to the hotness of the nuclear region relative to the cellular region, the nuclear chromatin content, and / or the temperature of the nuclear region. For example, the nuclear contrast feature (δh) may be associated with an average pixel intensity in the nuclear region that is higher than normal cytoplasmic intensity (e.g., the average pixel intensity in the cytoplasmic region of normal healthy cells in the field of view). In certain embodiments, the nuclear contrast feature and / or the nuclear region feature is acquired at a wavelength corresponding to the maximum radiant power of the naturally emitted IR. The nuclear contrast feature may be an average of multiple nuclear contrast features acquired at multiple wavelengths of naturally emitted IR. The nuclear region feature may be an average of multiple nuclear region features acquired at multiple wavelengths of naturally emitted IR.

[0088] In certain embodiments, an information surface value (S) for each cell is calculated from the nuclear contrast feature (δh) and the spatial probability density of the nuclear region feature (a) (e.g., Ω(a), Ω(δh), Ω(δhxa), Ω(δh,a)). For example, the information surface value (S) may be:

number

[0089] In certain embodiments, cells in a target sample image are divided into one or more subgroups (e.g., characterized as subgroups) according to information surface values ​​(e.g., dimensionality-reduced information surface values). The subgroups may correspond to different cell cycle stages and / or different intrinsic cell cycle times (L). In certain embodiments, the intrinsic cell cycle times (L) are related to chromatin content in the cells. For example, the intrinsic cell cycle time (L) is a function of nuclear region features and nuclear contrast features.

[0090] In certain embodiments, the thermal and thermodynamic diagnostic parameters include a specificity index (SI). The specificity index is the ensemble average (SI) of the local specific heat between cells in each of the subgroups. <C L >). The subgroups may be determined, for example, by information surface values, extremization of information surface values, and / or one or more characteristic cell cycle times (L) of the information surface values. In certain embodiments, the specificity indicator is the integral of the ensemble average of the local specific heats over the cell cycle stages and / or characteristic cell cycle times, for example,

number

number

[0091] In certain embodiments, the normal state of the target tissue is determined by a diagnostic score (e.g., calculated from thermal and thermodynamic determination parameters). In certain embodiments, the diagnostic score is mapped onto a reference diagnostic scale. The reference diagnostic scale may include pre-assigned values ​​for various types of cancer and their respective subtypes, benign inflammation, and various normal health conditions. In certain embodiments, normal, healthy tissue has a low positive or negative score on the reference diagnostic scale. In certain embodiments, the diagnostic score (F) is a function of the thermal and thermodynamic diagnostic parameters.

number

[0092] 1, the diagnostic score for each visual field is a vector 108 of values, each representing a different measurement or characteristic. In this example, the vector has 15 values ​​(V1-V15), although different embodiments may have a different number of values.

[0093] Moving to step 3 of procedure 100 in FIG. 1, a cell cycle transformation function is constructed from a subset of values ​​selected from vector 108. In certain embodiments, the function includes weights, normalization, etc. The functions described herein can be constructed from at least a subset of the diagnostic scores (vectors of values) for the field. Each function uses a particular subset of the diagnostic score vector as input and calculates an output value called an index. The index (referred to as "diagnostic index" in FIG. 1) is then compared (or mapped) to a corresponding pre-calibrated reference scale to determine whether the index exceeds a threshold (or multiple thresholds defining one or more ranges on the reference scale) associated with a diagnostic test represented by a function (e.g., presence or absence of HER2) corresponding to that particular reference scale. Comparison of the diagnostic index to the reference scale (e.g., mapping the index value to the reference scale to see if it exceeds a threshold) determines whether the field associated with the associated diagnostic score contains a particular characteristic (e.g., presence of HER2, cell cycle dysregulation, or a characteristic quantified by any of the other seven cell cycle transformation function tests detailed herein). The result of comparing the marker against a reference scale determines whether the field associated with that marker should be highlighted in the backprojection (see, e.g., step 4 of FIG. 1; see also FIG. 5 below). In step 4 of the procedure in FIG. 1, the fields that exceed the threshold are graphically displayed (e.g., highlighted) by rendering a two-dimensional shape (e.g., a polygon) on the image that encloses all such fields. In certain embodiments, user interface features such as heat maps and pathology annotations are then presented, for example, via a graphically rendered overlay on the image of the biological sample. In the example shown in FIG. 1, a bold polygonal outline identifies the fields that exceed the threshold.

[0094] Further to the schematic diagram of FIG. 1, the block flow diagram of FIG. 3 illustrates a diagnostic procedure 300 for cytological analysis of an image of an in vitro biological sample, according to an exemplary embodiment. Data representing an image of the biological sample is acquired / received (302), the image including an area divided into one or more subsections (fields of view). The biological sample may include tissue, fluid, or other substances. Each subsection has a location identifier, such as, for example, (x,y) coordinates. A respective diagnostic score is then calculated for each of the one or more subsections (304). In certain embodiments, this involves the use of one or more cell cycle transformation functions (e.g., any of the nine specific functions (306a-306i) described herein). The cell cycle transformations calculated for the various subsections of the image are used to identify regions having morphometric features of clinical interest (308). For example, the function may be solved by mapping a relevant subset of diagnostic score values ​​to variables within the function and comparing / mapping the output of the function (the "diagnostic index") to a specific reference scale associated with that diagnostic test. The test is positive for a desired subsection (field of view) if the index value for that subsection exceeds a desired threshold (e.g., a predetermined threshold). In this manner, each of multiple subsections of the image of the biological specimen may be evaluated and a map of the region of interest may thus be obtained.

[0095] Cancer or any metabolic or inflammatory disease is known to alter various pathways related to key cell cycle proteins (e.g., cell cycle checkpoint proteins, DNA damage pathway proteins, etc.), thereby altering the overall and various phase-specific cell cycle patterns. Often these changes manifest as prominent phenotypic or morphological changes. Collectively, these pattern changes and their manifestations are referred to as cell cycle variants. Advantageously, the techniques described herein may characterize and calculate the extent of one or more of these cell cycle variants from snapshot images or image data (e.g., H&E stained FFPE tissue images) of any fixed tissue and cell population therein in whole or in part, by capturing the impact of these cell cycle variants on the corresponding / equivalent thermodynamic profile(s) of one or more cells. When the relevant cell cycle variants are characterized and appropriately measured / calculated on a vectorial continuous scale or as a function thereof, disease states (e.g., various grades of invasive cancer, DNA ploidy, etc.) and corresponding molecular (e.g., ER, PR, HER2, Ki67, BRCA, MMR, etc.) or immune (TIL) profiles can be identified. For example, cell cycle abnormalities such as S-phase dysregulation may be involved in cancer grade, and the higher the grade, the greater the degree of S-phase dysregulation.

[0096] FIG. 4 shows the characterization and calculation of such deformations (FIGS. 4C and 4D) in terms of functions of numerical vectors extracted from the input images (FIGS. 4A and 4B). Using a threshold on a reference scale (measuring the inverse of S-phase dysregulation as in our case), the calculated values ​​can be assigned to predictions of high versus low cancer grade (FIGS. 4E versus 4F). This calculation of predictive diagnostic values ​​(herein from a portion of the full image) can also be tagged along with the coordinates of all slide images / image data or images / image data containing larger masses of the same tissue, and these predictions can be presented as rectangular diagnostic boundaries drawn on the original larger image (containing larger masses of the same tissue) at the position specified by the above-mentioned coordinates (FIGS. 4E, 4F). All such selected fields containing areas of diagnostic interest are marked with a rectangular boundary (in red) respectively (FIG. 5B) and can be clustered and enclosed in a larger polygon (or convex hull of the coordinates of the selected fields) (in blue) as an extended area / diagnostic boundary of diagnostic interest (FIG. 5C). Similarly, by converting the continuous functions characterizing cell cycle transformations into a color map (0-1 scale, binned and mapped to a color scheme), the heatmap overlay can be rendered onto the original larger image (containing a larger mass of the same tissue) using the coordinates (Figure 5D). All these user interface (UI) features, diagnostic fields / areas of interest, extended diagnostic boundaries / extended areas of interest, and heatmaps, together with the prediction of the pathology as a corresponding pathology annotation (e.g., invasive carcinoma in Figure 5), are highly useful for pathology review. Currently, it is believed that there are no other methods (IHC, ISH, FISH, NGS, FACS, IF imaging, etc.) that can directly calculate cellular molecular profiles without the use of one or more molecular-specific tags (antibodies, fluorophores, etc.). Other methods usually require complex experimental procedures, and many such methods are purely qualitative or semi-quantitative.

[0097] Briefly summarized, Figures 4A and 4B show input images / image data / fields containing high-grade and low-grade cancer, respectively. Figures 4C and 4D show functional calculations of corresponding cell cycle variants (e.g., S-phase dysregulation) as diagnostic indicators using portions of vector arrays that numerically characterize the cellular / biological and biological information contained within the field, predicting the degree of pathology (e.g., high or low cancer grade, or cancer grade 3 (high), grade 2 (high), and grade 1 (low)). The curves represent the corresponding visualization of each variant in 2D (C_L,L) space, where (C_L,L) are quantities described throughout this specification. Figures 4E and 4F show the identification and annotation of regions of interest on a larger / whole slide image (containing large portions of the same tissue or the entire tissue) using scores as a function of (x,y) coordinates.

[0098] FIG. 5A shows the inputted whole slide image, and FIG. 5B and FIG. 5C show the whole slide image annotated with the field of view (shown in red) containing the region of diagnostic interest and the diagnostic boundary (shown in blue) containing the extended region of interest. In this case, the main or extended region of the diagnostic indicator is for pathology annotation referring to the tissue region containing invasive cancer. FIG. 5D shows a heat map with a color bar corresponding to the degree of invasive cancer at any (x,y) point in the whole slide image. As shown in FIG. 1 and FIG. 3, the exemplary diagnostic process uses the following 12 functions to predict a particular diagnostic entity as defined in Table 1 below. It should be noted that the variables referenced in the last column are described throughout this specification. [Table 1]

[0099] Function 1: Dysregulation of the S phase of the cell cycle may contribute to cancer grade. The higher the degree of dysregulation, the higher the risk of retaining damaged DNA and therefore the higher the risk of developing high-grade cancer. In certain embodiments, cancer grade, which is an important part of pathology diagnosis, is a measurement by which a pathologist evaluates how aggressive a cancer may be. Figures 3 and 4 present an exemplary general methodology of how any such dysregulation present in an image / portion of an image / image data may first be calculated as a diagnostic score (including a vector array that numerically characterizes the cellular / biological and biological information contained in the input image / image data) of pixel data (nuclear area and contrast), and then map a portion of the diagnostic score to a specific diagnostic index value and back-project it to the entire slide image using the original image (x,y) coordinates. Figure 6 presents a specific example of how S phase dysregulation capturing cancer grade can be calculated with function F1, which essentially measures the inverse of such dysregulation, so that above a certain threshold, a low cancer grade (grade 1) can be confirmed, and below the threshold, a high grade (grade 2 and grade 3) can be confirmed. It should be noted that multiple thresholds are used to separate high-grade cancers into grades 2 and 3. A unique aspect of the construction of Function 1 is that in (C_L,L) space, Function 1 exhibits a visible variance of curvature indicative of such dysregulation.

[0100] Figures 6A and 6B show input images / image data / fields containing high-grade and low-grade cancer, respectively. Figures 6C and 6D show function calculations of corresponding cell cycle deformations manifested in S-phase dysregulation as diagnostic indicators using a portion of the vector array that numerically characterizes the cellular / biological and biological information contained within the field, which predicts the degree of cancer grade (e.g., high / low or cancer grade 3 (high), grade 2 (high), and grade 1 (low)). The curves represent the corresponding visualization of each deformation in 2D (C_L,L) space. Figure 6E is a table showing the determination of high cancer grade vs. low cancer grade using function 1 below and above a certain threshold.

[0101] Function 2: Ki67 protein is a marker of cell activity and proliferation. Ki-67 protein is present in all active phases of the cell cycle (G1, S, G2, and mitosis), but absent in resting / quiescent cells (G0). The proportion / percentage of Ki67-positive tumor cells may correlate with the clinical course of cancer. For example, in breast cancer, Ki67 identifies a highly proliferative subset of ER-positive breast cancer patients that will benefit more from adjuvant chemotherapy. In essence, similar to cancer grade, the percentage of Ki67 marker measures the overall aggressiveness of the tumor in terms of tumor growth. Thus, the duration of the active phase of the cell cycle and its variants are naturally involved in the determination of Ki67 expression. Figures 3 and 4 explain the methodology of how any such deformations present in the image / portion of an image / image data can first be calculated as a diagnostic score (comprising a vector array that numerically characterizes the cellular / biological and biological information contained in the input image / image data) of pixel data (nuclear area and contrast), and then the portion of the diagnostic score can be mapped to a specific diagnostic index value and back-projected onto the entire slide image using the original image (x,y) coordinates. Figure 7 shows a concrete example of how the deformations of the duration of the cell cycle active phase capturing %Ki67 expression can be calculated with Function 2. This basically measures the inverse of M-phase specific cellular activity that affects cell cycle entry patterns, above a certain threshold a low %Ki67 can be confirmed, and below a threshold a high %Ki67 can be confirmed. A unique aspect of the construction of Function 2 is that in (C_L,L) space, Function 2 exhibits visibly distinct entry and exit point signatures indicative of such deformations.

[0102] Figures 7A and 7B show input images / image data / fields of cancer tissue containing high and low Ki67 expression, respectively. Figures 7C and 7D show function calculations of corresponding variations of cell cycle active phase length as a novel diagnostic indicator using a portion of the vector array that numerically characterizes the cellular / biological and biological information contained within the field, thereby predicting the degree of high and low %Ki67 expression. The curves represent the corresponding visualization of each variation in 2D (C_L,L) space. Figure 7E is a table showing the determination of high vs. low %Ki67 expression status of cancer using function 2 below and above a certain threshold.

[0103] Function 1 and Function 2 Outcomes / Random Variables: High grade and high %Ki67 expression may both be associated with high mitotic figures. Mitotic figures are cells in the process of dividing to produce two new cells. This process is called mitosis. Pathologists can see mitotic figures when they view tissue under a microscope. Pathologists also routinely report and score the number of mitotic figures present in tissue samples for the purpose of calculating cancer grade. Such regions of interest containing mitotic figures can be identified from the heatmaps of Function 1 or Function 2 by pointing to areas of high grade or high %Ki67 expression, for example, as shown in FIG. 8.

[0104] Figure 8A shows a heat map of cancer grade using Function 1 values ​​as a function of (x,y) coordinates across the slide image. Figure 8B shows a heat map of %Ki67 expression using Function 2 values ​​as a function of (x,y) coordinates across the slide image. Figure 8C shows examples of enlarged regions across the slide image containing mitotic figures, identified by looking for areas of high grade or high %Ki67 expression from the heat maps of Figures 8A and 8B.

[0105] Function 3: Ploidy of cancer cells refers to the amount of DNA they contain. When cells have a normal amount of DNA, they are said to be diploid. These cancers tend to grow and spread slowly. When the amount of DNA is abnormal (either numerically or structurally, by gaining or losing whole chromosomes or parts of chromosomes), the cells are called aneuploid. When cells contain more than one set of chromosomes, but always contain strictly multiple uniploid numbers (so that chromosomes are balanced), they are called polyploid. There is a good correlation between DNA ploidy and histological grade, and DNA ploidy provides clinically useful predictive information for some patients. Most low-stage tumors are diploid, while high-stage tumors are non-diploid. Cells carrying high DNA ploidy show significantly distorted / disordered nuclear and chromatin material. Herein, DNA ploidy refers to such deviations from the diploid state. Figure 9 shows a specific example of how nuclear size skewness and chromosomal instability, which characterize DNA ploidy, can be calculated with Function 3. This essentially measures cellular chromosomal instability that affects cell cycle exit patterns, such that above a certain threshold, high DNA ploidy can be identified, and below that threshold, low DNA ploidy can be identified. A unique aspect of the construction of Function 3 is that in (C_L,L) space, Function 3 exhibits visibly distinct exit signatures indicative of such variations.

[0106] Figures 9A and 9B present images / image data / fields of cancer tissue carrying low and high DNA ploidy, respectively. Figures 9C and 9D show the corresponding function calculations of nuclear size skewness and chromosomal instability as novel diagnostic indicators using a portion of the vector array that numerically characterizes the cellular / biological and biological information contained within the field, thereby predicting the degree of DNA ploidy. The curves represent the corresponding visualization of each transformation in 2D (C_L,L) space. Figure 9E is a table showing the determination of high DNA ploidy status vs. low DNA ploidy status of cancer using function 3 below and above a certain threshold.

[0107] Function 4: Tumor infiltrating lymphocytes (TILs) refers to all lymphocyte cell populations that have invaded tumor tissue. TILs have been described in many solid tumors, including breast cancer, and have emerged as important biomarkers in predicting treatment efficacy and outcome. Pathologists can see TILs under a microscope. Figure 10 shows an example of how stroma structural distortion and cell cycle arrest, which characterize stromal TIL expression, can be calculated with Function 4. This essentially measures deviations in stroma shape and cellular activity, so that above a certain threshold, low stromal TIL expression can be confirmed, and below the threshold, high stromal TIL expression can be confirmed. A unique aspect of the construction of Function 4 is that in (C_L,L) space, Function 4 exhibits visibly different shapes that indicate such deformations.

[0108] 10A and 10B show input images / image data / fields of cancer tissue carrying low and high stromal TILs, respectively. FIG. 10C and 10D show function calculations of corresponding architectural distortions and global measures of cell cycle arrest in the stroma as novel diagnostic indicators using a portion of the vector array that numerically characterizes the cellular / biological and biological information contained in the field, thereby predicting the degree of stromal TIL expression. The curves represent the corresponding visualization of each deformation in 2D (C_L,L) space. FIG. 10E is a table showing the determination of low vs. high stromal TIL status using function 4 below and above certain thresholds.

[0109] Function 5: HER2 / human epidermal growth factor receptor 2 is a protein that promotes the growth of cancer cells. In about one-fifth of breast cancers, the cancer cells have an extra copy of the gene that produces the HER2 protein. HER2 amplification / overexpression testing is a routine reflex test performed to measure the prognosis of cancer. HER2-positive cancers have a poor prognosis and can often metastasize to the brain. Figure 11 shows a concrete example of how function 5 can calculate a cell cycle G1 / S transition dysregulation and arrest signature that characterizes HER2 amplification / overexpression. This essentially measures the G1 / S transition-specific cellular activity deviation, such that above a certain threshold a HER2-positive state (amplified / overexpressed) can be confirmed, and below the threshold a HER2-negative state (non-amplified / non-overexpressed) can be confirmed. A unique aspect of the construction of function 5 is that in (C_L,L) space the function exhibits visibly different shapes indicative of such deformations.

[0110] 11A and 11B show images / image data / fields of cancer tissues with or without the presence / absence of HER2 amplification / overexpression (referred to as positive / negative), respectively. FIG. 11C and 11D show function calculations of corresponding cell cycle G1 / S dysregulation and arrest signatures as novel diagnostic indicators using a portion of the vector array that numerically characterizes the cellular / biological and biological information contained within the field, thereby predicting the degree of HER2 amplification / overexpression status. The curves represent the corresponding visualization of each transformation in 2D (C_L,L) space. FIG. 11E is a table showing the determination of positive versus negative HER2 status using function 5 below and above certain thresholds.

[0111] Function 6: Estrogen and progesterone receptors are found in breast cancer cells that depend on estrogen and related hormones to grow. Patients diagnosed with invasive breast cancer or recurrent breast cancer routinely have their tumors tested for estrogen (ER) and progesterone (PR) receptors, collectively referred to as hormone receptors (HR). Figure 12 shows an example of how Function 6 can calculate a cell cycle G1 / S transition dysregulation and arrest signature that characterizes HR overexpression. This essentially measures G1 / S transition-specific cellular activity deviations such that above a certain threshold a HR positive state (overexpression) can be confirmed, and below the threshold a HR negative state (non-overexpression). A unique aspect of the construction of Function 6 is that in (C_L,L) space Function 6 exhibits visibly different shapes indicative of such deformations.

[0112] Figures 12A and 12B show the input image / image data / field of cancer tissue, regardless of the absence / presence of HR overexpression (referred to as negative / positive), respectively. Figures 12C and 12D show the function calculation of the corresponding cell cycle G1 / S dysregulation and arrest signature as a novel diagnostic indicator using a portion of the vector array that numerically characterizes the cellular / biological and biological information contained in the field of view, thereby predicting the degree of HR overexpression status. The curves represent the corresponding visualization of each transformation in 2D (C_L,L) space. Figure 12E is a table showing the determination of HR positive status vs. HR negative status using function 6 below and above a certain threshold.

[0113] Function 7: Quiescent state is a state of reversible growth arrest in which cells are not actively dividing but retain the ability to re-enter the cell cycle upon receiving the appropriate stimuli. Quiescent cell populations are believed to be resting in the G0 phase of the cell cycle and can cause cancer to recur when chemotherapy or other targeted therapies are discontinued. Thus, in certain embodiments, such measurements of quiescent population load (QPL) can be important markers for evaluating prognosis or treatment efficacy. Figure 13 shows an example of how function 7 can calculate cell cycle G1 entry and G0 arrest signatures that characterize QPL. This essentially measures G1-specific cellular activity deviations such that above a certain threshold a high QPL state can be confirmed and below a threshold a low QPL state (non-amplified / non-overexpressed). A unique aspect of the construction of function 7 is that in (C_L,L) space function 7 exhibits visibly distinct shapes indicative of such variations.

[0114] Figures 13A and 13B show images / image data / fields of cancer tissue with low versus high quiescent population load (QPL). Figures 13C and 13D show function calculations of corresponding cell cycle G1 entry and G0 arrest signatures as novel diagnostic indicators using a portion of the vector array that numerically characterizes the cellular / biological and biological information contained in the field, thereby predicting the degree of quiescent population present in the tissue. The curves represent the corresponding visualization of each transformation in 2D (C_L,L) space. Figure 13E is a table showing the determination of low versus high QPL using function 7 below and above a certain threshold.

[0115] Function 8: "BRCA" is an abbreviation for "breast cancer genes". BRCA1 and BRCA2 are two different genes that, when carrying certain mutations, are known to affect the likelihood of developing breast cancer. Homologous recombination is a very precise DNA repair mechanism. When the BRCA gene is mutated or the tumor has other vulnerabilities that damage the DNA repair mechanism, the condition is identified as Homologous Recombination Deficient (HRD). When HRD is present, the cancer cells are particularly vulnerable to certain targeted therapies, such as PARP inhibitor therapy. Thus, in certain embodiments, identifying the presence of BRCA mutations or HRD is important to select the appropriate treatment. Figure 14 shows an example of how chromosomal instability signatures that characterize BRCA mutation and HRD status can be calculated in function 8. This essentially measures a divergent shift in cellular activity, which can confirm a BRCA-positive status (mutated) and, below a threshold, a BRCA-negative status (non-mutated). A particular aspect of the construction of function 8 is that in (C_L,L) space, function 8 exhibits visibly different shapes that indicate such deformation. HRD is calculated from the local variations of function 8.

[0116] Figure 14A and Figure 14B show the input image / image data / field of cancer tissue with wild type / mutated BRCA1 or BRCA2 genes (referred to as BRCA negative / positive), respectively. Figure 14C and Figure 14D show the function calculation of the corresponding chromosomal instability signature as a novel diagnostic index using a part of the vector array that numerically characterizes the cellular / biological and biological information contained in the field of view, which predicts the degree of BRCA mutation status. The curves represent the corresponding visualization of each transformation in 2D (C_L,L) space. Figure 14E is a table showing the determination of BRCA negative vs. positive status using function 8 below and above a certain threshold. Figure 14F shows how the HRD status can be calculated from the local variation / variability of function 8, where the HRD status is shown as a heat map as a function of (x,y) coordinates.

[0117] Function 9: Mismatch repair (MMR)-deficient cells usually have many DNA mutations, which can lead to cancer. The presence of functional tumor MMR deficiency can be assessed with any tumor microsatellite instability (MSI) test. MMR deficiency is common in colon cancer, other types of gastrointestinal cancer, and endometrial cancer, but can also be found in breast cancer, prostate cancer, bladder cancer, and thyroid cancer. Figure 15 shows an example where function 9 can calculate cell cycle arrest and chromosomal instability signatures that characterize MMR deficiency states. This essentially measures the variation in shape that represents cellular activity, so that above a certain threshold, an MR normal state can be confirmed, and below the threshold, an MMR deficiency state can be confirmed. A unique aspect of the construction of function 8 is that in (C_L,L) space, function 9 exhibits visibly different shapes that indicate such variations.

[0118] Figure 15A and Figure 15B show the input image / image data / field of view of cancer tissue with MMR-deficient vs. MMR-normal status. Figure 15C and Figure 15D show the function calculation of the corresponding cell cycle arrest and chromosomal instability signatures as novel diagnostic indicators using a portion of the vector array that numerically characterizes the cellular / biological and biological information contained in the field of view, thereby predicting the degree of MMR-deficient status. The curves represent the corresponding visualization of each transformation in 2D (C_L,L) space. Figure 15E is a table showing the determination of MMR-deficient vs. MMR-normal status using function 9 below and above a certain threshold.

[0119] Function 10: Malignant melanoma is a malignant tumor arising from melanocytes of the skin and has a high propensity to metastasize. In malignant melanoma, activating mutations of the oncogenes BRAF and NRAS are very common (e.g., the prevalence of BRAF600 mutations is nearly 50% and the prevalence of NRAS mutations is 20%). Identification of the presence of BRAF and / or NRAS mutations in melanoma specimens provides both prognostic and therapeutic guidance. Both of these mutations affect the mitogen-activated protein kinase (MAPK) pathway, promoting tumor growth and promoting disease progression. From a cell cycle perspective, this can directly lead to a highly dysregulated cell cycle, especially at cell cycle entry and G1 / S transition. Figure 28 shows an example of how dysregulation of cell cycle G1 entry and G1 / S transition, which characterizes BRAF and / or NRAS mutations, can be calculated from function 10. This essentially measures the inverse of the collective strength of those variations, such that above a threshold a BRAF and / or NRAS mutant state is confirmed, below the threshold a wild-type (common) state is confirmed. A unique aspect of the construction of function 10 is that in (C_L,L) space, function 10 exhibits a visible divergent signature indicative of such variations. Regions of interest containing BRAF and / or NRAS mutations can be identified from a heat map of function 10, as described with reference to FIG. 29.

[0120] Figure 28A and Figure 28B show the input image / image data / field of cancer tissue (in this case, malignant melanoma) carrying BRAF / NRAS mutant and wild type, respectively. Figure 28C and Figure 28D show the function calculation of the corresponding cell cycle G1 entry and G1 / S transformation as a novel diagnostic indicator using a part of the vector array that numerically characterizes the cellular / biological and biological information contained in the field of view, which predicts the degree of BRAF / NRAS mutation status. The curves represent the corresponding visualization of each transformation in 2D (C_L,L) space. Figure 28E is a table showing the determination of the BRAF / NRAS mutation status vs. wild type status of cancer (here, malignant melanoma) using function 10 below and above a certain threshold.

[0121] FIG. 29 shows a heat map of BRAF / NRAS mutation status in malignant melanoma using function 10 values ​​as a function of (x,y) coordinates across slide images (original image shown in inset).

[0122] Functions 11 and 12: Barrett's esophagus is a pathology of the esophagus with the complication of gastroesophageal reflux disease (GERD) and is associated with an increased risk of esophageal adenocarcinoma (cancer of the esophagus). Approximately 30 million people in North America suffer from GERD, which is the most common chronic gastrointestinal disease. Approximately 5% of patients with chronic GERD or inflammation of the esophagus are known to develop Barrett's esophagus. GERD is a condition in which stomach contents, including stomach acid, reflux into the esophagus. Barrett's esophagus is the result of chronic GERD, which can occur when the lining of the esophagus (including the squamous mucosa) heals abnormally and changes from skin cells (squamous cells) to cells with characteristics of enterocytes (e.g., specialized columnar cells with intestinal metaplasia). Due to the high risk of developing cancer, patients diagnosed with Barrett's esophagus may require ongoing surveillance, such as long-term follow-up / repeated esophagogastroduodenoscopy. In Barrett's esophagus, when precancerous cells (dysplasia) are diagnosed, endoscopic treatment can be recommended as a safe and effective method to prevent progression to cancer. Detection of dysplasia and the degree / level of dysplasia are important prognostic factors for Barrett's esophagus. From the viewpoint of cell cycle, the occurrence of dysplasia may result from the lack of tumor suppressor effect, which directly transitions to highly dysregulated cell cycle, especially in the cell cycle G1 / S and G2 / M transitions.

[0123] FIG. 31 shows a specific example of calculating dysregulation at cell cycle G1 / S and G2 / M transition stages (e.g., characterizing the presence / absence of dysplasia and the degree of dysplasia state) using function 11 and function 12, respectively. Function 11 and function 12 measure the inverse and direct correlation of the intensity of G1 / S and G2 / M transformation, respectively. In particular, above the threshold of function 11, the presence of dysplasia can be confirmed, and below the threshold of function 11, the absence of dysplasia can be confirmed. The degree of dysplasia (e.g., low, medium, or high) corresponds to a certain range of thresholds as the value of function 12 increases. A particular aspect of the construction of function 11 and function 12 is that in (C_L,L) space, function 11 and function 12 show visibly different divergent signatures indicative of G1 / S and G2 / M transformation.

[0124] 31A and 31B show input images / image data / fields of precancerous esophageal tissue (in this case, dysplastic Barrett's esophagus) and benign esophageal tissue (in this case, non-dysplastic Barrett's esophagus), respectively. FIG. 31C and 31D show function calculations of corresponding cell cycle G1 / S and G2 / M dysregulation (using a portion of vector arrays that numerically characterize the cellular / biological and biological information contained in the field) as novel diagnostic indicators, which predict (i) the presence or absence of dysplasia (e.g., using function 11), and (ii) the degree / degree of dysplasia (e.g., using function 12). The curves shown in FIG. 31C and 31D represent the corresponding visualization of each transformation in 2D (C_L,L) space. FIG. 31E is a table showing the determination of the presence or absence of esophageal dysplasia based on whether function 11 is below or above a certain threshold. FIG. 31F is a table showing the determination of the degree / level of dysplasia in a dysplastic esophagus using function 12 for a particular range of thresholds.

[0125] 17-24 show exemplary methods involving the calculation of measured cell cycle variations and the corresponding functions from Table 1 above (specifically, functions 1-9).

[0126] In certain embodiments, the network environment shown in Figure 30 of U.S. Patent No. 10,535,434 and described therein (e.g., at column 23) may be used with the methods and systems described herein. Additionally, in certain embodiments, the computing device shown in Figure 31 of U.S. Patent No. 10,535,434 and described therein (e.g., at columns 23-27) may be used with the methods and systems described herein.

[0127] In certain embodiments, one or more of the devices described in U.S. Pat. No. 10,535,434 may be used in the methods and systems described herein, such as, for example, an exemplary device for detecting molecular signatures of cancer cells from skin lesions (FIGS. 20-22 of U.S. Pat. No. 10,535,434), an exemplary whole-body scanner for detecting cancer (FIG. 23 of U.S. Pat. No. 10,535,434), an exemplary device for detecting cancer from an internal site of a subject (e.g., not from the skin) (FIG. 24 of U.S. Pat. No. 10,535,434), an exemplary device for detecting breast cancer using a biopsy needle (FIG. 25 of U.S. Pat. No. 10,535,434), an exemplary device for detecting the borders of cancer during surgery (FIG. 26 of U.S. Pat. No. 10,535,434), and the exemplary endoscopy, colonoscopy, and colposcopy devices described in U.S. Pat. No. 10,535,434.

[0128] In certain embodiments, the methods and systems described herein may be used for real-time or near real-time microscopy, for example, for in vivo or ex vivo (e.g., biopsy) analysis during surgery. Components may include, for example, real-time video surveillance devices, scanners, sensors, virtual reality (VR) or augmented reality (AR) headsets or other displays, etc., for rendering or displaying overlays generated as described herein in real-time or near real-time.

[0129] 27 is a schematic diagram of a system 2700 and associated methods for acquiring and analyzing digital images of a biological sample, according to an exemplary embodiment. As shown in item 2710, the system includes one or more scanners, sensors, VR or AR headsets, other displays, and associated equipment for acquiring digital images of the biological sample and displaying the images to one or more users. The associated equipment may include, for example, endoscopes, syringes, sponges, forceps, scalpels, microscope slides, etc. The system also includes a processor, network, and associated components 2730 as described in FIGS. 30 and 31 of U.S. Pat. No. 10,535,434.

[0130] Element 2720 of FIG. 27 illustrates an exemplary step of a method involving features of system 2700 of FIG. 27. In an exemplary method, a user scans a sample, e.g., ex vivo, e.g., obtained by biopsy as part of a surgical or non-surgical procedure. In other particular embodiments, the sample is imaged (e.g., scanned) in vivo (e.g., without isolating the tissue or other biological sample from the body). In certain embodiments, the acquisition or access of the biological sample is performed for real-time or near real-time image analysis, e.g., during a surgical procedure to identify cancer borders or for another medical purpose. In other embodiments, the acquisition or access of the biological sample is performed for non-real-time analysis.

[0131] The digital image may be a still image, may be part of an array or series of images, or may be part of a video stream (e.g., a live video stream) of images of the biological sample. The digital image is transmitted to a processor of a computing device for analysis. For example, the processor, network, and related components 2730 of Figures 30 and 31 of U.S. Patent No. 10,535,434 may be used to perform any of the steps of the digital image analysis methods described herein, e.g., method 2740 of Figure 16. The processor renders a report or interactive results (e.g., overlay) that are displayed to the user via the device 2710. The processor may receive user input and update the interactive results accordingly, e.g., by receiving and displaying annotations on the overlay and highlighting areas of interest identified by the user via the user interface device (e.g., by clicking a mouse, touching a touchpad display, hovering a cursor over a particular location, etc.). Graphical user interface widgets, e.g., text boxes, sliders, dropdown menus, buttons, icons, menus, radio buttons, scroll bars, windows, etc., may be presented to the user for selection and interaction via the display.

[0132] Although the devices, systems, and methods have been particularly shown and described with reference to certain preferred embodiments, it should be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the subject matter recited in the appended claims.

[0133] FIG. 30 is a flow diagram of an exemplary process for automatically assessing the measurements of cell cycle G1 / S dysregulation and cell cycle G1 phase entry dysregulation for each of at least one subsection of a digital image. For convenience, the process will be described as being performed by one or more computer systems at one or more locations. First, the system receives a digital image of a biological sample by a processor. Then, the system automatically assesses the measurements of cell cycle G1 / S dysregulation and cell cycle G1 phase entry dysregulation for at least one subsection of the digital image to identify regions containing one or more of BRAF and NRAS mutations.

[0134] In some implementations, the system identifies regions that contain only BRAF gene mutations.

[0135] In some implementations, the system identifies regions that contain only NRAS gene mutations.

[0136] In some implementations, the system identifies regions that contain mutations in both the BRAF and NRAS genes.

[0137] The system can be applied to any of a variety of possible applications, for example, the system can be applied to perform prediction of BRAF / NRAS mutation status from digitized whole-slide images of H&E stained melanomas.

[0138] By comparison, in 2020, there were more than 300,000 new cases of malignant melanoma worldwide, resulting in approximately 57,000 deaths. In the United States, it is the fifth most common cancer, with an overall 5-year survival rate of 93.7% and a less favorable 31.9% survival rate for patients with distant disease. The most frequent genetic abnormalities in cutaneous melanoma are BRAF and NRAS mutations, which guide treatment and prognostic therapies. Testing is usually centered around sequencing and immunohistochemistry, which can be time-consuming, tissue-destructive, costly, and not universally accessible. The system herein circumvents these issues.

[0139] To demonstrate the efficacy of the system described herein, a study was designed to use the system on unannotated whole slide images (WSIs) of H&E stained melanoma. Unlike traditional artificial intelligence-based approaches, the system uses underlying predictive criteria based on deterministic, hard-coded observational relationships on a continuous scale (digital biomarkers). The system mapped these features to tumor-associated variations in biological pathways and converted them into a numerical scale that allows for the identification of BRAF and NRAS mutation status. Seventy-one independent cases of cutaneous melanoma from biopsy, resection, and lymph node specimens (139 WSIs in total; 1–3 slides / case, median=2) were analyzed. There were roughly equal numbers of wild-type, BRAF, and NRAS mutated cases (n=22, 19, and 24, respectively).

[0140] The system demonstrated excellent performance in predicting wild-type vs. BRAF / NRAS mutated cases: Cohen's kappa = 0.90 (95% CI 0.78-1.01), p < 0.0001, 95% accuracy, 95% sensitivity, 95% specificity, 98% PPV, and 91% NPV. A patient case-level attrition rate of 8% (cases with no analyzable slides) was observed.

[0141] The system was able to classify melanoma specimens into wild-type versus BRAF / NRAS mutant cases with high accuracy. The system may streamline current clinical laboratory workflows by shortening diagnostic turnaround times and reducing costs while sparing tissue.

[0142] 32 is an exemplary flow diagram of a process for automatically assessing a measure of cell cycle G1 / S dysregulation for each of at least one subsection of a digital image. For convenience, the process will be described as being performed by one or more computer systems at one or more locations. First, the system receives, by a processor, a digital image of a biological sample. Then, the system automatically assesses a measure of cell cycle G1 / S dysregulation for each of at least one subsection of the digital image to identify areas of dysplasia.

[0143] 33 is a flow diagram of an exemplary process for automatically assessing a measure of cell cycle G2 / M dysregulation for each of at least one subsection of a digital image. For convenience, the process will be described as being performed by one or more computer systems at one or more locations. First, the system receives, by a processor, a digital image of a biological sample. Then, the system automatically assesses a measure of cell cycle G2 / M dysregulation for each of at least one subsection of the digital image to identify the degree of dysplasia in the area of ​​the digital image.

[0144] Embodiments of the subject matter and functional operations described herein can be implemented in digital electronic circuitry, tangibly embodied computer software or firmware, computer hardware (including structures disclosed herein and structural equivalents thereof), or one or more combinations thereof. Embodiments of the subject matter described herein can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions coded on a tangible, non-transitory storage medium for execution by or control of the operation of a data processing apparatus. The computer storage medium can be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or one or more combinations thereof. Alternatively or additionally, the program instructions can be coded in an artificially generated propagated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal, which is generated to code information for transmission to a receiving apparatus suitable for execution by a data processing apparatus.

[0145] A computer program (which may also be referred to or described as a program, software, software application, app, module, software module, script, or code) can be written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages) and can be deployed in any form (e.g., as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment). A program may correspond to a file in a file system, but need not. A program can be stored as part of a file that holds other programs or data, e.g., one or more scripts stored in a markup language document, a single file dedicated to the program in question, or multiple coordinated files, e.g., files that store one or more modules, subprograms, or portions of code. A computer program can be deployed to be executed on one computer, or on multiple computers at one site or distributed across multiple sites and interconnected by a data communications network.

[0146] The processes and logic flows described herein may be performed by one or more programmable computers executing one or more computer programs that perform functions by manipulating input data and generating output. The processes and logic flows may also be performed by special purpose logic circuitry, such as an FPGA or an ASIC, or by a combination of special purpose logic circuitry and one or more programmed computers.

[0147] A computer suitable for executing a computer program can be based on a general-purpose or special-purpose microprocessor, or both, or other types of central processing unit. Typically, the central processing unit will receive instructions and data from a read-only memory or a random-access memory, or both. The essential elements of a computer are a central processing unit for implementing or executing instructions, and one or more memory devices for storing instructions and data. The central processing unit and memory may be supplemented by, or incorporated in, special-purpose logic circuitry. Typically, a computer will also include, or be operatively connected to, one or more mass storage devices (e.g., magnetic, magneto-optical, or optical disks) for storing data, or to receive data therefrom, or to transfer data thereto, or both. However, a computer need not have such devices. In addition, a computer can incorporate other devices, such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a global positioning system (GPS) receiver, or a portable storage device (e.g., a universal serial bus (USB) flash drive, to name just a few).

[0148] Computer-readable media suitable for storing computer program instructions and data include all types of non-volatile memory, media, and memory devices (including, by way of example, semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magnetic optical disks; and CD-ROM and DVD-ROM disks).

[0149] To provide for user interaction, embodiments of the subject matter described herein can be implemented with a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user, and a keyboard and pointing device (e.g., a mouse or trackball) by which the user can provide input to the computer. Other types of devices can be used to provide for user interaction as well, for example, feedback provided to the user can take any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback), and input from the user can be received in any form, including acoustic, speech, or tactile input. In addition, a computer can interact with a user by sending documents to a device used by the user and receiving documents from it, for example, by sending a web page to a web browser on a user's device in response to a request received from the web browser. A computer can also interact with a user by sending a text message or other form of message to a personal device (e.g., a smartphone running a messaging application) and receiving a response message from the user in return.

[0150] A data processing apparatus for implementing machine learning models may also include special purpose hardware accelerator units for handling a common and computationally intensive part of the machine learning training or generation, i.e. inference, workload.

[0151] The machine learning model can be implemented and deployed using a machine learning framework (e.g., the TensorFlow framework or the Jax framework).

[0152] An embodiment of the subject matter described herein may be implemented in a computing system that includes a back-end component (e.g., as a data server), or includes a middleware component (e.g., an application server), or includes a front-end component (e.g., a client computer having a graphical user interface, a web browser, or an app through which a user may interact with an implementation of the subject matter described herein), or any combination of one or more such back-end, middleware, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks are a local area network (LAN) and a wide area network (WAN), e.g., the Internet.

[0153] A computing system may include clients and servers. Typically, clients and servers are remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. In some embodiments, a server sends data (e.g., HTML pages) to a user device, for example, to display the data on the device acting as a client and receive user input from interaction with the device. Data generated at the user device (e.g., results of user interaction) can be received at the server from the device.

[0154] Although this specification contains many specific implementation details, these cannot be interpreted as limitations on the scope of any invention or what may be claimed, but rather should not be interpreted as descriptions of features that may be specific to a particular embodiment of a particular invention. Certain features described in the context of separate embodiments herein may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented in multiple embodiments individually or in any suitable subcombination. Furthermore, although features may be described above as functioning in a particular combination and may be claimed on their own from the outset, one or more features from a claimed combination may in some cases be deleted from the combination, and the claimed combination may be directed to a subcombination or a variation of the subcombination.

[0155] Similarly, although operations are depicted in the figures and claimed in a particular order, this should not be understood as requiring that such operations be performed in the particular order or sequential order shown, or that all of the operations shown be performed, to achieve desirable results. In certain situations, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system modules and components in the above embodiments should not be understood as requiring such separation in all embodiments, and it should be understood that the program components and systems described may generally be integrated into a single software product or packaged into multiple software products.

[0156] Specific embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims. For example, the actions recited in the claims can be performed in a different order to still achieve desirable results. By way of example, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some cases, multitasking and parallel processing may be advantageous.< / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c> < / c>

Claims

1. 1. A computer-implemented method comprising: receiving, by a processor of a computing device, a digital image of a biological sample, the digital image including an area divided into a plurality of subsections; and further comprising: calculating, by the processor, a diagnostic score for a first subsection of the plurality of subsections, the diagnostic score comprising a plurality of values ​​collectively related to nuclear heat and nuclear area of ​​one or more cells imaged in the first subsection of the biological specimen; and executing, by the processor, a cell cycle transformation function to generate a diagnostic index corresponding to the first subsection, wherein the cell cycle transformation function identifies one or more cell cycle transformations based on a subset of the plurality of values ​​of the diagnostic score for the first subsection, and further comprising: mapping, by the processor, the diagnostic indicator of the first subsection to a reference scale to determine whether the diagnostic indicator of the first subsection exceeds a threshold of the reference scale of the cell cycle variation; and determining, by the processor, the presence of one or more cell cycle anomalies associated with the first subsection based on the diagnostic indicator exceeding the threshold.

2. The computer-implemented method of claim 1 , further comprising graphically indicating the first subsection based on the diagnostic indicator exceeding the threshold in an area of ​​the digital image.

3. 3. The computer-implemented method of claim 2, wherein the graphically indicating includes generating a two-dimensional shape to be rendered on the digital image, the two-dimensional shape including a boundary surrounding the first subsection based on the diagnostic indicator exceeding the threshold.

4. The computer-implemented method of claim 3 , further comprising generating one or more pathology annotations associated with the boundary.

5. 4. The computer-implemented method of claim 3, wherein the boundary further encloses one or more additional subsections of the plurality of subsections based on their respective diagnostic indicators exceeding the threshold.

6. The computer-implemented method of claim 5 , wherein the boundary excludes one or more subsections for which a corresponding diagnostic index does not exceed the threshold.

7. 7. The computer-implemented method of claim 6, further comprising generating a heat map within the two-dimensional shape, the heat map indicating the degree(s) to which the diagnostic index of the corresponding first subsection and the respective diagnostic index of the corresponding one or more additional subsections exceed the threshold.

8. The computer-implemented method of claim 6 , wherein the two-dimensional shape comprises a convex hull polygon.

9. 2. The computer-implemented method of claim 1, wherein said executing comprises mapping said subset of said plurality of values ​​of said diagnostic score to corresponding variables in said cell cycle transformation function.

10. 2. The computer-implemented method of claim 1, wherein the identified degree of presence of a cell cycle variant in the first subsection is proportional to the degree to which the diagnostic indicator exceeds the threshold.

11. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates dysregulation of cell cycle S phase to identify regions of high-grade cancer.

12. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates an active cell cycle and identifies regions of high Ki67 expression.

13. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates nuclear size skewness and chromosomal instability to identify regions of high DNA ploidy.

14. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates architectural distortion and cell cycle arrest in the stroma to identify regions of stromal rich TILs.

15. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates a cell cycle G1 / S dysregulation and arrest signature to identify HER2 positive / amplified regions.

16. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates a cell cycle G1 / S dysregulation and arrest signature to identify HR positive / overexpression regions.

17. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates cell cycle G1 entry and G0 arrest signatures to identify regions harboring high quiescent population load (QPL).

18. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates chromosomal instability and identifies BRCA positive and HRD carrying regions.

19. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates cell cycle arrest and chromosomal instability to identify regions with MMR dysfunction.

20. 4. The computer-implemented method of claim 3, wherein the graphically indicating comprises rendering an overlay on the digital image, the overlay graphically identifying one or more regions of the digital image corresponding to cellular abnormalities.

21. 2. The computer-implemented method of claim 1, further comprising automatically identifying, by the processor, one or more edges of the biological specimen depicted in the digital image, and restricting the two-dimensional shape to an area within the one or more edges of the biological specimen.

22. The computer-implemented method of claim 1 , wherein the image of the biological sample is a stained tissue image.

23. 2. The computer-implemented method of claim 1, further comprising: executing, by the processor, one or more cell cycle transformation functions to generate a diagnostic index corresponding to each of the plurality of divided subsections of the digital image, wherein the one or more cell cycle transformation functions determine one or more diagnostic entities selected from the group consisting of histological grade of cancer, mitotic figures, Ki67 status, DNA ploidy, stromal TIL status, HER2 status, HR status, quiescent population burden (QPL), BRCA mutation or HRD status, MMR function deficiency status, and any combination of the foregoing.

24. 2. The computer-implemented method of claim 1, further comprising: executing, by the processor, one or more cell cycle transformation functions to generate a corresponding diagnostic index for each of the plurality of divided subsections of the digital image, wherein the one or more cell cycle transformation functions calculate one or more surrogate biological signatures selected from the group consisting of cell cycle DNA synthesis phase (S phase) dysregulation, length of the active phase of the cell cycle, M-phase activity, nuclear size skewness and chromosomal instability, global measures of interstitial architectural distortion and cell cycle arrest, cell cycle G1 / S dysregulation and arrest signatures, cell cycle G1 entry and G0 arrest signatures, chromosomal instability, cell cycle arrest, and any combination of the above.

25. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; and automatically assessing, by the processor, a measure of cell cycle S-phase dysregulation for each of at least one subsection of the digital image to identify regions of high-grade cancer.

26. 26. The method of claim 25, wherein the automatically evaluating comprises calculating a curvature variance of a geometric (<C>) junction for each of the at least one subsection.

27. 26. The method of claim 25, comprising using said measure of cell cycle S-phase dysregulation to identify one or more regions of said digital image that contain mitotic figures.

28. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; and automatically assessing, by said processor, a measure of cell cycle activity in each of at least one subsection of said digital image to identify areas of high Ki67 expression.

29. 29. The method of claim 28, wherein the automatically evaluating comprises calculating (C_L-initial skewness) x (C_L-maximum skewness) for each of the at least one subsection.

30. 30. The method of claim 28, comprising using said measure of cell cycle activity to identify one or more regions of said digital image that contain mitotic figures.

31. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; and automatically assessing, by the processor, measures of nuclear size skewness and chromosomal instability for each of at least one subsection of the digital image to identify regions of high DNA ploidy.

32. 32. The method of claim 31, wherein said automatically evaluating comprises calculating a percentage of tissue area with high maximum skewness for each of said at least one subsection (C_L).

33. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; and automatically assessing, by the processor, measures of stromal architectural distortion and cell cycle arrest in each of at least one subsection of the digital image to identify regions of stromal rich TIL.

34. 34. The method of claim 33, wherein the automatically evaluating comprises calculating a shape (<C>) deviation for each of the at least one subsection.

35. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; The method comprises automatically assessing, by the processor, measurements of cell cycle G1 / S dysregulation and arrest signatures for each of at least one subsection of the digital image to identify HER2 positive / amplified regions.

36. 36. The method of claim 35, wherein the automatically evaluating comprises calculating a shape (<C_L>) divergence for each of the at least one subsection.

37. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; The method comprises automatically assessing, by the processor, measurements of cell cycle G1 / S dysregulation and arrest signatures for each of at least one subsection of the digital image to identify HR positive / overexpression regions.

38. 38. The method of claim 37, wherein the automatically evaluating comprises calculating a shape (<C>) divergence for each of the at least one subsection.

39. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; The method comprises automatically assessing, by the processor, measurements of cell cycle G1 entry and G0 arrest signatures for each of at least one subsection of the digital image to identify regions possessing high quiescent population load (QPL).

40. 40. The method of claim 39, wherein said automatically evaluating comprises calculating (<C>max-<C>min) x (<C>max skewness) for each of said at least one subsection.

41. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; and automatically assessing, by said processor, a measure of chromosomal instability for each of at least one subsection of said digital image to identify BRCA positive and HRD carrying regions.

42. 42. The method of claim 41, wherein the automatically evaluating comprises calculating shape (<C>) boundary value jumps for each of the at least one subsection.

43. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; and automatically assessing, by the processor, measures of cell cycle arrest and chromosomal instability for each of at least one subsection of the digital image to identify regions with MMR dysfunction.

44. 44. The method of claim 43, wherein the automatically evaluating comprises calculating a shape (<C>) cross-sectional gradient variation for each of the at least one subsection.

45. 1. A computer-implemented method comprising: determining, by a processor of a computing device, the presence of one or more cell cycle variants associated with a first subsection of a digital image of a biological sample, the digital image including an area divided into a plurality of subsections; the determining is based on a diagnostic index exceeding a threshold; the diagnostic score for the first subsection comprises a plurality of values ​​collectively related to nuclear heat and nuclear area of ​​one or more cells imaged in the first subsection of the biological specimen; The computer-implemented method, wherein the diagnostic indicator for the first subsection is determined using a cell cycle variant function that identifies one or more cell cycle variants based on a subset of a plurality of values ​​of the diagnostic score for the first subsection.

46. 46. ​​The method of claim 45, wherein the one or more cell cycle transformation functions calculate a surrogate biological signature selected from the group consisting of cell cycle DNA synthesis phase (S phase) dysregulation, length of the active phase of the cell cycle, M-phase activity, nuclear size skewness and chromosomal instability, global measures of interstitial architectural distortion and cell cycle arrest, cell cycle G1 / S dysregulation and arrest signatures, cell cycle G1 entry and G0 arrest signatures, chromosomal instability, cell cycle arrest, and any combination of the above.

47. 46. ​​The method of claim 45, wherein the cell cycle transformation function determines a diagnostic entity selected from the group consisting of cancer histological grade, mitotic figures, Ki67 status, DNA ploidy, stromal TIL status, HER2 status, HR status, quiescent population burden (QPL), BRCA mutation or HRD status, MMR function deficiency status, and any combination of the above.

48. 46. ​​The method of claim 45, further comprising automatically identifying and graphically rendering, by the processor, a two-dimensional shape including a boundary surrounding the first subsection and one or more of the additional subsections of the digital image based on the diagnostic indicators for each of the first subsections and each of the additional subsections exceeding the threshold.

49. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates cell cycle G1 / S dysregulation and cell cycle G1 phase entry dysregulation to identify regions containing one or more of BRAF and NRAS gene mutations.

50. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; The method comprises automatically assessing, by the processor, measures of cell cycle G1 / S dysregulation and cell cycle G1 phase entry dysregulation for each of at least one subsection of the digital image to identify regions containing one or more of BRAF and NRAS mutations.

51. 51. The method of claim 50, wherein the automatically evaluating comprises calculating a shape(<C>max, <C>min, <C>) divergence for each of the at least one subsection.

52. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates cell cycle G1 / S dysregulation to identify areas of dysplasia.

53. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; and automatically assessing, by said processor, a measure of cell cycle G1 / S dysregulation in each of at least one subsection of said digital image to identify areas of dysplasia.

54. 54. The method of claim 53, wherein the automatically evaluating comprises calculating a shape(<C_L>, <C>) divergence for each of the at least one subsection.

55. 2. The computer-implemented method of claim 1, wherein the cell cycle transformation function calculates cell cycle G2 / M dysregulation to identify the degree of dysplasia in a region.

56. 1. A method comprising: receiving, by a processor of a computing device, a digital image of a biological sample; The method includes automatically assessing, by the processor, a measure of cell cycle G2 / M dysregulation in each of at least one subsection of the digital image to identify a degree of dysplasia in an area of ​​the digital image.

57. 57. The method of claim 56, wherein said automatically evaluating comprises calculating a shape (<C_L>, <C>max) divergence and a (<C>max-<C>min) shift for each of said at least one subsection.

58. 1. A system comprising: one or more computers; and and one or more storage devices communicatively connected to the one or more computers, the one or more storage devices storing instructions that, when executed by the one or more computers, cause the one or more computers to perform operations of the respective method described in any one of claims 1 to 57.

59. One or more non-transitory computer storage media storing instructions that, when executed by one or more computers, cause the one or more computers to perform the operations of each of the methods recited in any one of claims 1 to 57.

Images