A computational system for multiparameter cellular and subcellular imaging data in situ or in vitro: Pathology Spatial Analysis Platform

Abstract

To provide a comprehensive computational systems pathology spatial analysis computer platform capable of integrating, visualizing and modeling high-dimensional in situ or in vitro cellular and subcellular resolved imaging data.SOLUTION: In a computational systems pathology spatial analysis platform, a processing apparatus 25 includes a spatial heterogeneity quantification component 30, a microdomain identification component 35, and a weighted graph component 40 for constructing a weighted graph for multi-parameter cellular and subcellular imaging data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to digital pathology, and in particular to a Comprehensive Computational System for Pathology Spatial Analysis (CSPSA) computer platform capable of integrating, visualizing, and modeling high-dimensional cellular and subcellular resolution image data in situ or in vitro. [Background technology]

[0002] Digital pathology refers to the acquisition, storage, and display of histologically stained tissue samples, initially gaining traction in niche applications such as second-opinion telepathology, immunostain interpretation, and intraoperative telepathology. Large volumes of patient data, typically consisting of 3–50 slides, are generated from biopsy specimens and visually evaluated by pathologists under a microscope; however, digital technology allows for this evaluation on a high-resolution monitor. Due to the manual intervention involved, current workflow practices are time-consuming, error-prone, and subjective.

[0003] Cancer is a heterogeneous disease. On hematoxylin-eosin (H&E)-stained histological images, heterogeneity is characterized by the presence of various histological structures, such as carcinoma in situ, invasive carcinoma, adipose tissue, blood vessels, and normal ducts. Furthermore, in many malignancies, molecular and cellular heterogeneity is a prominent feature between tumors in different patients, between different sites of neoplasia in the same patient, and within a single tumor. Intratumor heterogeneity involves phenotypically distinct clonal subpopulations of cancer cells and other cell types that comprise the tumor microenvironment (TME). These clonal subpopulations and other cell types include local bone marrow-derived stromal stem and progenitor cells, subclasses of immune-inflammatory cells that are either tumor-promoting or tumor-lethal, cancer-associated fibroblasts, endothelial cells, and pericytes. The TME can be viewed as an evolving ecosystem in which cancer cells engage in heterotypic interactions with these other cell types and utilize available resources to proliferate and survive. From this perspective, the spatial interactions between cell types within the TME (i.e., spatial heterogeneity) are thought to be one of the major factors behind disease progression and treatment resistance. Therefore, it is imperative to characterize spatial heterogeneity within the TME to accurately define specific disease subtypes and identify optimal treatment courses for individual patients.

[0004] To date, intratumor heterogeneity has been investigated using three major approaches. The first approach involves taking core samples from specific regions of the tumor and measuring population averages. Sample heterogeneity is measured by analyzing multiple cores within the tumor using multiple techniques, such as whole-exome sequencing, epigenetics, proteomics, and metabolomics. The second approach involves "single-cell analysis" using the methods described above, RNASeq, imaging, or flow cytometry after isolation of cells from tissue. The third approach utilizes the spatial resolution of optical microscopy imaging to maintain spatial context and measure cellular biomarkers in situ in combination with molecular-specific labeling. Biomarkers can identify cell type, activation state (e.g., phosphorylation of target proteins), and intracellular function. While each of these approaches offers some level of effectiveness, they each have various drawbacks and limitations.

[0005] Furthermore, one of the biggest problems in assessing the clinical significance of tumor heterogeneity is the lack of advanced tools for the spatial analysis of multiparameter cellular and subcellular imaging data. Summary of the Invention

[0006] In one embodiment, a method for analyzing disease progression from multiparameter cellular and subcellular image data obtained from multiple tissue samples from multiple patients or from several multicellular in vitro models is provided. The method includes generating a global quantification of spatial heterogeneity among cells of several different predetermined phenotypes in the multiparameter cellular and subcellular image data, identifying a plurality of microdomains for the multiple tissue samples based on the global quantification, each microdomain associated with a respective one of the multiple tissue samples, and constructing a weighted graph for the multiparameter cellular and subcellular image data. The weighted graph has a plurality of nodes and a plurality of edges, each located between a pair of the nodes, where each node in the weighted graph is a specific one of the microdomains, and the edge between each pair of microdomains in the weighted graph indicates the degree of similarity between that pair of microdomains.

[0007] In another embodiment, a system for analyzing disease progression from multi-parameter cellular and subcellular image data obtained from multiple tissue samples from multiple patients or from several multi-cellular in vitro models is provided, comprising: a processing device including: (i) a spatial heterogeneity quantification component configured to generate a global quantification of spatial heterogeneity among cells of several different predetermined phenotypes in the multi-parameter cellular and subcellular image data; (ii) a microdomain identification component configured to identify a plurality of microdomains for the plurality of tissue samples based on the global quantification, each microdomain associated with a respective one of the plurality of tissue samples; and (iii) a weighted graph component that constructs a weighted graph for the multi-parameter cellular and subcellular image data, the weighted graph having a plurality of nodes and a plurality of edges, each located between a pair of nodes, where each node in the weighted graph is a particular one of the microdomains and the edge between each pair of microdomains in the weighted graph indicates a degree of similarity between that pair of microdomains.

[0008] In yet another embodiment, a method for generating a spatially-based representation of intercellular communication from multiparameter cellular and subcellular image data obtained from multiple tissue samples from multiple patients or multiple multicellular in vitro models is provided. The method includes generating a quantification of spatial heterogeneity between cells of several different predetermined phenotypes within the multiparameter cellular and subcellular image data and identifying microdomains for one of the several tissue samples based on the quantification, wherein the multiparameter cellular and subcellular image data is phenotyped to identify the several different predetermined phenotypes, and constructing a communication graph for the microdomains. Each phenotype is a node in the communication graph, and an edge between each pair of phenotypes in the communication graph indicates the influence of one phenotype in the pair on the presence of the other phenotype in the pair.

[0009] In yet another embodiment, a computer system is provided for generating a spatially-based representation of heterogeneous cell-to-cell communication from multi-parameter cellular and subcellular image data obtained from multiple tissue samples from multiple patients or multiple multicellular in vitro models. The system includes a processing device including: (i) a spatial heterogeneity quantification component configured to generate a quantification of spatial heterogeneity between cells of several different predetermined phenotypes in the multi-parameter cellular and subcellular image data, the spatial heterogeneity quantification component performing cell phenotyping on the multi-parameter cellular and subcellular image data to identify the specific different predetermined phenotypes; (ii) a microdomain identification component configured to identify a plurality of microdomains for the several tissue samples based on the quantification, each microdomain associated with one of the several tissue samples; and (iii) a communication network component configured to construct a communication graph for a selected one of the plurality of microdomains, each phenotype being a node in the communication graph, and an edge between each pair of phenotypes in the communication graph indicating the influence of one phenotype in the pair on the presence of the other phenotype in the pair.

[0010] In yet another aspect, there is provided a method for creating a personalized medical strategy for a specific patient, the specific patient being one of several patients, the several patients being associated with multi-parameter cellular and subcellular image data obtained from several tissue samples from the several patients. The method includes the steps of: identifying a plurality of microdomains in one of the several tissue samples associated with the specific patient, the plurality of microdomains being based on the multi-parameter cellular and subcellular image data; generating a heterogeneous intercellular communication network for each of the plurality of microdomains, each heterogeneous intercellular communication network including a representation of the heterogeneous intercellular communication based on spatial information of the microdomain; quantifying spatial and temporal interdependencies of the microdomains based on the heterogeneous intercellular communication network; and formulating a medical strategy for the specific patient based on the quantified results.

[0011] In yet another embodiment, a method for representing the time evolution of a disease in a particular patient is provided, where the particular patient is one of several patients, and the several patients are associated with multi-parameter cellular and subcellular image data obtained from several tissue samples from the several patients. The method includes generating a geometric representation of a disease landscape for the particular patient, the geometric representation including a plurality of points, each point of the geometric representation (i) describing the disease status of the particular patient at a particular time point and based on a particular one of the several tissue samples associated with the selected patient, (ii) based on a microdomain in the particular one of the several tissue samples based on the multi-parameter cellular and subcellular image data, and (iii) including a representation of a heterogeneous intercellular communication network for the microdomain based on spatial information for the microdomain. [Brief explanation of the drawings]

[0012] [Figure 1]FIG. 1 is a schematic diagram of an exemplary Computational System Pathology Spatial Analysis (CSPSA) platform for in situ or in vitro multiparameter cellular and subcellular image data, according to one embodiment of the disclosed concepts. [Figure 2] FIG. 2 is a schematic diagram of an exemplary hyperplexed image stack. [Figure 3] FIG. 3 is a flowchart illustrating a method for analyzing tumor progression / evolution from multi-parameter cellular and subcellular image data obtained from multiple tumor sections from a patient cohort, which method can be implemented on the CSPSA platform of FIG. 1 in accordance with an exemplary embodiment of the disclosed concepts. [Figure 4] FIG. 4 is a schematic diagram of an exemplary global space map. [Figure 5] FIG. 5 is a schematic diagram of each of the predetermined dominant biomarker intensity patterns in an exemplary embodiment of the disclosed concepts. [Figure 6] FIG. 6 is a schematic illustration of a cellular spatially dependent image according to one particular exemplary embodiment of the disclosed concepts. [Figure 7] FIG. 7 is a schematic diagram of an exemplary microdomain of an exemplary tissue section according to one exemplary embodiment of the disclosed concepts. [Figure 8] FIG. 8 is a schematic diagram of a weighted microdomain graph constructed for multi-parameter cellular and subcellular image data in accordance with an exemplary embodiment of the disclosed concepts. [Figure 9] FIG. 9 is a flow chart illustrating a method for generating spatially informed representations of cross-species intercellular communication from multi-parameter cellular and subcellular image data according to another exemplary embodiment of the disclosed concepts. [Figure 10] FIG. 10 is a schematic diagram of an exemplary communication graph according to one exemplary embodiment of the disclosed concepts. [Figure 11]FIG. 11 is a schematic diagram of a particular communication graph generated according to a particular exemplary embodiment where the phenotypes are tumor cells, lymphocytes, macrophages, stroma, and necrosis. [Figure 12] FIG. 12 is a schematic diagram of a personalized medicine strategy according to one embodiment of the disclosed concepts. [Figure 13] FIG. 13 is a schematic diagram of an exemplary cancer landscape according to one embodiment of the disclosed concepts. DETAILED DESCRIPTION OF THE INVENTION

[0013] As used herein, the singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise.

[0014] As used herein, a description of two or more parts or components being "coupled" means that the parts are coupled or operate together, to the extent that a connection occurs, either directly or indirectly, i.e., through one or more intermediate parts or components.

[0015] As used herein, the term "several" means one or an integer greater than one (ie, a plurality).

[0016] As used herein, the terms "component" and "system" are intended to refer to a computer-related entity, whether hardware, a combination of hardware and software, software, or software in execution. For example, a component may be, but is not limited to, a process running on a processor, a processor, an object, an executable file, a thread of execution, a program, and / or a computer. For example, both an application running on a server and the server may be considered a component. One or more components may reside within a process and / or thread of execution, and components may be localized on one computer and / or distributed among two or more computers. While some methods of displaying information to a user have been illustrated and described with specific diagrams or graphs as screenshots, those skilled in the relevant art will recognize that various alternatives may be employed.

[0017] As used herein, the term "multiparameter cellular and subcellular image data" refers to data obtained by generating several images from several sections of tissue, providing information about multiple measurable parameters at the cellular or subcellular level in those sections of tissue. Multiparameter cellular and subcellular image data can be generated by several different imaging techniques, including, but not limited to, (1) transmitted light imaging by IHC using several biomarkers, (2) immunofluorescence imaging, including both multiplex imaging (1-7 biomarkers) and hyperplex imaging (more than 7 biomarkers), (3) toponome imaging, (4) matrix-assisted laser desorption / ionization mass spectrometry imaging (MALDI MSI), (5) complementary spatial imaging, such as FISH, MxFISH, FISHSEQ, or CyTOF, (6) multiparameter ion beam imaging, and (7) in vitro imaging of experimental models. Additionally, multi-parameter cellular and subcellular image data may be generated, for example, but not limited to, by generating several biomarker images from several sections of tissue by labeling each section of tissue with multiple different biomarkers.

[0018] As used herein, the term "spatial map" refers to a representation of several quantified spatial statistics that indicate the relationships between cells of different given phenotypes in a set of multi-parameter cellular and subcellular imaging data, e.g., a collection of data and / or a representation of spatial statistics in a visually perceptible format. For example, without limitation, a spatial map may be a point-wise mutual information (PMI) map generated by the methods described in PCT Application No. PCT / US2016 / 036825 and U.S. Patent Application Publication No. 2018 / 0204085, both of which are entitled "Systems and Methods for Finding Regions of Interest in Hematoxylin and Eosin (H&E) Stained Tissue Images and Quantifying Intratumor Cellular Spatial Heterogeneity in Multiplexed / Hyperplexed Fluorescence Tissue Images," the disclosures of which are incorporated herein by reference.

[0019] As used herein, the term "microdomain" refers to a spatial arrangement (or motif) of phenotypically distinct cells in a tissue sample that arises from spatial intratumor heterogeneity and is associated with one or more outcome-specific variables (e.g., time to recurrence). Microdomains may be identified from multiparameter cellular and subcellular imaging data according to any of several known or later-developed methods. Such methods include those described in U.S. Provisional Application No. 62 / 675,832, entitled "Predicting the Recurrence Risk of Cancer Patients From Primary Tumors with Multiplexed Immunofluorescence Biomarkers and Their Spatial Correlation Statistics," and PCT Application No. PCT / US2019 / 033662, entitled "System and Method for Predicting the Risk of Cancer Recurrence From Spatial Multi-Parameter Cellular and Sub-Cellular Imaging Data for Tumors by Identifying Emergent Spatial Domain Networks Associated With Recurrence," which are incorporated herein by reference, and / or "Spagnolo, et al., Platform for Quantitative Evaluation of Spatial Intratumoral Heterogeneity in Multiplexed Fluorescence Images, Cancer Res. 2017 Nov 1;77(21):e71-e74," and / or the public domain THRIVE (Tumor Examples of methods include, but are not limited to, methods implemented in the Heterogeneity Research Interactive Visualization Environment (Heterogeneity Research Interactive Visualization Environment) software.The method uses spatially resolved correlations between biomarkers as covariates in multivariate survival models of prognostic data (e.g., recurrence) to construct maps of the spatial organization of cancer recurrence in multiplexed tissue samples. These maps delineate microdomains associated with recurrence and metastatic progression.

[0020] Directional terms used herein, such as top, bottom, left, right, upper, lower, front, rear, and derivatives thereof, refer to the orientation of the elements illustrated in the drawings and do not limit the scope of the claims, unless expressly stated.

[0021] The disclosed concepts will be described below, for purposes of explanation, with reference to numerous specific details in order to provide a thorough understanding of the present invention. However, it will be apparent that the disclosed concepts can be practiced without these specific details without departing from the spirit and scope of the present invention.

[0022] The disclosed concept provides a comprehensive computational system for pathology spatial analysis (CSPSA) platform capable of integrating, visualizing, and / or modeling high-dimensional in situ cellular and subcellular resolved imaging data. The CSPSA platform of the disclosed concept combines a core set of existing tools for cellular phenotyping and spatial analysis with an advanced toolset for inferring spatial heterogeneous intercellular (intercellular) and intracellular communication patterns and, in the case of cancer, constructing tumor (or other disease) evolutionary trees from precancerous origins to metastatic endpoints. This combination of tools can be applied to both research and clinical purposes, such as quantifying spatial intratumor heterogeneity in tissue samples and in vitro models and correlating it with prognostic data, building diagnostic and prognostic applications, and designing personalized therapeutic strategies and drug discovery. Furthermore, the platform can be enhanced by adding region-specific genomics and spatial transcriptomics data to the in situ cellular and subcellular high-resolution images.

[0023] FIG. 1 is a schematic diagram of an exemplary Computational System for Pathology Spatial Analysis (CSPSA) platform 5 for in situ multiparameter cellular and subcellular image data based on an embodiment of the disclosed concepts, on which various methods described herein may be implemented. As shown in FIG. 1 , the CSPSA platform 5 is a computing device configured to receive and store specific multiparameter cellular and subcellular image data 10 (e.g., for a large patient cohort including tumors of recurrent and non-recurrent cancers, or for individual patients or individual tumors) and process the data 10 as described herein. In a non-limiting exemplary embodiment, the multiparameter cellular and subcellular image data 10 is generated using multiplexed or hyperplexed immunofluorescence imaging, although it will be understood that other imaging techniques, such as those described elsewhere herein, may also be used. For example, the multiparameter cellular and subcellular image data 10 may be based on an exemplary hyperplexed image stack 12 for all patient data in a cohort, as depicted in FIG. 2 .

[0024] The CSPSA platform 5 may be, for example, but not limited to, a PC, laptop computer, tablet computer, smartphone, or other suitable computing device configured to perform the functions described herein. The CSPSA platform 5 includes an input device 15 (e.g., a keyboard), a display 20 (e.g., an LCD), and a processing unit 25. A user can use the input device 15 to provide input to the processing unit 25, and the processing unit 25 provides output signals to the display 20 so that the display 20 can display information to the user (e.g., a spatial map or other spatial dependency image, a micro-domain image, a weighted graph image, and / or a communication network graph image), as described in detail herein. The processing unit 25 includes a processor and memory. The processor may be, for example, but not limited to, a microprocessor (μP), a microcontroller, an application-specific integrated circuit (ASIC), or other suitable processing device, interfaced with the memory. The memory may be one or more of various types of internal and / or external storage media, such as RAM, ROM, EPROM, EEPROM, FLASH®, or other computer-readable media providing storage registers, for data storage, such as an internal storage area of ​​the computer, and may be volatile or non-volatile memory. The memory stores several routines executable by the processor, including routines for implementing the disclosed concepts, as described herein.In particular, the processing device 25 includes a spatial heterogeneity quantification component 30 configured to generate a global quantification of cell-to-cell spatial heterogeneity of several different predetermined phenotypes in the multi-parameter cellular and subcellular image data 10 as described herein; a microdomain identification component 35 configured to identify a plurality of microdomains as described herein for a plurality of tumor sections from the multi-parameter cellular and subcellular image data 10 based on the global quantification generated by the spatial heterogeneity quantification component 30; a weighted graph component 40 configured to construct a weighted graph of the multi-parameter cellular and subcellular image data 10 as described herein; and a communication network component 45 configured to construct a communication network graph for the multi-parameter cellular and subcellular image data 10 as described herein.

[0025] 3 is a flowchart illustrating a method for analyzing tumor progression / evolution from multi-parameter cellular and subcellular image data 10 obtained from multiple tumor sections of a patient cohort, which method may be implemented in a CSPSA platform 5 in accordance with an exemplary embodiment of the disclosed concepts. However, it will be understood that this is intended to be exemplary only, and that the method steps illustrated in FIG. 3 may be implemented in other configurations and / or platforms.

[0026] The method begins at step 50, where spatial heterogeneity quantification component 30 provides a global quantification of spatial heterogeneity in multi-parameter cellular and subcellular image data 10. In an exemplary embodiment, the global quantification provided at step 50 is a global spatial map of multi-parameter cellular and subcellular image data 10. In one particular implementation, the global spatial map is a global PMI map 52, as shown in FIG. 4, generated in the manner described in PCT Application No. PCT / US2016 / 036825 and U.S. Patent Application Publication No. 2018 / 0204085, which are referred to elsewhere herein and incorporated by reference. In this exemplary implementation, multi-parameter cellular and subcellular image data 10 is generated by labeling each section of tissue with multiple different biomarkers (e.g., ER, PR, and HER2) to generate several biomarker images from several sections of tissue. As described in the aforementioned application, the PMI map 52 is generated by first performing cell segmentation on the multi-parameter cellular and subcellular image data 10 (i.e., for each "slide") thereof. Any of several suitable cell segmentation algorithms, now known or later developed, may be employed. Spatial location and biomarker intensity data are then obtained for each cell, and each cell is assigned to one of predetermined phenotypes (each phenotype being a predetermined dominant biomarker intensity pattern) based on the cell's biomarker intensity configuration. FIG. 5 shows, in an exemplary embodiment, a schematic representation 56 of each of the predetermined dominant biomarker intensity patterns, numbered 1 through 8. In an exemplary embodiment, each schematic representation 56 is provided with a unique color so that the schematic representations can be easily distinguished from one another. The cell assignments described herein and the schematic representation shown in FIG. 5 can be used to generate a cellular spatial dependency image that visually indicates the heterogeneity of a target tissue sample. FIG. 6 shows a cellular spatial dependency image 58 based on one particular exemplary embodiment of the disclosed concepts. As shown in FIG. 6, a cellular spatial dependency image 58 uses a schematic representation 56 to show the spatial dependency between cells in a subject slide.A spatial network is then constructed to represent the composition of the intensity patterns of key biomarkers in the subject's slide, and the heterogeneity of the subject's slide is quantified by generating a PMI map 52, as shown in Figure 4. In an exemplary embodiment, the spatial network and PMI map 52 are generated as follows:

[0027] A network is constructed for a subject slide to represent the spatial organization of biomarker patterns in the biomarker images (i.e., tissue / tumor samples) of the subject slide. Constructing a spatial network for a tumor sample essentially connects cellular biomarker intensity data (network nodes) with spatial data (network edges). The premise in constructing the network is that cells have the ability to communicate with nearby cells up to a certain limit, e.g., up to 250 μm, and that the ability of a cell to communicate within that limit depends on the distance between the cells. Therefore, in an exemplary embodiment, a probability distribution is calculated for the distance between a cell in the subject slide and its 10 nearest neighbors. A hard limit is selected based on the median value of this distribution multiplied by 1.5 (to estimate the standard deviation), and cells in the network are only connected within this limit. Edges between cells in the network are then weighted by the distance between adjacent cells.

[0028] In an exemplary embodiment, point-wise mutual information (PMI) is then used to assess the association between each pair of biomarker patterns in the dictionary, and therefore between different cellular phenotypes, for the subject's slides. This metric captures general statistical associations, both linear and nonlinear, and previous work has used linear metrics such as Spearman's rho coefficient. Once PMI is calculated for each pair of biomarker patterns, the overall measure of association in the subject's slide data is displayed in a PMI map 52. The use of PMI in this embodiment is merely exemplary, and other methods of defining spatial relationships, such as those described in Spagnolo DM, Al-Kofahi Y, Zhu P, Lezon TR, Gough A, Stern AM, Lee AV, Ginty F, Sarachan B, Taylor DL, Chennubhotla SC, Platform for Quantitative Evaluation of Spatial Intratumoral Heterogeneity in Multiplexed Fluorescence Images, Cancer Res. 2017 Nov 1 and Nguyen, L., Tosun, B., Fine, J., Lee, A., Taylor, L., Chennubhotla, C. (2017), Spatial statistics for segmenting histological structures in H&E stained tissue images, IEEE Trans Med Imaging. 2017 Mar 16, may be employed in connection with the disclosed concepts.

[0029] The exemplary PMI map 52 describes the relationship between different cellular phenotypes within the microenvironment of a subject's slide. In particular, entries in the PMI map 52 indicate how frequently a particular spatial interaction between two phenotypes (referenced by row and column numbers) occurs in a dataset compared to the interaction predicted by a random (or background) distribution across all phenotypes. Entries in a first color, such as red, indicate a strong spatial association between the phenotypes, while entries in a second color, such as black, indicate a lack of colocalization (a weak spatial association between the phenotypes). Other colors may be used to indicate other associations. For example, PMI entries colored in a third color, such as green, indicate an association that is no better than the random distribution of cellular phenotypes across the dataset. Additionally, the PMI map 52 can indicate anti-associations with entries colored in a fourth color, such as blue (e.g., if phenotype 1 rarely occurs spatially near phenotype 3).

[0030] Referring again to FIG. 3 , following step 50, the method then proceeds to step 55. In step 55, the microdomain identification component 35 identifies and locates one or more microdomains (e.g., as described elsewhere herein) in each tumor section in the multi-parameter cellular and subcellular image data 10 based on the global quantification provided in step 50. For example, a cancer cell type and a specific immune cell type coexisting with it may have a high spatial co-occurrence value. Using combinations of cancer and immune cells as seed points, microdomains can be generated in any tissue section by growing a spatial network of cells around the seed (e.g., approximately 100 cells) based on a distance cutoff, by mining the global quantification (e.g., spatial map) and propagating phenotypic associations. Such an exemplary microdomain 62 is illustrated in the exemplary tissue section 64 shown in FIG. 7 .

[0031] Next, in step 60, a local quantification of spatial heterogeneity (e.g., a local spatial map, such as a local PMI map) is determined for each microdomain identified in step 55. Furthermore, each of the local quantifications determined for each tissue section is used to define a degree of similarity between pairs of microdomains in the tissue section. In particular, in an exemplary embodiment, given two microdomains, A and B, (i) the difference in relative abundance of cells in microdomain A and microdomain B and (ii) their local quantifications (e.g., PMI maps) are used to define a degree of similarity between pairs of microdomains in the tissue section. A and PMI B A distance function is defined that is a combination of differences in the local spatial map (such as σ, ...

[0032] Next, in step 65, a weighted microdomain graph 66, shown in exemplary form in Figure 8, is constructed for the multi-parameter cellular and subcellular image data 10. Each node 68 of the weighted microdomain graph 66 is one of the microdomains generated in step 55, and the edges 72 connecting each pair of microdomains are weighted by their determined similarity (e.g., the weighted similarity values ​​described above). The weighted microdomain graph 66 may then be displayed on the display 20.

[0033] Next, as shown in step 70 of FIG. 3 , the weighted microdomain graph 66 can be further analyzed to understand phenotypic progression during the metastatic process. Note that while ground truth labels exist for each entire tissue section of the multi-parameter cellular and subcellular image data 10 as originating from either recurrent (R) or non-recurrent (NR) patients, no a priori knowledge is available to ascertain which microdomains best distinguish between the two categories. In an exemplary embodiment, the labels R and NR are transferred from the tissue section to the corresponding microdomains. Several iterations of belief propagation are then performed to refine the labels and isolate cliques of microdomains in the weighted microdomain graph 66 that are uniformly R or NR. With improved confidence in label assignment, a betweenness centrality measure can be used to identify microdomains that are most commonly transferred on the path from NR to R. To mimic evolutionary trajectories, a random walk of the weighted microdomain graph 66 can be performed, where it is known that some microdomains are only found in either observed or unobserved metastatic data. Comparing the relative probabilities of paths from NR to R can identify classes of events important for metastasis. The mean first-passage time of the random walk can then be used to identify metastasis-prone microdomains as sets of nodes for which the random walk has a 0.5 probability of reaching either R or NR. Note that a metastasis-prone microdomain is a midpoint on the evolutionary trajectory from NR to R. Note that in an exemplary embodiment, the steps just described are performed by components of the CSPSA platform 5, and the results are displayed to the user on the display 20 of the CSPSA platform 5. This aspect of the disclosed concepts thus provides the ability to define, identify, and compare phenotypes (phenotypic evolution) important for metastasis, as described herein.Additionally, it will be understood that cancer is only one example of an area in which the disclosed concepts may be applied, and that the disclosed concepts may also be applied to other diseased tissues, such as neurodegenerative diseases, metabolic diseases, inflammatory diseases, among others.

[0034] 9 is a flowchart illustrating a method for generating a spatially informed representation of heterogeneous intercellular communication from multi-parameter cellular and subcellular image data 10, which in this embodiment may be obtained from several tumor sections from a patient, and may be implemented in a CSPSA platform 5 according to another exemplary embodiment of the disclosed concepts. However, this is intended to be illustrative only, and it will be understood that the method steps illustrated in FIG. 9 may be implemented in other configurations and / or platforms.

[0035] The method begins at step 75, where the spatial heterogeneity quantification component 30 provides quantification of spatial heterogeneity in the multi-parameter cellular and subcellular image data 10, as described in detail elsewhere herein. As described elsewhere herein, the quantification of spatial heterogeneity performed at step 75 includes performing cellular phenotyping on the multi-parameter cellular and subcellular image data 10 to identify distinct predetermined phenotypes. In an exemplary embodiment, the quantification provided at step 75 is a spatial map of the multi-parameter cellular and subcellular image data 10. In certain embodiments, the spatial map is generated in the manner described in PCT Application No. PCT / US2016 / 036825 and U.S. Patent Application Publication No. 2018 / 0204085, which are referred to elsewhere herein and incorporated by reference. In this exemplary embodiment, the multi-parameter cellular and subcellular image data 10 may be generated by generating several biomarker images from several sections of tissue by labeling each section of tissue with multiple different biomarkers (e.g., ER, PR, and HER2). In a non-limiting exemplary embodiment, the multi-parameter cellular and subcellular image data 10 is generated using multiplexed or hyperplexed immunofluorescence imaging, although it will be understood that other imaging techniques, such as those described elsewhere herein, may also be used.

[0036] Next, at step 80, microdomain identification component 35 identifies and locates one or more microdomains as described herein in each of the tumor sections of multi-parameter cellular and subcellular image data 10 based on the quantification provided at step 75. Next, at step 85, one or more of the identified microdomains are selected. The method then proceeds to step 90.

[0037] In step 90, the communication network component 45 constructs a communication graph for each of the selected microdomains. The communication graph may be displayed on the display 20. An exemplary communication graph 92 is shown in FIG. 10. As shown in FIG. 10, in an exemplary embodiment, each phenotype is a node 94 of the communication graph 92, and an edge 96 between each pair of phenotypes (each pair of nodes 94) in the communication graph 92 indicates the influence of one phenotype of the pair on the presence of the other phenotype of the pair. In an exemplary embodiment, each phenotype (each node 94) in the communication graph 92 is represented by a data vector obtained from the multi-parameter cellular and subcellular image data 10. Furthermore, the edges 96 of the communication graph 92 in the exemplary embodiment are generated by determining, for each pair of phenotypes, the value of a linear or non-linear correlation coefficient between the data vectors of the pair of phenotypes to establish a numerical relationship between the pair of phenotypes (after removing the confounding effects of phenotypes other than the pair of phenotypes). In addition, for each numerical relationship, a directionality is preferably established based on several receptor-ligand databases of related biomarkers. Thus, in this embodiment, the edges 96 between each pair of nodes 94 indicate the determined numerical relationship and the determined directionality. In an exemplary embodiment, the edges 96 may be colored to represent a particular correlation coefficient value (e.g., the degree of positive or negative correlation).

[0038] Additionally, in exemplary embodiments, each data vector of each node 94 is a vector of expression values ​​of several predetermined biomarkers, where the predetermined biomarkers are the specific biomarkers used in generating the multi-parameter cellular and subcellular image data 10. For example, each vector of expression values ​​may be a biomarker intensity pattern for the predetermined biomarkers determined in step 75. Each data vector may also be analyzed to biologically interpret and determine several activation states 98 for each of the phenotypes (each node 94), which activation states 98 are included in the communication graph 92. For example, the biomarker ALDH1 is a surrogate for tumor cell stemness, and the biomarker PD-L1 determines mutational burden.

[0039] 11 is a schematic diagram of a particular communication graph generated according to a particular exemplary embodiment in which the phenotypes (nodes 94) are tumor cell, lymphocyte, macrophage, stroma, and necrosis. The influences 96 of various edges and the various activation states 98 are also shown.

[0040] One outcome of the mapping described herein is to infer disease progression pathways based on the local tumor microenvironment (TME) and then list known molecular targets in those pathways. This can then be followed by using machine learning tools (e.g., Balestra Web) to predict drug-target interactions based on spatial relationships. This could lead to the development of new treatments as well as drug repurposing. Further detailed below are several exemplary clinical applications (e.g., applications of the CSPA platform) that can incorporate various aspects of the disclosed concepts, including: (1) drug discovery and personalized medicine strategies using spatially modulated computational systems (pathology); (2) cancer / disease landscapes, which are geometric representations of multiparametric readouts of a patient's cancer / disease state; and (3) applications of the CSPA platform for in vitro models for basic research and clinical translation.

[0041] In certain exemplary embodiments, the disclosed concepts can be used for drug discovery and personalized medicine strategies using spatially modulated computational systems pathology. More specifically, spatial intratumor heterogeneity is a key feature in determining the temporal evolution of cancer and the fate of patients. This heterogeneity is reflected in the diversity of heterogeneous intercellular communication networks embedded in microdomains, making the resulting systems biology patient-dependent. Current treatment strategies are designed for average patients and therefore do not reflect patient-specific systems biology. Patient-specific systems pathology begins with identifying microdomains / regions of interest, characterizing their underlying communication networks, and accurately quantifying microdomain interdependencies both spatially and temporally. This knowledge is also essential for new drug design strategies. This strategy can also be enhanced by utilizing region-specific genomics.

[0042] Figure 12 is a schematic illustration of a personalized medicine strategy based on this aspect of the disclosed concepts, ranging from tissue sections to personalized therapeutic treatments. From left to right, Figure 12 shows: i) multi-parameter imaging at cellular / subcellular resolution; ii) identification of regions of interest (microdomains) within the image; iii) reconstruction of intra-microdomain communication networks using specific connection weights and spatial relationships between cells / microdomains; iv) combining the information to extract accurate systems biology; and v) prescription of personalized therapeutic strategies.

[0043] As shown in FIG. 12 , in the illustrated exemplary embodiment, for an individual patient, microdomain 1 and microdomain 2 of the sample were found to be phenotypically distinct. A phenotyping algorithm, as described herein, was applied to the tissue sample. A heterogeneous intercellular network was then constructed for each microdomain, as described herein. Prognostic tests applied separately to microdomain 1 and microdomain 2 revealed that both microdomains were associated with a low risk of recurrence. However, when the microdomains were in close proximity, the risk of recurrence increased dramatically, suggesting a flow of information between microdomain 1 and microdomain 2. Thus, in this aspect of the disclosed concepts, the heterogeneous intercellular network informs which pathways are activated and what the relationship between the two networks is. For example, WNT signaling is upregulated in microdomain 2 but downregulated in microdomain 1, and TGFβ is upregulated in microdomain 1 but downregulated in microdomain 2.

[0044] In this aspect of the disclosed concepts, pathways and microdomain networks are identified, and thus the information flows that must be inhibited to slow the progression of cancer in a particular individual are further identified. This information can be used to develop personalized medicine strategies. In another specific exemplary embodiment, the disclosed concepts can be used to accurately describe the time evolution of a disease, such as cancer, in a particular patient. Thus, the disclosed concepts define a cancer landscape, which is a geometric representation of a multi-parametric readout of the patient's cancer state. Each point in the cancer landscape represents the patient's cancer state, such as pre-cancerous, early stage, or invasive. The path followed in this landscape describes the time evolution of the disease for that particular patient. Current treatments use cancer landscapes averaged over a patient population. Unless individual patient profiles perfectly match the average profile, predictions made with average models may be inaccurate.

[0045] Figure 13 shows an example cancer landscape with a geometric representation, where locations on the landscape represent specific example patient states. Using colon cancer progression as an example, Figure 13 shows the depression labeled 1 representing the precancerous state, the depression labeled 2 representing the small polyp state, the depression labeled 3 representing the large polyp state, and the depression labeled 4 representing the invasive colon cancer state. The arrows in the geometric representation indicate the path a patient takes from precancerous to invasive cancer. From this model, the dynamics of cancer progression can be inferred.

[0046] According to one aspect of this embodiment of the disclosed concepts, a retrospective patient cohort is sought to build a comprehensive library of microdomains described herein for tissue evolution from pre-cancer to metastasis. A specific heterogeneous intercellular communication network, as described herein, is associated with each microdomain. Additionally, each heterogeneous intercellular communication network has a systems biology model in the form of a system of ordinary differential equations. The system of ordinary differential equations defines the kinetics of the cancer landscape. In prospective studies, this kinetics helps predict the time evolution of the microdomain from pre-cancer to metastasis. The kinetic model also serves as a surrogate for a synthetic tissue developmental model.

[0047] In yet another specific exemplary embodiment, the disclosed concepts may be used in combination with in vitro models for basic research and clinical translation. More specifically, in two specific applications of the disclosed concepts discussed above, the platform of the disclosed concepts was applied to in situ, hyperplexed, single-cell resolution imaging of solid tumors. In this embodiment, the same platform is applied to image data from in vitro microphysiological models. Multicellular in vitro models enable the study of spatiotemporal cellular heterogeneity and heterogeneous intercellular communication that recapitulates human tissues, applicable to investigating mechanisms of disease progression in vitro, drug testing, and characterizing the structural organization and content of these models for use in transplantation.

[0048] In vitro models include 2D cultures, 3D spheroids, organoids, and biomimetic microphysiological systems. The goal of these systems has been to approach physiologically relevant biomimetics. 2D cultures are cell cultures grown on a surface. They develop 2D communication networks, so information exchange is limited to two dimensions. 3D spheroids are spatially growing clusters of cells. Although cells grow in an artificial environment, they better mimic in vivo growth conditions because cells can interact and grow in all directions. Organoids are very small yet self-organizing three-dimensional tissue cultures. They are typically derived from stem cells. Organoids can be engineered to reproduce much of the complexity of an organ or can be induced to express selected aspects of the organ, such as the generation of only specific cell types. Regenerating organ function, even partially, means reproducing systems biology and heterogeneous intercellular communication in controlled in vitro conditions. Biomimetic microphysiological systems are the first step towards mimicking organ function in the context of organoids. 3D microfluidic channel cell culture chips mimic the full range of activity, mechanics, and physiological responses of whole organs, allowing for a more accurate representation of systems biology and heterogeneous cell-to-cell communication networks.

[0049] In one or more aspects of this particular embodiment, spatiotemporal cellular heterogeneity and heterogeneous intercellular communication may be monitored by constructing a timeline of images and comparing the evolution of inferred networks in the model. The level of complexity of in vitro systems, from 2D cultures to biomimetic MPS, reflects the complexity of the systems biology models that can be developed. While applicable to all in vitro models, biomimetic models constructed by layering or bioprinting various cell types in appropriate 3D tissues will benefit most from defining spatial relationships within the model. The same type of systems pathology defined in the two specific applications of the disclosed concepts discussed above will be useful here. The goal is to demonstrate that the in vitro model reflects the organizational and spatial relationships identified in the in situ tissue / organ under study and that systems biology can be reproduced in vitro. Models may be established and studied in a "normal" healthy state, or may be established as disease models using cells from diseased patients and / or induced pluripotent stem cells (iPSCs) from the patient that have been differentiated and matured into the cell types required for the model. At different time points and after selected treatments, the models may be fixed, embedded, sectioned, and labeled in the same manner as patient tissue. Hyperplex imaging methods may then be applied to the labeled tissue sections. Computational system pathology analysis of the disclosed concepts described herein may then be applied to tissue sections obtained from the model.

[0050] Finally, although the foregoing has described image data obtained from tumor sections, it will be understood that the disclosed concepts are applicable to image data obtained from other types of tissue sections, as well as image data obtained from solid, un-sectioned samples using imaging modalities that can penetrate un-sectioned samples.

[0051] In the claims, signs placed between parentheses shall not be construed as limiting the scope of the claim. The word "comprises" or "includes" does not exclude the presence of elements or steps other than those listed in a claim. In a device claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The word "a" preceding an element does not exclude the presence of a plurality of such elements. In any device claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The mere fact that certain elements are recited in mutually different dependent claims does not indicate that these elements cannot be used in combination.

[0052] While the present invention has been described in detail for purposes of illustration based on what are presently considered to be the most practical and preferred embodiments, it should be understood that such detail is for the purpose only and that the invention is not limited to the disclosed embodiments, but on the contrary, is intended to cover modifications and equivalent arrangements within the spirit and scope of the appended claims. For example, it should be understood that the present invention contemplates that one or more features of any embodiment can be combined with one or more features of any other embodiment, to the extent possible.

Claims

1. 1. A method of generating a personalized medicine strategy for a particular patient, wherein the particular patient is one of several patients, the several patients being associated with multi-parameter cellular and subcellular image data obtained from several tissue samples from the several patients; identifying a plurality of microdomains in one of the several tissue samples associated with the particular patient, the plurality of microdomains being based on the multi-parameter cellular and subcellular image data; generating a heterogeneous intercellular communication network for each of the plurality of microdomains, wherein each heterogeneous intercellular communication network includes a representation of heterogeneous intercellular communication based on spatial information of the microdomain; Quantifying the spatial and temporal interdependence of the microdomains based on a plurality of heterogeneous intercellular communication networks; designing a personalized medicine for the particular patient based on said quantification; A method comprising:

2. 2. The method of claim 1, wherein each microdomain is identified by generating a quantification of spatial heterogeneity between cells of several different predetermined phenotypes in the multi-parameter cellular and subcellular image data and identifying a microdomain for one of the several tissue samples based on the quantification, the method comprising performing phenotyping on the multi-parameter cellular and subcellular image data to identify the several different predetermined phenotypes.

3. 2. The method of claim 1, wherein each heterogeneous intercellular communication network comprises a communication graph of the microdomain, each phenotype being a node of the communication graph, and an edge between each pair of phenotypes in the communication graph indicates the influence of one phenotype in the pair on the presence of the other phenotype in the pair.

4. 10. The method of claim 1, wherein each microdomain confers a low risk of a prognostic variable of interest, but the spatial proximity and relationship between microdomains increases the overall risk of the prognostic variable of interest.

5. 2. The method of claim 1, wherein the plurality of intercellular communication networks indicates which pathways are activated within the plurality of microdomains and what the relationships are between the plurality of intercellular communication networks across the plurality of microdomains.

6. 10. The method of claim 1, wherein the heterogeneous intercellular communication network defines the information flow that needs to be inhibited by a drug to slow or reverse disease progression in the particular patient.

Images