Analytical systems and methods

Abstract

A computer-implemented method of analyzing assays performed at respective assay sites of an array or microarray comprising a plurality of assay sites, the method comprising: receiving at least one image of the assay sites of the array or microarray; for each of the assay sites, processing the at least one image to determine at least one measure representative of a degree of a reaction at the assay site; for each of the assay sites, determining one or more parameters for the assay site, wherein the one or more parameters for at least one of the assay sites of the array or microarray are different from the one or more parameters for at least one other of the assay sites of the array or microarray; and for each of the assay sites, determining a size of the reaction at the assay site from the at least one measure for the assay site and the one or more parameters for the assay site. Also disclosed is a computerized analysis system configured accordingly.

Description

Technical Field

[0001] This disclosure relates to analytical systems and methods for analyzing reactions, such as analytical systems and methods for analyzing determinations. Background Technology

[0002] Analytical systems can be used to automate the analysis of chemical determinations. Such determinations are typically performed in a conventional array of printed areas configured to contain reactants and test samples. These determinations typically test for the presence or level of an analyte in the test sample. The response, usually given as a change in the degree of opacity, color, size, or other detectable variation, relates to the presence or level of the analyte. Analytical systems are equipped with sensors, typically digital cameras, to identify the response in each print in the array. In this way, for example, a digital camera image can be viewed, and the response (e.g., the degree of change in color or opacity) can be identified to determine the presence or level of the analyte in any given printed area.

[0003] Typically, the same type of measurement is performed in each printed area, but it may be desirable to perform different types of measurements in different printed areas of the array.

[0004] Manually reviewing the responses shown in digital camera images is possible, allowing for a degree of expertise and human judgment, but it is often slow, labor-intensive, and prone to human error. Automated response determination can be faster and more productive, but it can sometimes be associated with errors, especially if there are unforeseen or inexplicable effects or if the response is marginal or critical (often referred to as an edge case).

[0005] At least some examples of this disclosure seek to improve the automated analysis of measurements performed on an array or microarray comprising multiple printed areas, particularly in cases where different measurements are performed in different printed areas. Summary of the Invention

[0006] Various aspects of the invention are defined in the independent claims. Preferred features are defined in the dependent claims.

[0007] According to a first example of this disclosure, a computer-implemented method is provided for analyzing multiple reactions performed at corresponding sites in an array comprising multiple sites, the method comprising:

[0008] At least one image from multiple sites of the receiving array;

[0009] For each of a plurality of sites, at least one image is processed to determine at least one image metric representing the degree or magnitude of the response at that site;

[0010] For each of a plurality of loci, one or more parameters are determined for that locus, wherein the parameters for at least one of the loci are different from the parameters for at least one of the other loci.

[0011] For each of a plurality of sites, the size of the response is determined based on at least one image metric used for that site and one or more parameters used for that site.

[0012] The array can be, includes, or is included in a microarray, such as a multiplexed microarray or a hybrid array.

[0013] The reaction may be or includes a assay. The sites of the array may be or include assay sites, wells, printed areas, or reaction sites of the array. The reaction may include chemical and / or biological reactions. The assay may include one or more reagents at the site. The assay may include adding a sample to a reagent. The presence of an analyte in the sample may cause a reaction with the reagent. The reaction may produce a spot at the site, the spot being generated by a measurable change caused by the reaction and indicating that measurable change. At least one property of the printed material may cause a reaction that produces a spot, characterized by a change in one or more of opacity, intensity, color, and / or size, which may indicate the extent or magnitude of the reaction. One or more standards and / or parameters for the site may be, include, or represent standards and / or parameters of the spot or representing an image portion of the spot. The magnitude of the reaction may include or indicate a spot grade. The determination of the reaction size may be, be included in, or may include a grading or determination of the spot grade. The magnitude of the reaction or the spot grade may be qualitative, such as whether a reaction has occurred, or quantitative, such as a value indicating the magnitude or extent of the reaction. At least one measure for the site may be or include a measure of the spot.

[0014] At least one image may include one or more images of all the measurement sites among the multiple measurement sites, or a collection of images that individually cover one or some of the measurement sites among the multiple measurement sites but collectively cover all the measurement sites among the multiple measurement sites.

[0015] Criteria and / or parameters may include at least one criterion having an associated threshold or range. At least one criterion with an appropriate threshold or range may indicate the size of the reaction or the blob grade. The size of the reaction or the blob grade may include one or more of the following: a reaction, no reaction, or one or more values ​​representing the size of the reaction, and / or an indication of uncertainty regarding the size of the reaction. An indication of uncertainty regarding the size of the reaction at a site may be or include determining that at least one image metric determined for that site falls between a range of values ​​indicating no reaction and a range of values ​​indicating a reaction. Determining the size of the reaction at a site may include comparing at least one metric used for that site with a reference function or one or more values ​​that include or depend on the criteria and / or parameters used for that site. Determining the size of the reaction at a site may include determining whether at least one metric used for that site is above or below at least one threshold, or within or outside at least one range, or meets or does not meet at least one of these criteria, thereby determining the size or extent of the reaction or the blob grade.

[0016] The criteria and / or parameters used for a site may depend on one or more of the following: the type of assay performed at the site, the location of the site on the array, and / or the type of sample at the site.

[0017] The criteria and / or parameters used for a site can be unique, or at least customized, adjusted, or selected for that site. The criteria and / or parameters can be preset, for example, determined before analyzing multiple reactions. The criteria or parameters can be used and / or specific to the type of assay performed at the site, the site's location on the array, and / or the sample type at the site. The method may include, for at least one or each site, determining or identifying the type of assay performed at the site, the site's location on the array, and the sample type at the site, and selectively applying criteria or parameters specific to the determined assay type, site location on the array, and / or sample type for the corresponding site.

[0018] The standards and / or parameters used for sites can be fixed. The standards and / or parameters for each given assay type, site location on the array, and / or sample type can be fixed. The standards and / or parameters used for sites can be selected from multiple predetermined standards and / or parameters associated with different assay types, site locations on the array, and / or sample types, but once selected, they can be fixed. The selection of standards and / or parameters for at least one or each corresponding site can include selecting predetermined parameters and predetermined standards and / or parameters for the assay type, site location on the array, and / or sample type of that corresponding site. Multiple sites (e.g., assay sites or printed areas) can be included in an array of sites. Sites can be arranged in rows. Each row of sites can be staggered relative to at least one or each of the adjacent rows of sites. The array can include alternating staggered rows of sites, wherein each alternative row of sites can be aligned. The array can conform to the American National Standards Institute (ANSI) / Laboratory Automation and Screening Association (SLAS) standards. The array can include one or more control sites, which can be located at one or more or each of the outer corners of the array or the array of sites. Control sites can be placed at the vertices of a parallelogram.

[0019] At least one measure for a site may include, depend on, or represent the pixel intensity of an image region representing at least a portion of the site (e.g., a blob at the site). At least one measure for a site may include, depend on, or represent the average pixel intensity of an image region representing at least a portion of the site (e.g., a blob at the site). At least one measure for a site may include, depend on, or represent the difference between the pixel intensity or average pixel intensity of an image region representing at least a portion of the blob at the site and the background pixel intensity or average background pixel intensity (e.g., for an image portion representing a region that is not a blob). At least one measure for a site may include, depend on, or represent a measurement of the diffusion value or standard deviation of the background pixel intensity.

[0020] At least one metric for a site may be or include the signal-to-standard-deviation ratio (SSR), for example, the difference between the average pixel intensity of at least a portion of the image region representing the spot at that site and the average pixel intensity of at least a portion of the background, divided by a measure of the diffusion value or standard deviation of the background pixel intensity of at least a portion of the background.

[0021] At least one metric for a site may include, depend on, or represent a measurement of the diffusion value or standard deviation of pixel intensity, or an average pixel intensity representing at least a portion of the image region of the spot at that site. At least one metric for a site may include, depend on, or represent a percentage or fraction of pixels representing the spot at that site, which are at or above a pixel intensity threshold.

[0022] The method may include at least one of the following: cropping an image around a site in an array. The method may also include meshing the image into segments, wherein each segment encloses a blob and / or site.

[0023] The determination of one or more or each of at least one metric of cropping, meshing, and / or speckling may include the determination of parameters using associated cropping, meshing, and / or metric, which may be shared with one or more or each or all other specks or sites having the same assay type; for example, common parameters are common to a assay type. The criteria and / or parameters used in determining the response size of a site based on at least one image metric used for the site may be unique to that site.

[0024] This method may include, for example, preprocessing or cleaning the image by filtering or smoothing it. The filtering or smoothing can be configured to selectively remove noise from the image.

[0025] Cropping at least one image may include determining a cropping region of the image, which may be a subset of the total region of the image, wherein multiple sites are shown. Cropping an image may include cropping the image around multiple sites on an array. The method may include identifying one or more or each control site, which may define the outer corner of the array of sites on the array, and may define or be used to define the corner of the cropping region of the image.

[0026] This method may include detecting or identifying an image portion corresponding to a blob in a reaction site. The method may include, for example, using object recognition techniques to detect or identify the shape of a cropped image that matches a description of the blob. Detecting or identifying the image portion corresponding to a blob in a reaction site may be based on one or more parameters, such as determining a specificity parameter or a site-specific parameter.

[0027] The method may include filtering identified or recognized blobs in an image. The method may also include filtering out identified or recognized blobs where at least one metric is below a threshold, which may be a preset or predetermined threshold and may be included in or derived from at least one parameter in the parameters (e.g., a determination-specificity parameter or a site-specificity parameter).

[0028] The method may include identifying the location of a reference point, which may be a control site. The reference point (e.g., a control site) can be identified by identifying the extreme or farthest sites shown in the image and optionally applying a known or predicted geometry of the array. The method may include filtering out identified or determined spots using a spatial threshold based on the known or predicted geometry of the array (e.g., a known or predicted arrangement of sites on the array, which may include an arrangement relative to the reference point (e.g., a control site)). The method may include filtering out those spots that are not within the area defined by the control site at each corner, e.g., outside the area where a site was expected to be found.

[0029] The method may include meshing an image, which may include determining a grid comprising multiple segments, the grid being based on the determined locations of reference points (e.g., control sites) and / or the determined locations of other sites or spots and / or the expected geometry of the sites on an array. Each segment may include a single corresponding determination site of the array. These cells may be non-overlapping. These cells may correspond to the arrangement of sites on the array. Meshing may be based on one or more parameters, such as determination-specific parameters or site-specific parameters.

[0030] Determining at least one measure of a site may include at least one measure determined directly from an image. Determining at least one measure of a reaction site may include indirectly determining at least one measure from at least one measure determined directly from an image. Determining at least one measure of a site may be based on one or more parameters, such as determining specificity or site-specific parameters.

[0031] Non-exhaustive examples of metrics include: average image background pixel value or intensity, average standard deviation or diffusion value of image background value or intensity, average and / or standard deviation or diffusion value of pixel value or intensity of a blob, circumferential radius of a blob, shape radius and / or isocircumference coefficient of a blob, edge sharpness or diffusion value of a blob, circumferential signal-to-standard deviation ratio, circumferential average, circumferential standard deviation, circumferential background average and / or standard deviation or diffusion value, circumferential intensity value above the background, average, standard deviation and / or diffusion value of image background pixel value or intensity of a blob, signal-to-standard deviation ratio (SSR) of a blob and / or the difference between the pixel value or intensity of at least a portion of the blob and the pixel value or intensity of the background, etc.

[0032] Determining one or more parameters for a site may include identifying or determining relevant parameters associated with a measurement performed at the relevant site and / or associated with a specific site, and applying those parameters to that site. Parameters may be predetermined for a given array, measurement type, sample type, location, position within the array, and / or site, etc.

[0033] In one example, the parameters for the location can be retrieved from an input file or entered by the user using a user input device such as a keyboard and / or the like.

[0034] The method may include identifying or receiving one or more identifiers that indicate at least one or each site in an array or on an array, or a measurement or type of measurement performed at at least one or each site in an array. Determination of one or more parameters for a site may include, for example, looking up or determining the parameter from a lookup table, database, data warehouse, or function that associates the parameter with an array, measurement type, sample type, and / or site based on the determined identifier. One or more indicators may be received from an input device and / or from a data warehouse, for example, as part of a configuration file. Input devices may include user input devices (e.g., a keyboard for receiving user input), machine-readable data devices (e.g., a scanner for barcodes or QR codes or other machine-readable codes, an RFID tag reader, and / or an optical tag reader, etc.). One or more indicators may be received from a Laboratory Information Management System (LIMS). The method may include reading or otherwise receiving the identifier, and may include subsequently using the identifier to access one or more criteria and / or parameters associated with the identifier, for example, from a data warehouse.

[0035] Parameters can be based on past measurements, historical data, experimental measurements, modeling data, or predictive data and / or similar data. The determination of one or more parameters for a site may include using machine learning, artificial intelligence, or some other learned or trained algorithm or function and / or using historical data or modeled data and / or the like to determine parameters for one or more or each different measurement type and / or site.

[0036] These parameters can include global parameters. Global parameters can be parameters that are functions of the measurement; that is, global parameters can vary with different measurements and / or be associated with the corresponding measurement type. The same value of a global parameter can be applied to multiple sites or spots. One or more of the following can be performed based on global parameters: image meshing, image cropping, and / or determination of one or more measures of the magnitude or extent of the response at each site representing the site.

[0037] These parameters may include unique parameters. Unique parameters can be parameters that are functions of a specific site or spot on the array, for example, as a function of the assay type and the site or spot on the array. The value of the unique criterion can vary individually on a site-by-site basis. At least one or more, or all, of the parameters used to determine the magnitude of the response at a assay site based on at least one image metric used for the corresponding assay site can be unique parameters.

[0038] Non-exhaustive examples of parameters that can be used as global standards include: the value of the maximum average intensity of the image background and / or the maximum standard deviation or diffusion value of the image background intensity, etc.

[0039] Parameters used for sites on an array can specify which values ​​in at least one metric indicate a response or no response at that site on the array. Parameters can specify one or more thresholds, criteria, and / or ranges that specify the magnitude or extent of the response, for example, whether a spot indicates a response, no response, or indeterminacy.

[0040] Examples of non-exhaustive parameters include one or more, or each of the following: the maximum average intensity value for a reaction site, the maximum standard deviation or diffusion value for the intensity of a reaction site, the limit for the signal-to-standard deviation ratio for a reaction site, the threshold for pixel values ​​or intensity on the background of a reaction site that is considered empty or filled, the maximum and / or minimum circumferential radius or shape radius for a detection shape considered as a blob, the maximum acceptable background cv value for validating SSRs, the SSR value range for blob reactivity, the SSR value range for blob non-reactivity, the average range for blob reactivity, the average range for blob non-reactivity, the maximum acceptable background cv value for validating SSRs, the SSR value range for circumferential reactivity, the SSR value range for circumferential non-reactivity, the delta value range for circumferential reactivity, and / or the delta value range for circumferential non-reactivity, etc.

[0041] Determining the magnitude of a response at a assay site or blob may include applying one or more measures and parameters for the reaction site or blob to one or more logic tests, wherein the results of the one or more logic tests indicate whether a response has occurred at the site, the magnitude or extent of the response at the site, the activity and / or the magnitude or extent of the activity at the site, or the presence of an analyte at the site. Determining the activity and / or the magnitude of the activity at the site may include determining the presence of detectable substances, such as blobs, the shape of non-background effects, but without knowing whether the detected activity is a blob or a response. Logic tests may include comparing one or more measures for the site with one or more parameters for the site to determine the magnitude of the response.

[0042] A second example according to this disclosure is a processing system configured to implement the method of the first example to parse one or more measurements performed at corresponding measurement sites from a plurality of measurement sites in an array.

[0043] The processing system can be configured to receive at least one image of a measurement site in an array. The processing system can be configured to process at least one image for each measurement site to determine at least one metric representing the magnitude or extent of the response at that measurement site. The processing system can be configured to determine one or more parameters for each measurement site, wherein the parameters for at least one measurement site in the array are different from the parameters for at least one other measurement site in the array. The processing system can be configured to determine the magnitude of the response at each measurement site based on at least one metric and one or more parameters for that measurement site.

[0044] The processing system may include data storage. The processing system may include a communication module. The processing system may include one or more output devices. The processing system may include one or more input devices.

[0045] The processing system may include one or more processors, which may be single-core or multi-core processors. The one or more processors may include one or more central processing units, graphics processing units, math coprocessors, and / or tensor processing units, etc. Data storage may include solid-state memory, magnetic memory, and / or optical memory, etc. The communication unit may be configured to communicate via wired and / or wireless communication, and may be configured to communicate with remote and / or local systems, for example, via networks such as LANs, WANs, the Internet, one or more cellular networks, Ethernet, and / or fiber optic networks. At least one output device may include a display or other visual output device, an audio output device, and / or a haptic output device.

[0046] At least one input device may include a keyboard, touchscreen, trackball, touchpad, joystick, and / or voice recognition-based input device. Input devices may include machine-readable data devices such as scanners for barcodes or QR codes or other machine-readable codes, RFID tag readers, and / or optical tag readers. The processing system may be configured to read or otherwise receive codes associated with an array or sites on an array, and may include subsequent use of the code to access, for example, one or more criteria and / or parameters associated with that code from a data warehouse.

[0047] The processing system can be configured to look up or determine parameters from configuration files, lookup tables, databases, data warehouses, or functions that associate relevant parameters with relevant assay types, sample types, and / or loci. The processing system can be configured to communicate with at least one remote processing system. Remote processing systems may include servers, cloud computing resources, workstations, and / or personal computers, etc. Remote processing systems may include remote data storage.

[0048] A third example according to this disclosure is an analysis system for analyzing one or more measurements performed at corresponding measurement sites in a plurality of measurement sites of an array, the analysis system including the processing system of the second example described above. The analysis system may include at least one imaging device configured or configurable to collect at least one image of the array and transmit the at least one image to the processing system.

[0049] The processing system can be configured to receive at least one image of a measurement site in an array. The processing system can be configured to process at least one image for each measurement site to determine at least one metric representing the magnitude or extent of the response at that measurement site. The processing system can be configured to determine one or more parameters for each measurement site, wherein the parameters for at least one measurement site in the array are different from the parameters for at least one other measurement site in the array. The processing system can be configured to determine the magnitude of the response at each measurement site based on at least one metric and one or more parameters for that measurement site.

[0050] A fourth example according to this disclosure is a computer program product including instructions that, when implemented on a processing system, cause the processing system to perform the method of the first example of this disclosure. The computer program product may be embodied on a non-transient and / or tangible computer-readable medium.

[0051] In any other aspect, example, or embodiment of this disclosure, the individual features and / or combinations of features defined above in any aspect, example, or embodiment of this disclosure or in any specific embodiment of this disclosure may be used separately and individually, alone, or in combination with any other defined features.

[0052] Furthermore, this disclosure is intended to cover any device configured to perform any feature described herein in connection with the methods of using or producing any device feature described herein and / or the methods of using or manufacturing any device feature described herein. Attached Figure Description

[0053] These and other aspects of this disclosure will be described by way of example only, with reference to the accompanying drawings, wherein:

[0054] Figure 1 Schematic diagram of a computerized measurement and analysis system;

[0055] Figure 2 :use Figure 1 Examples of annotated images measured on an array collected by the system;

[0056] Figures 3 to 7 A graph showing the changes in the measured properties of a specific metric used in the spot grading process;

[0057] Figure 8 : By Figure 1 An overview of the measurements and analyses performed by the computerized measurement and analysis system shown;

[0058] Figure 9A :constitute Figure 8 An overview of the image analysis shown as part of the measurement analysis;

[0059] Figure 9B : Figure 9A A detailed flowchart of image analysis;

[0060] Figure 10 : Figure 9A and Figure 9B Another flowchart of image analysis;

[0061] Figure 11 A schematic diagram of the mesh matching process as part of the trimming process;

[0062] Figure 12 : used to pass Figure 11 A schematic diagram of the mesh of the microarray generated by the meshing process;

[0063] Figure 13 Determine the composition Figure 9A , Figure 9B and Figure 10 A flowchart of the determination of measurement process as part of image analysis;

[0064] Figure 14 As Figure 17 A flowchart of the segmentation process for determining a portion of the metric shown;

[0065] Figure 15 :exist Figure 8 A flowchart of the operation of the hierarchical rule engine used in the measurement and analysis;

[0066] Figure 16 The diagram illustrates the method used to determine relative to Figure 15 The diagram illustrates a flowchart of the logic used in the hierarchical rule engine to determine the applicability of analysis;

[0067] Figure 17 The diagram shows the situation relative to... Figure 15 The diagram illustrates the flowchart of the spot determination logic used in the hierarchical rule engine;

[0068] Figure 18 The diagram shows the situation relative to... Figure 15 The flowchart shown is a process for the object detection logic used in the hierarchical rule engine.

[0069] Figure 19 The diagram shows the situation relative to... Figure 15 The flowchart shown is a diagram of the bright object determination logic used in the hierarchical rule engine;

[0070] Figure 20 The diagram shows the situation relative to... Figure 15 The flowchart illustrating the speckle classification logic used in the hierarchical rule engine is shown; and

[0071] Figure 21 The diagram shows the situation relative to... Figure 15 The diagram illustrates the flowchart of the spot classification confirmation logic used in the hierarchical rule engine. Detailed Implementation

[0072] Figure 1 A measurement analysis system 5 is shown for analyzing measurements performed in an array 10 (e.g., a multiplexed microarray or hybrid array) comprising multiple measurement sites or printed regions 15, wherein individual measurements can be performed at each measurement site 15 or printed region. System 5 includes one or more sensors, in this example in the form of a digital camera 17, configured to collect images of the array 10. Although... Figure 1 Only one camera 17 is shown, but more than one camera 17 may be provided. In this example, camera 17 is configured to capture images of the entire array 10; for example, the field of view 18 of camera 17 includes the entire array 10 or at least all of the measurement sites 15 of array 10. However, camera 17 may be configured to capture images of only some of the measurement sites 15 of array 10, or different cameras 17 may be configured to capture images of different subsets of the measurement sites 15 of array 10 or to capture array 10 from different angles. System 5 includes an analysis system 20 that receives images collected by camera 10 and is configured to analyze the images to determine the state of the response at a specific measurement site 15.

[0073] The analysis system 20 includes a processing system 25, a data storage unit 30, a communication module 35, one or more output devices 45, and one or more user input devices 40. The processing system 25 includes one or more processors, which may be single-core or multi-core processors. The one or more processors include one or more central processing units and optionally include one or more graphics processing units, math coprocessors, and / or tensor processing units, etc. The data storage unit 30 may include solid-state memory, magnetic memory, and / or optical memory, etc. The communication unit 35 may be configured to communicate via wired and / or wireless communication. In this example, the communication unit 35 is configured to communicate with remote and / or local systems, for example, via networks such as LAN, WAN, the Internet, one or more cellular networks, Ethernet, and / or fiber optic networks. At least one output device 45 may include a display or other visual output device, an audio output device, and / or a haptic output device. At least one input device 40 may include one or more of the following: a keyboard, a touchscreen, a trackball, a touchpad, a joystick, a voice recognition-based input device, an RFID tag reader, and / or a barcode or QR code reader, etc.

[0074] The analysis system 20 can be configured to communicate with at least one remote processing system 50. The remote processing system 50 may include servers, cloud computing resources, workstations, and / or personal computers, etc. The remote processing system 50 includes a remote data storage device 55. In this way, the processing system 25 can be used to perform any or all of the method steps described herein, particularly method steps related to data processing and / or data storage, and the at least one remote processing system 50 or the method steps described herein may be distributed between the processing system 25 and the at least one remote processing system 50.

[0075] Depend on Figure 1 An example of images collected by the system from array 10 in Figure 2 As shown in the diagram. Some assay sites 15a include spots representing a reaction (e.g., a reaction indicating the presence of an analyte), some assay sites 15b indicate no reaction, which may indicate a lack of analyte, and some assay sites 15c are empty. Spots are generally associated with variations in opacity or color or the like, representing the size or extent of the reaction. Importantly, the inventors have identified cases in which different assay sites 15a, 15a' of array 10 are used to perform reactions with different reaction intensities. Different reaction intensities can be associated with different assays, i.e., different types of assays are performed at different assay sites 15, and at least one assay or each assay can be associated with a different reaction intensity of at least one other assay. In another example, the reaction intensity can be sample-dependent or can vary depending on the assay site.

[0076] Analysis of the reaction at each assay site 15 determines one or more measures for the reaction at assay site 15, and logic dependent on one or more parameters (e.g., comparison with a threshold) is applied to produce a measurement of the reaction state (e.g., whether a reaction has occurred (i.e., qualitative determination), or a value for the magnitude or degree of the reaction or activity (i.e., quantitative determination), or an indication of active or inactive or similar substances, etc.). This can be achieved by selecting a threshold for all reactions in the reaction, for example, selecting assays and samples for all assay sites at assay sites. However, the inventors have noted that the dependence of each analytical property on a given measure can vary significantly and differently for different assays.

[0077] This can be seen from Figures 3 to 7 From this, we can see that Figures 3 to 7 Each example illustrates the variation of different properties (accuracy, specificity, and sensitivity in this case) and exemplary metrics (marginal metrics in this case) for different determinations. Thus, in these examples, if for... Figure 3 The measurement shown here sets a common edge metric threshold, which may require 20 or more edge metric thresholds. However, for Figure 4 As shown in the diagram, an edge metric threshold of 20 or more may result in undesirable low sensitivity for the measurement, while an edge metric threshold of 5 to 10 may be more preferable. Figures 3 to 7 As can be seen in each of the measurements shown, different measurement thresholds may be better for different measurements. After identifying these issues, the measurement analysis system 5 is specifically configured to allow for enhanced analysis of the array, which is used to perform different measurements at different measurement sites 15 in the same array 10.

[0078] exist Figure 8 The document outlines a computerized method for analyzing array 10 to determine the magnitude or extent of a response at assay site 15 of array 10. Although this method can be... Figure 1 The processing system 25 shown is used to perform the process, but one or more steps or each step can be performed by a remote processing system 50, or different steps can be performed by both the processing system 25 and the remote processing system 50, so that the processing is effectively distributed among them.

[0079] Step 705 indicates the performance of the measurement at each of the plurality of measurement sites 15 in the microarray 10. The performance of individual measurements is known in the art; however, in the present case, different measurements are performed at at least some of the measurement sites 15 in the microarray 10 and captured in the image 805 collected by the measurement analysis system 5 in step 710.

[0080] Algorithm 810 provides image analysis as part of the measurement analysis, and Figure 9A , Figure 9B and Figure 10 This is shown in further detail below.

[0081] Processing system 25 and / or remote processing system 50 receive images 805 of array 10 collected by camera 17 (see image 805). Figure 1 Processing system 25 and / or remote processing system 50 implement... Figure 8 and Figure 9A The algorithm 810 shown processes image 805 to determine a measure indicating the magnitude or extent of the response at each assay site 15. Algorithm 810 uses various parameters 815 to determine this measure.

[0082] The parameter 815 used to determine the metric is global and can be applied regardless of the measurement being performed. However, the metric analysis performed by the grading engine uses specific and non-general parameters; that is, different parameters can be used for different measurement sites 15 or groups or subsets of measurement sites 15 within the same array 10. For example, different measurement types can be associated with different parameters. Because array 10 is used to perform different measurements included in image 805, the parameters used for at least one measurement site 15 will differ from the parameters used for at least one other measurement site 15 within the same array 10. The parameters used for each measurement type are typically predetermined, and appropriate predetermined parameters are selected for that site based on the type of measurement performed at each different measurement site. The measurement type performed at each different measurement site and / or the parameters associated with that measurement type are typically provided in the configuration file. In terms of completeness, the provision of parameters is not inherently limited to this, and in other feasible examples, the type of measurement performed in each different printed area can be determined based on user input received from one of the input devices, such as a keyboard or a reader of an identifier (e.g., a QR code, RFID tag, holographic tag, or barcode) provided on or with the array. For example, the type of measurement for each measurement site 15 can be input by the user using a keyboard or using an identifier of the array that can be read by a reader, and pre-stored data for that measurement type can be retrieved from a database, lookup table, or other form of data file that can be stored in the data storage 30 and / or remote data storage 55, etc. Additionally or alternatively, parameters can vary based on sample type, sample size, location of individual measurement sites in the array 10, and / or other factors that affect the intensity of the reaction or the magnitude or extent of the response to the reaction. In this case, it should be understood that parameters for each measurement site 15 or a group of measurement sites 15 can be determined in a manner similar to that described above.

[0083] Algorithm 810 is configured to: parse an image; identify regions of the image associated with different assay sites 15; determine one or more measures for the regions of the image associated with different assay sites 15; and determine the magnitude or degree of the response to the different assay sites 15 (e.g., a value for the magnitude or degree of reactivity or non-reactivity or reactivity or activity) based on the measures and parameter 815.

[0084] like Figure 9A , Figure 9B and Figure 10As shown, the output from algorithm 810 may include a metric 820 indicating the magnitude or extent of the response at one or more or each of assay sites 15. The metric for assay site 15 is processed by a grading engine to determine the magnitude or extent of the response (e.g., spot grade) 720. This metric may be a simple indication of whether there is a response or not, or it may be more granular, for example, giving a value for the magnitude or extent of the response at the assay site. The determined spot grade can then be presented for user interpretation 725. The output may also include one or more errors and / or alerts 825, for example, if the algorithm fails to determine the magnitude or extent of the response at a given assay site 15, or if some other error occurs during the process.

[0085] Figure 9B The text shows that in Figure 8 and Figure 9A A more detailed overview of the process shown is provided below. Figure 9B As shown, the process has images 805 collected by the array 10 from the camera 17 as input.

[0086] Cropping 905 inputs image 805 to tightly crop image 805 to a region of image 805 that includes assay sites 15 (and any other features used in the process, such as control assay sites / spots), based on cropping parameters 910 used to control the cropping process 905. The microarray 10 can be provided in standardized sizes, scales, and layouts. The cropping process aims to reduce the region of the original image 805 to an image region associated with the array of assay sites 15 and any other features used in the process (e.g., control assay sites 15 / spots). Cropping processes are generally known, and any suitable cropping process can be used. Examples of suitable cropping processes include identifying corner control spots and using these corner control spots to determine the array region in the image where the array of assay sites 15 is located. This region is defined based on the positions of the four control spots and the safety margin necessary to take into account spot placement tolerances. The image can then be cropped around this array region.

[0087] The cropped image output from cropping stage 905 is meshed 915, which divides the cropped image into segments 925 (see...). Figure 2Each segment 925 of the image includes a single measurement site 15. Gridding 915 is performed based on gridding parameters 920 used to control the gridding process 915. Gridding parameters 915 can be common parameters specific to the measurements in the experiment. A grid can be overlaid on the image. This includes determining the most probable location of the grid based on the position of corner spots. The segment 925 of the grid model corresponding to the corner control spots is known to the array 10, thus allowing the grid model to be applied to the image 805 using the determined location of the corner control spots, as shown. Figure 11 As shown in the image.

[0088] The meshing process 915 includes the geometric definition of the layout of the array based on the measurement sites and constructs a model of the array of measurement sites 15 in array 10 based on the positional tolerances of the spots in measurement sites 15. The mesh is in... Figure 12 As shown, the grid is organized by rows and columns of segments 925, each segment 925 comprising a single measurement site 15. The print field corresponding to each individual measurement site 15 comprises a single spot, and the spot does not contact adjacent spots. The result is a grid organized by rows and columns of print fields corresponding to the segments 925 of the grid. The grid model (grid descriptor) is stored in data storage 30 and / or remote data storage 55.

[0089] Once image fragment 925 is generated, a spot measurement 930 can be generated for a given measurement site 15 based on measurement generation parameters 935.

[0090] The blob metric 930 can be any metric of the image associated with the blob, which can be used to characterize the blob indicating a response in a given determination site 15. For example, at least one of the blob metrics 930 for a given blob may include or represent the pixel intensity of an image region representing the blob. A specific example of this is a blob metric 930 for a given blob that includes or represents the average pixel intensity of at least a portion of the image region representing the response site and / or the difference between the average pixel intensity of the image region representing the blob and the average pixel intensity of the background. A specific example of a suitable metric for each blob is the standard deviation ratio (SSR), which is a measure of the diffusion value (e.g., standard deviation) of the background pixel intensity of at least a portion of the image representing the given blob, divided entirely by the difference between the average pixel intensity of at least a portion of the image representing the given blob and the average pixel intensity of at least a portion of the background.

[0091] The blob measurement calculation 930 can be designed such that the measurement calculation is repeatable and generally the same regardless of the blob to be resolved, for example, by using an average pixel intensity value and / or by using a mask to select pixels from specified regions of the blob and background. Thus, the measurement generation parameters 935 can also generally be generic regardless of the measurement type or common among the blobs.

[0092] Once a spot metric 930 has been determined for any spot of interest, a grading engine 940 can be used to evaluate that spot metric to determine the magnitude or extent of the response at the corresponding assay site 15, which includes the corresponding spot. The grading engine 940 applies spot-specific (also assay-specific and / or sample-type specific and / or analogous) grading parameters 945 and / or grading rules 950 to the spot metric determined for a given spot to determine the magnitude or extent of the response (e.g., spot grade) for that given spot / assay site 15. As described above, since different spots / assay sites 15 are used for different assays and / or sample types, the grading parameters 945 and / or rules 950 applied by the grading engine 940 for one or more spots / assay sites 15 are made different from those applied for some other spots / assay sites 15 to determine the magnitude or extent of the response (spot grade) for each spot / assay site 15.

[0093] In this way, common or general parameters are applied where appropriate, such as for one or more of the calculations of trimming 905, gridding 915, and / or spot measurement 930, which facilitates efficient handling of processes to which individual parameterization has little impact. Conversely, customized rules 950 and / or parameters 945 are determined and used by the grading engine 940 for different spots / measurement sites 15 and / or a different set of spots / measurement sites 15 on the same array 10, where parameters specific to the measurement, sample type, or other factors that may significantly affect the performance and quality of the measurement in progress are used. However, it will be understood that in other examples, one or more, or each, of the calculations of trimming 905, gridding 915, and / or spot measurement 930 may be performed using measurement site 15-specific parameters (e.g., measurement, spot location, and / or sample type specificity).

[0094] exist Figure 10 The diagram shows the method used to illustrate. Figure 9BAn alternative flowchart of the method is shown, illustrating the input image 805 provided and loaded 1005. Parameters in the form of a construct mesh descriptor 1007 are generated and used to crop 905 and mesh 915 of the input image 805. A metric 1012 for each segment 925 of the mesh is determined (step 1008) and output as an image analysis output file to a blob grading engine 940, which uses the metric to grade the blobs in each segment 925 / well 15. The results of the image analysis are also output (step 1014). The output in step 1014 may include, for example, the determined metric 1012 and any additional image analysis data (e.g., error indications) that may be useful for or requested by further investigation.

[0095] The segment 925 of the mesh model corresponding to corner reference spot 1747 is known for array 10, thus allowing the mesh model to be applied to image 805 using the determined locations of corner reference spot 1747 and other spots, such as... Figure 11 and Figure 12 As shown in the image.

[0096] exist Figure 13 The text shows that in Figure 9B and Figure 10 Further details on the generation of the metric 930 / 1008 shown are provided. Metric 930 / 1008 is determined based on the image of each segment 925 / printed domain in the gridded image. The metric 930 / 1008 for segment (printed domain) 925 is input into a grading rule engine 940, which uses the metric to determine the grade or reactivity of each measurement performed in the relevant segment (printed domain) 925.

[0097] like Figure 13 As shown, the metric generation process includes calculating a shape (e.g., a circular shape) metric 1732 1805. The shape metric is identified by using object recognition techniques to identify the shape (e.g., but not limited to a circular shape) of the reaction spots in the fragment, and then determined. The shape metric may include one or more of the following: center coordinates, circumferential radius, circumferential mean (i.e., the average pixel value of the circumferential region defined by the circumferential coordinates and radius), circumferential standard deviation (i.e., the standard deviation of pixel values ​​in the circumferential region), circumferential background mean (i.e., the average pixel value in the background region), circumferential background standard deviation (i.e., the standard deviation of pixel values ​​in the background region), circumferential SSR (the ratio of signal to standard deviation for the circumferential region), and / or any alternative metric indicating any of the above.

[0098] The metric generation process also includes calculating a domain metric 1810 for each segment (printed domain). Domain metrics can include measures such as the domain mean (the average of the pixel values ​​in the region defined by the coordinates and radius of the entire segment) and the domain standard deviation (the standard deviation of the pixel values ​​in the region defined by the coordinates and radius of the entire segment), and / or similar metrics. Domain metrics are calculated based on the position and size of individual segments of the mesh.

[0099] In addition to calculating domain metric 1810, the process also includes calculating a default domain metric 1815 for each segment (printed domain), where the default domain is a circle with a default radius placed at the center of the respective segment. Examples of default domain metrics include one or more, or each of, the following: the average pixel intensity of the default domain, the standard deviation of the pixel intensity of the default domain, the average pixel intensity of a region in the background area of ​​the default domain, the standard deviation of the pixel intensity of a region in the background area of ​​the default domain, and the signal-to-standard-deviation ratio (SSR) of the default domain and / or the like.

[0100] The metric generation process further includes calculating blob metrics. Blob metrics are extracted from each segment (printed domain) using a segmentation method. The purpose of the segmentation step is to detect the presence of an object (e.g., some form of activity or "blob," which may or may not be caused by the measurement response and / or an indicator of the size of the measurement response) in the segment under consideration (printed domain), and if present, to determine the location and size of the object. Appropriate metrics are calculated for both the object and the surrounding background. If no object is found in the segment (printed domain), no blob metric is generated. If an object is detected, a blob metric is determined and subsequently used in the analysis performed by the grading rule engine to determine whether the object is a blob, and if so, to determine the blob's grading. Examples of possible blob metrics include: blob center coordinates, blob radius, average pixel values ​​in the region defined by the blob mean and blob radius, standard deviation of pixel values ​​in the region defined by the blob mean and blob radius, average and / or standard deviation of the region defined by the background, blob threshold, blob edge metric, blob isoperimeter coefficient, and / or the ratio of the blob's signal to standard deviation, etc.

[0101] The determination of the blob measurement begins with segmentation step 1820. Segmentation step 1820 is... Figure 14 The details are as follows.

[0102] Figure 14The segmentation process 2005 shown includes the creation 2405 of an eroded image. This includes applying structuring elements of a size and shape set according to erosion parameters 2410, which include data setting the shape and size of the erosion elements. The erosion elements are matrices placed on each image pixel, and the value of the eroded pixel is the smallest input pixel found on that matrix. Erosion is used to determine an evaluation of the local background. The structuring elements can be selected such that, after erosion, the center of the segment reflects the local background.

[0103] For each segment 925, Figure 14 The segmentation process shown includes determining a pixel-intensity-based metric for segment 925. In this example, the metric for each segment 925 is an IQ3 metric 2415 calculated for the corresponding segment 925. The IQ3 metric for segment 925 is the pixel intensity of the image at the center of that segment 925.

[0104] exist Figure 14 In step 2430, thresholding is applied to segment 925. Thresholding converts a color or grayscale image into a black and white image by applying a threshold to pixel intensity; pixels above the threshold are identified as white, and pixels below the threshold are identified as black. The threshold can optionally be dynamically determined (i.e., adaptive thresholding is applied), or it can be provided as a parameter.

[0105] exist Figure 14 In step 2435, the thresholded image output from step 2430 is parsed to determine any contours in the image. Various techniques can be used to identify contours and edges, and an appropriate technique can be selected.

[0106] exist Figure 14 In step 2440, the contours found in step 2435 are examined segment by segment on the basis of fragment 925 to see if they are close to the boundary width threshold of the boundary of fragment 925 (which is provided as input parameter 2005). If so, they are removed. If more than one contour remains after this processing, the largest contour is selected for further processing in step 2450, and the remaining contours are discarded. If no contour is found for a given fragment 925, the fragment 925 is marked as "not segmented" or "no spot found", and no further segmentation processing is performed relative to the fragment 925.

[0107] For those segments 925 that are not marked as "unsegmented", in step 2455, the convex hulls 2705 of the spots 2710 in those segments are determined. The convex hull or convex hull 2705 of a spot is the smallest possible convex shape that includes the spot 2710.

[0108] In step 2460, the minimum closed circle is determined for the corresponding segment. The minimum closed circle is the smallest possible circular shape including the convex hull. The center coordinates and radius of the minimum closed circle are outputs of the segmentation stage.

[0109] For each segment 925, if no spots are found in the segmentation process 1820 (i.e., not segmented), a spot metric is not determined for that segment 925. However, if spots are determined to exist in a given segment 925, the segment 925 is segmented into a spot portion and a background portion in step 1825.

[0110] The image portion representing the measurement site 15 associated with segment 925 can be segmented into a portion representing the spots and a portion representing the background. The portion representing the spots is used to calculate the average intensity of the spots, for example, by taking the average of the intensity values ​​of all pixels included in the region representing the spots.

[0111] The portion representing the background is used to calculate the average intensity and standard deviation (or other measure of diffusion) of the background in the portion of the fragment that does not include the spots (i.e., the average and standard deviation of the pixel intensity values ​​of the pixels representing the background).

[0112] For segments or spots that have been identified as having been segmented (i.e., there is a discoloration or some form of activity), the process can proceed to determine the spot grade or the size or extent of the reaction required to assess each spot or assay site 15 that has not yet been considered non-reactive or a technical error.

[0113] Figure 13 Step 1830 in the process includes pixel filtering of the image to remove atypical pixels with extremely high or low intensity. Image artifacts typically have very high intensity, and this process removes any remaining residue of these artifacts. Pixels with very low intensity can indicate noise. Pixel filtering applies high pixel intensity thresholds and low pixel intensity thresholds, which are parameters of the process, to both the blob pixel group and the background pixel group, and removes those pixels that are below the low intensity threshold and above the high intensity threshold.

[0114] In step 1835, the spot metric used in the spot grading / determination of the size or extent of the reaction is determined. Various metrics can be used. One metric is the signal-to-standard deviation ratio (SSR), which can be calculated for each spot / fragment 925 / assay site 15 using the following formula:

[0115]

[0116] in:

[0117] Mean (Signal) is the average pixel intensity in the filtered spot region;

[0118] Mean(Background) is the average pixel intensity in the filtered background region; and

[0119] StD (Background) is the standard deviation of pixel intensity in the filtered background area.

[0120] Examples of other metrics that can be used as a complement to or alternative to SSR include IQ1, IQ2, IQ3, and IQ4, which are:

[0121]

[0122] Among them, A Signal It is the spot area obtained using the spot radius, and A spot It is the spot area obtained from the segment before pixel filtering (which is a measure of the number of pixels outside the spot radius), or the spot area within the spot radius excluding any pixels that have been removed by filtering (which is a measure of the number of missing pixels within the spot radius).

[0123] IQ3 = the pixel intensity of the image obtained at the center of segment 925.

[0124] IQ4 is the ratio of the percentage of pixels at or above the intensity of a target to the percentage of pixels at the camera's saturation value (typically 255).

[0125] Based on the teachings of this disclosure, other suitable measures may be apparent to those skilled in the art. Any determined measures are output to the grading rule engine 940. The grading rule engine employs these measures, such as circumferential, domain, and / or blobular measures for each measurement site, which typically indicate image characteristics, and uses these measures to determine the blob grade or the size or extent of the reaction, which indicates the characteristics of the measurement performed at the relevant measurement site.

[0126] Hierarchical rule engine 940 based on Figure 15 The operation is performed according to the logic outlined in the document. Figure 15 This is a flowchart outlining the rules used in the spot grading process (i.e., the generation of the measure 930 of spot / measurement site 15) and the operation of the grading engine 940. The application of individual rules by the grading engine 940 will be described in more detail below with respect to Figures 24 to 29.

[0127] In this example, the grading engine 940 uses grading parameters 945 and rules 950 (such as...) to determine the spot grade or the size or extent of the reaction. Figure 9B(As shown) can be global or unique. Global parameters are parameters specific to the assay type, while unique parameters are limited to a function of both the assay and the location of the spot / assay site 15, that is, the parameter depends on the assay site 15 / spot and depends on the assay.

[0128] The blob grading performed by the grading rule engine 940 considers an image 805 of the array of measurement sites 15 in array 10 (appropriately cropped 905 and meshed 915 into fragments) and a metric for each fragment 925 / measurement site 15, and delivers the grading results for the corresponding blob / measurement site 15 / fragment 925. In the specific example shown in Figure 23, the grading engine 940 applies a set of six rules for each fragment 925 / measurement site 15 of interest, namely:

[0129] • Rule 00: Rule 1205 for identifying images unsuitable for analysis;

[0130] Rule 01: Rule 1210 identifies whether fragment 925 is empty or includes an object;

[0131] • Rule 02: For example, rule 1215, which identifies whether segment 925 contains spots by segmentation or circumferential search;

[0132] • Rule 03: Classify spots as reactive, non-reactive, or technically incorrect, or rule 1220: The degree of reaction in spot / test site 15 is indeterminate.

[0133] Rule 04: Combines rule 1225 for detecting bright objects; and

[0134] Rule 05: Confirm the result of Rule 1230.

[0135] While a useful set of rules 1205-1230 is provided above for grading spots / reactions at determination site 15 using the grading rule engine 940, in other examples, some but not all of rules 1205-1230 may be applied, or additional or alternative rules may be applied. (See also...) Figures 16 to 21 To further describe the operation of the system.

[0136] exist Figure 16 The diagram shows an example of a flowchart for determining whether an image is suitable for analysis. If the background of the image in array 10 is sufficiently dark and uniform, then... Figure 16The process or rule shown can be used to determine whether an image of the array of measurement sites 15 in array 10 is suitable for analysis. The process uses a metric 1405 for the entire image, where, in this example, the metric includes the average pixel intensity of the background portion of the entire image and the standard deviation (or some other appropriate measure of diffusion) of the pixel intensity of the background portion of the entire image. The rule uses global parameters (1410), including a maximum acceptable value (or threshold) for the average image background and a maximum acceptable value (or threshold) for the standard deviation or diffusion value of the image. The process includes an implementation rule (1415) that determines whether the average pixel intensity of the background portion of the entire image exceeds the maximum acceptable value (or threshold) for the average image background or whether the standard deviation (or some other appropriate measure of diffusion) of the pixel intensity of the background portion of the entire image exceeds the maximum acceptable value (or threshold) for the standard deviation or diffusion value of the image, respectively. If so, a technical error is flagged 1420, and the process for that image ends 1425. Otherwise, the image is determined to be suitable for analysis, and the entire process implemented by the grading engine 940 continues to rule 1215, which... Figure 17 As shown in the image.

[0137] Figure 17 This is a schematic diagram of process 2205 for identifying whether an object detected in fragment 925 is a spot indicating a reaction. If the object cannot be identified as a spot with the required certainty, the process continues to... Figure 18 The process 2305 determines whether fragment 925 is empty. Figure 17 The process 2205 determines that if an object in a given fragment 925 satisfies specific conditions related to the detectability, shape, size, and location of the spot in the fragment 925, then the object is a spot indicating a reaction.

[0138] The process includes receiving parameters 2215, such as the maximum and minimum spot radii, the maximum and minimum isoperimeter coefficients, and the maximum and minimum spot edge measurements. These parameters are unique for a single measurement site / ongoing measurement.

[0139] The process involves applying rule 2220 to determine whether the metrics found in the fragment that support the identification of objects in a given fragment indicate a suitable response for grading. If so, the graded response can be uniquely determined for the given fragment.

[0140] The application of rule 2220 utilizes circumference and domain metric 2210, for example, relative to the above. Figure 13The metrics generated during the described metric generation process. Application of rule 2220 may include, for example, determining whether a circumference was found in a given fragment 925. If a circumference was found for a fragment in the analysis, it is determined that a spot indicating a reaction exists in that fragment, and the radius and location of the circumference are determined.

[0141] Additional or alternative rules can be used to determine the presence of an indicator spot. For example, the application of rule 2220 can also include determining, based on SSR, whether an object in the form of a non-circular "spot" has been found. If so, various measures of the "spot" are determined, such as radius, edge measure, isopiece factor, and location. If all spot criteria are met, it is determined that an indicator spot is present in the fragment, wherein the spot criteria include, for example, one or more of the following: spot radius greater than or equal to the minimum spot radius and less than or equal to the maximum spot radius; spot isopiece factor greater than or equal to the minimum spot isopiece factor and less than or equal to the maximum spot isopiece factor; spot edge measure greater than or equal to the minimum spot edge measure and less than or equal to the maximum spot edge measure, and / or similar. If all spot criteria are met, it is also determined that a spot has been detected.

[0142] If a circle is identified in fragment 925, or if all the color spot criteria are met, then apply. Figure 19 The process 2420 shown checks whether the spots identified in process 2205 can be identified as bright object artifacts rather than spots indicating a reaction.

[0143] If no circle is found and no color spot criteria are met, then apply Figure 18 Process 2305 determines whether a fragment is empty.

[0144] If segment 925 is determined to neither include a circumference nor meet the color spot criteria, process 2305 determines whether segment 925 includes an object or is empty. It does this by determining 2310 whether segment 925 is empty (i.e., the average pixel value is below a certain level) or whether it is uniformly filled with gray levels that do not allow objects to be distinguished (i.e., no transitions or edges are detected) and are consistent with the gray levels of the entire array image. Determination 2310 includes comparing a specific metric of segment 925 (e.g., SSR value) with one or more thresholds set to parameter 2312 (e.g., segment SSR limit parameter). This check can be specific to a printed area or measurement within segment 925.

[0145] For example, if the SSR value of segment 925 is less than the segment SSR limit parameter, the average pixel value of segment 925 is less than the maximum average pixel value parameter, the standard deviation (or other diffusion value measure) of segment 925 is less than the maximum standard deviation parameter, and the detection indication value of the object with respect to background noise minus the measurement of background noise is less than the maximum threshold parameter, then segment 925 is determined to be empty 2315, and the process for segment 925 ends 2320. Otherwise, determination 2325 is a technical error, i.e., the domain includes the object, but it cannot be considered sufficiently definitively that the object qualifies as a spot indicating the response.

[0146] As mentioned above Figure 17 The process described in the diagram, if Figure 17 If the process identifies objects within a given segment 925 that can potentially be identified as spots indicating a reaction, then the process includes, as step 2420, removing bright object artifacts from each segment 925. Figure 19 More details are shown below.

[0147] like Figure 19 As shown, the bright object artifact removal step 2420 includes applying rule 2505 to determine whether an object in the segment is a bright object anomalous by comparing a pixel intensity metric with one or more thresholds set to parameter 2425 and / or derived from other metrics. As an example, rule 2505 could specify that an object identified in segment 925 is considered a bright object anomalous if the signal average is higher than a detection threshold calculated based on the IQ3 value or if the signal standard deviation is higher than a value defined by a unique parameter.

[0148] A threshold is defined using a detection gradient and a detection intercept, which are provided as unique parameters 2425 for a given segment 925, and the detection threshold is defined as the detection gradient of segment 925 multiplied by the IQ3 value of segment 925 plus the detection intercept of segment 925. For example, rule 2505 can specify that an object in a given segment 925 can be identified as a bright object anomalous if any of the following criteria apply:

[0149] (1) The average spot signal (e.g., pixel value) of the spot is greater than or equal to the detection threshold;

[0150] (2) The standard deviation of the speckle signal or other measured diffusion values ​​is greater than the relevant threshold provided as the only parameter 2425; or

[0151] (3) The average value of the circumferential signal is greater than or equal to the detection threshold;

[0152] (4) The standard deviation of the circumferential signal or other diffusion values ​​measured are greater than the relevant threshold provided as the only parameter 2425.

[0153] If none of these criteria are met, the process proceeds as step 2510. Figure 20 The process 2025 shown applies rules to classify the spots. If any of the above criteria apply, a bright object anomaly is detected, and a technical error 2515 is recorded for segment 925 (i.e., the detected object may be anomalous); and process 2420 is repeated for another segment 925 until all segments 925 in image 10 have been parsed.

[0154] It will be understood that bright object anomaly detection can use only one or some of the criteria mentioned above and / or can use different criteria. In alternative or additional steps, the process can remove any pixels with pixel intensities above the detection threshold, rather than rejecting the entire segment 925 on the grounds of technical error.

[0155] As mentioned above, regarding Figure 19 ,if Figure 19 If the process determines that the object is not a bright object artifact, then apply... Figure 20 The process shown in Figure 2025 applies rules to classify spots.

[0156] exist Figure 20 In the process 2025 shown, the spot grade / reactivity of each assay site 15 / spot / fragment 925 is based on the use of Figure 13 The measurement and determination and / or determination of site-specific parameters 15 are determined by the process shown. Spot grading / reactivity determination is performed only on spots / fragments 925 that are not rejected by any of the aforementioned rules applied by the grading engine 940.

[0157] The determination of spot grade / response level based on SSR is provided as a useful example. It will be understood that the determination of spot grade / response level may additionally or alternatively be based on other measures, such as, but not limited to, IQ1, IQ2, IQ3 and / or IQ4, etc.

[0158] In this example, four thresholds were set as assay-dependent parameters and optionally as spot-specific and / or sample-type-dependent parameters. The thresholds were increments of Tel, T1, Th, and Teh. Based on Table 1 below, the spot grade / reactivity level for a given spot / assay site 15 / fragment 925 was determined using SSR:

[0159] >Teh Technical error [Teh-Th] Reactive ]Th to Tl[ Uncertain [Tl to Tel] non-reactive <Tel Technical error

[0160] Table 1

[0161] Procedure 2025 includes applying rule 3105 to determine the spot grade / reactivity of each spot / assay site 15 / fragment 925, such as Figure 20 As shown in the figure. According to these rules 3105, spots will be classified as reactive, non-reactive, or indeterminate based on a metric reflecting their intensity (e.g., SSR or delta (delta is the difference between the average spot or circumferential pixel intensity value and the average background pixel intensity value)).

[0162] In the specific examples shown in Table 1, if a fragment's SSR value is below the minimum threshold Tel or above the maximum threshold, it is determined to be a technical error. If the SSR is above the minimum threshold Tel but below the upper non-reactive threshold T1, it is determined to be a non-reactive classification. If the SSR is above the upper non-reactive threshold T1 but below the lower non-reactive threshold Th, it is determined to be an uncertain state, indicating that it may not be possible to determine the classification as reactive or non-reactive with sufficient certainty. If the SSR is between the lower reactive threshold Th and the upper reactive threshold Teh, it is determined to be a reactive classification. Although the specific examples given in Table 1 use SSR as a metric, it will be understood that the same concept can be applied to other metrics.

[0163] Advantageously, rule 3105 may include logic for dynamically switching between metrics used to determine blob values / reactivity levels to select the most appropriate metric, such as switching based on the standard deviation of background pixel intensity or other measures of diffusion value. In this example, if SSR is found to be unsuitable, the logic uses SSR as the default and switches to another metric, such as delta. The switching between SSR and delta is based on a measurement of the diffusion value of the background value, for example, by using a background CV value, which is 100 multiplied by the standard deviation of the background and then divided by the average of the background.

[0164] For example, in some cases, the background around the spot is non-uniform, causing the background standard deviation to be much larger than the mean. As a result, the SSR is greatly reduced, which can lead to incorrect determination of non-reactivity, even when the spot clearly indicates a reaction. In this case (invalid SSR), an alternative spot reactivity assessment using delta is adopted, and CV is used instead of standard deviation to assess background uniformity, so as to take into account the average level of the background relative to the standard deviation.

[0165] The threshold parameters 3110 used to determine the spot grade / reactivity level are all determination-dependent and optionally also dependent on the spot location and / or sample type; that is, they are different for different determination sites 15 in the same array 10. The required threshold parameters 3110 include the maximum acceptable background CV value, the reactive SSR value range, the non-reactive SSR value range, the reactive spot delta value range, and / or the non-reactive delta value range, etc.

[0166] According to the above... Figure 17 The process described in (Rule 2) can determine spots by segmentation or by circumferential detection. The applied threshold parameter 3110 can also depend on whether the spot is determined to be circumferential through circumferential search or by segmentation process, with different threshold parameters applied for each.

[0167] The application of rule 3105 begins by determining whether the background CV value is less than the maximum acceptable background CV value. If so, the process continues using the SSR metric. If not, the process continues using the delta metric, thereby providing the dynamic switching described above.

[0168] If the SSR measurement is used to proceed, it is determined whether the SSR (whether for a spot or a circumference, depending on how the spot was determined) is within the reactive SSR value range (this is a threshold parameter), and if so, the spot / assay site 15 / fragment 925 is determined to indicate that the associated assay is reactive 3115. If the SSR is not within the reactive SSR value range, it is determined whether the SSR is within the non-reactive SSR value range (this is also a threshold parameter), and if so, the spot / assay site 15 / fragment 925 is determined to indicate that it is non-reactive or unreactive 3120. If the SSR of the spot is determined to be outside both the reactive and non-reactive SSR value ranges, the spot / assay site 15 / fragment 925 is determined to be indeterminate or a technical error has occurred 3125.

[0169] If it is determined that a delta metric should be used, a similar determination process as described above is used, but instead of using SSR and the threshold parameter 3110 associated with SSR, delta and the threshold parameter 3110 associated with SSR are used instead of delta and the threshold parameter 3110 associated with SSR.

[0170] Regardless of Figure 20 The process shown in 2025 will determine how things are handled. Figure 21 The inspection step 3205 is shown. Inspection step 3205 includes applying rule 3215 to the inspection based on... Figure 17 The determination of spot grade / reactivity is made by the spot qualification step 2025 shown. Figure 21 The inspection step 3205 shown is an optional step and can be turned on or off by setting an appropriate inspection parameter value 3210. If Figure 20 The spot qualification step 2025 determined that the spot / test site 15 / fragment 925 should be classified as "non-reactive" 3120, and the same spot / test site 15 / fragment 925 was... Figure 18 If the process in Figure 26 is indicated as "empty" 2315, then rule 3215 identifies spot / assay site 15 / fragment 925 as non-reactive 3220. However, if spot qualification step 2025 determines that spot / assay site 15 / fragment 925 should be classified as "non-reactive" 3120, and the same spot / assay site 15 / fragment 925 was not indicated as "empty" 2315 through the process in Figure 26, then "technical error" 3225 is output for spot / assay site 15 / fragment 925 instead of "non-reactive". If spot qualification step 2025 determines that spot / assay site 15 / fragment 925 should be classified as "reactive" 3115, this is maintained by rule 3215, and then a "reactive" determination 3230 is output.

[0171] Advantageously, at least in the above respects Figure 20 The threshold parameters used in the described spot qualification process 2025 are individually variable, depending on one or more of the following factors: the ongoing assay, the specific location of assay site 15 on assay 10, and / or the type of sample being analyzed. In this way, the validity of assay determination can be improved by selecting a better threshold for any given assay, assay site 15, and / or sample type, regardless of which other (different) assays are performed on the same array 10. That is, to perform different assays at different assay sites 15 on the same array 10, a slight compromise needs to be considered in setting the threshold parameters.

[0172] Performing various steps, either individually or in combination, before spot qualification can improve the accuracy or efficiency of determining the test results, though it is not always necessary.

[0173] Although specific examples have been described above, it will be understood that this is provided to give those skilled in the art a possible way to put the invention into practice, but modifications to the methods and apparatus described above are possible within the scope of the claims.

[0174] For example, various metrics such as SSR, delta, and IQ3 were used, but it will be understood that alternative metrics can be used. Furthermore, although various techniques for detecting spots in determination site 15, such as edge detection, thresholding, and shape (e.g., circumferential shape) recognition, have been described above, it will be understood that alternative techniques can be used.

[0175] The digital camera 17 is advantageously used to collect measured images, but it will be understood that other sensor devices that can be used to determine the measure of the degree of response can be used, such as thermal imagers and ultraviolet or infrared sensors.

[0176] The method steps of this invention can be executed by one or more programmable processors, which implement a computer program to perform the functions of the invention by manipulating input data and generating outputs. The steps of the method can also be performed by special-purpose logic circuitry, such as an FPGA (Field-Programmable Gate Array) or ASIC (Application-Specific Integrated Circuit) or other custom circuitry. Processors suitable for implementing computer programs include CPUs and microprocessors, as well as any one or more processors. Typically, the processor receives instructions and data from read-only memory or random access memory, or both. The basic elements of a computer are a processor for executing instructions and one or more memory devices for storing instructions and data. Typically, a computer will also include one or more mass storage devices (e.g., magnetic disks, magneto-optical disks, or optical disks) for storing data, or be operatively coupled to receive data from or transfer data to said one or more mass storage devices, or both. Information carriers suitable for embodying computer program instructions and data include all forms of non-volatile memory, including, as examples, semiconductor memory devices such as EPROMs, EEPROMs, solid-state memories (e.g., SSDs and flash memory devices); magnetic disks, such as internal hard disks or removable disks; magneto-optical disks; and CD-ROMs and DVD-ROMs. Processors and memory can be supplemented by dedicated logic circuits or incorporated into dedicated logic circuits.

[0177] The steps of the method can be performed by a local processing device on a remote processing device, or some steps can be performed on a local processing device while others can be performed on a remote device in a distributed processing setup.

[0178] To provide interaction with the user, the present invention can be implemented on a device having a screen for displaying information to the user (e.g., a CRT (cathode ray tube), plasma, LED (light-emitting diode), or LCD (liquid crystal display) monitor) and an input device through which the user can provide input to the computer (e.g., a keyboard, touchscreen, mouse, and trackball). Other types of devices can be used; for example, the feedback provided to the user can be any form of sensory feedback, such as visual, auditory, or tactile feedback; and input from the user can be received in any form, including acoustic, voice, or tactile input.

[0179] Suitable processing devices for performing the methods described above may include mobile, fixed, or network-enabled devices that include or are configured to implement a controller or processing system. The device may be, include, or be incorporated into a mobile phone, smartphone, PDA, tablet computer, laptop computer, etc. The controller or processing system may be implemented by a suitable program or application (app) running on the device. The device may include at least one processor, such as a central processing unit (CPU), a mathematics coprocessor (MCP), and / or a graphics processing unit (GPU). The at least one processor may be a single-core or multi-core processor. The device may include memory and / or other data storage, which may be implemented in DRAM (Dynamic Random Access Memory), SSD (Solid State Drive), HDD (Hard Disk Drive), or other suitable magnetic, optical, and / or electronic storage devices. At least one processor and / or memory and / or data storage may be locally located, for example, provided in a single device or in multiple devices communicating in a single location, or may be distributed across several local and / or remote devices. The device may include a communication module, such as a wireless and / or wired communication module. The communication module can be configured to communicate via cellular communication networks, Wi-Fi, Bluetooth, ZigBee, Near Field Communication (NFC), IR, satellite communication, and / or other internet-enabled networks. The communication module can be configured to communicate via Ethernet or other wired networks or connections, via telecommunications networks such as POTS, PSTN, DSL, ADSL, optical carrier lines and / or ISDN links or networks and / or the like, via the cloud, and / or via the internet or other suitable data bearer networks. The communication module can be configured to communicate via optical communication such as optical wireless communication (OWC), optical free-space communication, or Li-Fi, or via fiber optics and / or the like. The device and / or controller, or at least one processor or processing unit, can be configured to communicate with a remote server or data warehouse via the communication module. The controller or processing unit may include or utilize at least one processor, memory, and / or other data storage and / or the communication module of the device.

Claims

1. A computer-implemented method for analyzing measurements performed at corresponding measurement sites in an array comprising multiple measurement sites; The plurality of assay sites are arranged in multiple rows and columns, and each assay site includes a trap or spot that forms a reaction site; the method includes: Receive at least one image, which collectively or individually images the plurality of measurement sites of the array, and process the at least one image to determine at least one measure representing the degree of response at the measurement site; For each of the measurement sites, one or more parameters are identified for that measurement site, wherein the one or more parameters are predetermined for a given measurement type, sample type, and location of the measurement site, and the one or more parameters for at least one measurement site in the array are different from the one or more parameters for at least one other measurement site in the array; and For each of the assay sites, the magnitude of the response at that assay site is determined based on at least one metric used for that assay site and one or more parameters used for that assay site; wherein: At least one of the parameters includes one or more criteria, thresholds, or ranges; Each criterion, threshold, or range indicates a different degree of responsiveness; and Determining the magnitude of the response at each of the measurement sites includes determining whether at least one metric used for the measurement site is higher or lower than one or more thresholds used for the measurement site, or within or outside one or more ranges used for the measurement site, or meets or does not meet at least one of the criteria used for the measurement site, in order to determine the degree of the response.

2. The method according to claim 1, wherein, The at least one metric for a site includes or represents: the pixel intensity of a region of the image representing the spot at the site, the spot being formed by the reaction of the analyte and having properties indicating the degree of the reaction.

3. The method according to claim 2, wherein, The at least one metric used for a site includes a signal-to-standard deviation ratio, or delta, where delta is the difference between the average pixel intensity value of the spot and the average background pixel intensity value.

4. The method of claim 1, comprising at least one of the following: Cropping the image around the said site; and / or The image is meshed into fragments, wherein, Each fragment will close the spot and / or the assay site.

5. The method according to claim 4, wherein, At least one of the cropping, gridding, and / or determination of the one or more measurements of the measurement site uses a common parameter, which is shared with one or more or each or all other spots or sites having the same measurement type.

6. The method according to claim 1, wherein, The parameter used for measurement is unique to the site, and the measurement is used to determine the magnitude of the response at the measurement site.

7. The method according to claim 2, wherein, The method includes detecting or identifying a portion of the image corresponding to the spots in the reaction sites, including one or both of the following: Detect or identify shapes in the image that include circumferences with diameters within predetermined intervals; and / or Edge detection used to determine the edges of the spots in the at least one image.

8. The method of claim 7, comprising: Filter out identified spots with pixel intensity measurements below a threshold, and / or spots not within a predetermined geometric region; the predetermined geometric region corresponds to the location of the measurement site and is located based on at least one other identified spot, control spot, or reference point of the array.

9. The method according to claim 1, wherein, The parameters used to determine the magnitude of the response at the measurement site include at least one of parameters received from a user input device and / or parameters retrieved from a data warehouse corresponding to one or more identifiers associated with the array, or with one or more or each of the measurement sites, or with the measurement performed here.

10. The method according to claim 9, wherein, The identifier is obtained by an input device, which includes at least one of the following user input devices for receiving user input: a barcode reader, a QR code reader or other machine-readable code reader, an RFID tag reader and / or an infrared signal reader.

11. The method according to claim 1, wherein, Determining the magnitude of the response at the assay site includes performing one or more logical tests on the at least one metric used for the assay site and one or more parameters used for the assay site, wherein the result of the one or more logical tests is the magnitude of the response at the assay site, the magnitude of the response including an indication of at least one of the following: Whether a reaction occurred at the measured site; The extent of the reaction at the measured site; Whether there is activity at the measured site; The degree of activity at the measured site; and / or To determine whether analyte is present at the measurement site.

12. The method of claim 1, wherein the array is a microarray.

13. A processing system comprising at least one processing device, a data storage device, and a communication system, the communication system being configured to receive an image and output an indication of the magnitude of a response at a measurement site; the processing device being configured to perform the method of any of the preceding claims.

14. An analysis system for analyzing one or more measurements performed at corresponding measurement sites of a plurality of measurement sites in an array; The measurement sites are arranged in rows and columns, and each measurement site includes a trap or spot; the analysis system includes: The processing system according to claim 13; At least one imaging device is configured or configurable to collect at least one image of the array and transmit the at least one image to the processing system; and An output device is configured to receive an indication of the magnitude of the reaction at the measurement site from the processing system, and to output the indication of the magnitude of the reaction at the measurement site.

15. The analysis system of claim 14, wherein the array is a microarray.

16. A computer program product comprising instructions that, when executed on a processing system, cause the processing system to perform the method according to any one of claims 1 to 12.

Images