Application development environment for biological sample assessment processing

Abstract

A system and method for developing applications (Apps) for automated assessment and analysis of processed biological samples. Such samples are obtained, combined with nutrient media and incubated. The incubated samples are imaged and the image information is classified according to predetermined criteria. The classified image information is then evaluated according to Apps derived from classified historical image information in a data base. The classified historical image information is compared with the classified image information to provide guidance on further processing of the biological sample through Apps tailored to process provide sample process guidance tailored to the classifications assigned to the image information.

Claims

1. A method for processing a biological sample comprising: obtaining a biological sample; combining the biological sample with nutrient media; incubating the biological sample; obtaining a digital image of the incubated biological sample; classifying the digital image according to analytical criteria selected from the group consisting of specimen origin information, clinical sample criteria, process materials and process conditions; obtaining data from historical digital images of incubated biological samples on nutrient media that share at least one of the analytical criteria assigned to the digital image; and using the historical digital image data to output an instruction to a user for further processing of the biological sample, wherein the further processing is selected from the group consisting of identification testing, antibiotic susceptibility testing or both.

2. The method of claim 1, wherein the analytical criteria comprises specimen origin information, and wherein the specimen origin information comprises geographic information regarding the biological sample source and the type of biological sample.

3. The method of claim 1, wherein the analytical criteria comprises process materials, and wherein the process materials comprise type of nutrient media.

4. The method of claim 1, wherein the historical digital image data is classified by at least one of specimen type, organism taxa or culture media type.

5. The method of claim 1, further comprising, after obtaining the digital image of the incubated biological sample, analyzing the digital image data and determining, from the analyzed data if the digital image reflects microbial growth.

6. The method of claim 5, wherein, in response to a determining that the digital image does not show microbial growth, outputting an indication of no microbial growth.

7. The method of claim 5, wherein, in response to determining that there is an indication of microbial growth, determining if the specimen is a sterile specimen and identifying one or more coordinates of a colony of microorganisms in the digital image that was a basis for the indication of microbial growth.

8. The method of claim 7, wherein, in response to determining that the specimen is sterile, indicating that the specimen is a high value positive and sending instructions to further process the specimen.

9. The method of claim 8, further comprising communicating the coordinates of a colony to a module that communicates the coordinates to a picking apparatus that will pick the colony from the biological sample.

10. The method of claim 9, further comprising transferring the biological sample to the picking apparatus and picking the colony from the biological sample wherein the steps of transferring and picking are controlled by the module.

11. The method of claim 7, wherein, in response to determining that the specimen is not sterile, comparing the historical digital image data with the digital image data to identify a specific predetermined species of microorganism in the digital image of the incubated biological sample.

12. The method of claim 11, wherein the comparing is performed by a module, and, if the module determines that the specific predetermined species of microorganism is present in the image data, the module reports the identification of the specific predetermined species.

13. The method of claim 12, further comprising flagging the specimen for further review.

14. The method of claim 7, further comprising comparing the historical digital image data with the digital image of the incubated biological sample and determining an amount of microbial growth wherein the comparing and determining steps are performed in a module in communication with an imaging apparatus that obtained the digital image of the incubated biological sample.

15. The method of claim 7, further comprising determining if the microbial growth on a plate is one of pure colonies, predominant colonies, or complex colonies by communicating the digital image of the incubated biological sample to a module that determines a growth level as a vector of three probabilities.

16. The method of claim 15, wherein, in response to a determination that the colony is pure, the method further comprises: reporting, from the module, that the plate is pure; and determining, by the module, if the growth exceeds a predetermined threshold growth.

17. The method of claim 16, wherein, if the module determines that the growth exceeds the predetermined threshold growth, the method further comprises: identifying the sample as a high value positive; alerting a user of the high value positive; identifying the coordinates of the high value positive; communicating the coordinates of a colony to a module that communicates the coordinates to a picking apparatus that will pick the colony from the biological sample; and transferring the biological sample to the picking apparatus and picking the colony from the biological sample wherein the steps of transferring and picking are controlled by the module.

18. The method of claim 17, wherein, if the module determines that the growth does not exceed the predetermined threshold the module provides a presumptive identification of the colony based on comparing an image of the colony from them digital image of the incubated biological sample provided to the module with the historical digital image data accessed by them module, the module performing the further step of: reporting the presumptive ID to a user.

19. The method of claim 15, wherein, in response to a determination that the colony is predominant, the module provides a presumptive identification of the colony based on comparing an image of the colony from them digital image of the incubated biological sample provided to the module with the historical digital image data accessed by them module, the module performing the further step of: reporting the presumptive ID to a user.

20. The method of claim 19, the method further comprising: determining, by the module, if the growth exceeds a predetermined threshold growth.

21. The method of claim 20, wherein, if the module determines that the growth exceeds the predetermined threshold growth, the method further comprises: identifying the sample as a high value positive; alerting a user of the high value positive; identifying the coordinates of the high value positive; communicating the coordinates of a colony to a module that communicates the coordinates to a picking apparatus that will pick the colony from the biological sample; transferring the biological sample to the picking apparatus and picking the colony from the biological sample wherein the steps of transferring and picking are controlled by the module; and alerting a user that further review is required.

22. The method of claim 20, wherein, if the module determines that the growth does not exceed the predetermined threshold the module performs the further step of: reporting the presumptive ID to a user.

23. The method of claim 15, wherein, in response to a determination that the colony is complex, the method further comprises: reporting, from the module, that the plate is complex; and determining, by the module, if the growth exceeds a predetermined threshold growth.

24. The method of claim 23, wherein, if the module determines that the growth exceeds the predetermined threshold growth, the method further comprises: alerting a user that further review is required.

25. The method of claim 17, wherein, if the module determines that the growth does not exceed the predetermined threshold the module performs the further step of: reporting to a user that the complex sample does not meet or exceed the predetermined threshold growth.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a block diagram illustrating an example system environment for biological image processing application development.

(2) FIG. 2 is an example flow diagram of an implementation with example Apps and their processes in a development environment according to an aspect of the disclosure. FIG. 2A is the portion of FIG. 2 that illustrates the growth and key ID Evaluation Modules. FIG. 2B is the portion of FIG. 2 that illustrates the quadrant quantitation and semi-quantitation modules.

(3) FIG. 3 is a table illustrating an example of target organisms with presumptive identifications and the associated media.

(4) FIG. 4 is table illustrating example targeted media with the associated specimen types.

(5) FIG. 5 is a schematic of processes that may be implemented in some versions of the present technology, such as in a processing system 101 of in FIG. 1, where the processing system may access data and/or algorithms of a Data Lake as discussed herein to develop Apps with a training process, validation process and/or clinical submission process. These processes may in turn employ information (e.g., updated data and algorithms) also derived from the Data Lake.

DETAILED DESCRIPTION

(6) The present disclosure provides apparatus and methods of an environment for developing imaging applications for identifying and analyzing biological specimens such as microbial growth. Many of the methods described herein can be fully or partially automated, such as being integrated as part of a fully or partially automated laboratory workflow.

(7) This document provides a description of the design and implementation of a system that accelerates delivery of automated imaging capabilities to biological imaging systems such as the BD Kiestra System. The imaging capabilities of such systems are enabled by a suite of hardware, software, analytical algorithms, and clinical rules. An example of one such commercialized system includes one or more digital cameras (e.g., 4 MP) with multiple illumination configurations that generate an optimized and standardized image based on an appropriate platform. The systems described herein are capable of being implemented in other optical systems for imaging microbiology samples. There are many such commercially available systems, which are not described in detail herein. One example may be the BD Kiestra ReadA Compact intelligent incubation and imaging system. The Kiestra ReadA compact is an automated incubator with an integrated camera and plate transport system that enables automated imaging of plates. The ReadA compact is commercially available. The ReadA compact also has integrated plate import and plate export devices that couple the incubator to other instruments for manipulation. Therefore, in some embodiments, in response to the analysis of the digital images by one or more of the Apps, the Apps can issue instructions and control the incubation of the relevant specimen under evaluation by the App. Other example systems include those described in PCT Publication No. WO2015/114121 and U.S. Patent Publication 2015/0299639, the entirety of which is incorporated by reference herein. Such optical imaging platforms are well known to those skilled in the art and not described in detail herein.

(8) A series of Apps for the system can provide analysis of the images from most specimens, generated at various predetermined times, such as in the ReadA Compact. The system can enable downstream actioning of these plates-including automated release of no-growth and/or negative plates, and automated characterization of colonies for definitive ID and AST analysis.

(9) At a high level, the imaging analysis tools (Imaging Apps) can be deployed to enable a collection of different results that provide substantial value to the lab and/or actionable clinical results. These Apps utilize one or more image analysis algorithms (Modules), as well as a set of rules that provide information on how to apply the Modules on specific media types and/or specific specimens. Certain Apps can also have associated Expert Systems, which overlay an additional set of rules on more basic App determinations, typically regulatory/clinical guidance, that inform recommendations for actioning and interpreting the result.

(10) Developing Imaging Apps can utilize an iterative approach. A data lake is developed using either specimen information acquisition or images of clinical specimens being processed as normal practice in the clinical lab, or both. Algorithms are developed to model conclusions/instructions/outputs such as those illustrated in FIG. 2 (FIG. 2A and FIG. 2B) that can be drawn from the truths and classification information associated with a specimen image evaluated by the App. Validation and verification (V&V) of the App is performed by using a collection of predefined images with certain metadata requirements from the data lake that are marked for V&V analysis. The Apps are then subjected to a clinical study (which includes processing specimens to obtain images of those samples and evaluating the images of the specimens using the App) to both verify the App and train the App. The clinical studies do not require that a specific clinical site is used for the clinical study

(11) However, a biological image processing application development system 100, such as one illustrated in FIG. 1, may be implemented to leverage an exponential strategy with several components. The system 100 includes one or more of: a) a set of toolbox applications, such as for processing system 101 with one or more processors, having generic algorithms and Modules that enable the App and can be rapidly matured or optimized (for a type of specimen, for example) by applying Artificial Intelligence algorithms such as neural networks, artificial neural networks, and other deep learning algorithms. Artificial intelligence may be implemented to automatically determine what attributes of an image, and with what algorithms, to best provide the desired Module capability. b) critical to the development and deployment of the Apps described herein is a developed database 102, denominated a Data Lake herein, that comprises images of clinical specimens with linked manual/standard analysis of truth (i.e. facts regarding the mage such as colony quantitation, IDs, result interpretation, etc.) and image conditions, and in some cases patient demographic data (suitably disidentified). The images carry classification information in the form of metadata so that only relevant image data is used to develop a particular App. The Data Lake may be populated by clinical lab collaboration such as with imaging systems 104 that may optionally supply data to the database via a network.

(12) Where the Data Lake is stored is largely a matter of design choice. The Data Lake can be stored locally or in a cloud. The data in the Data Lake can be partitioned. For example, the data in the Data Lake could be segmented depending on how the data is accessed and/or used. In one embodiment, one segment of the data in the Data Lake could be for algorithm training, another segment could be used for data verification or validation and yet another could be used for clinical submission.

(13) The database of the system can store and provide classified data of these clinical images and also linked data, that can be pulled individually as appropriate for algorithm development, formal verification and validation, or clinical submission on demand. The App can be established with a level of global standardization including lab protocols, media types, imaging time, quantitation scoring, etc. Resources can supplement the generation of images and reference data (specimens and/or spiked/contrived samples) for specimens/organisms/conditions that are infrequently encountered in the clinical setting. This proactive strategy for data generation allows an almost on-demand prioritization and development of particular Apps for specific specimen or media types. The system architecture permits integration of the new modules into existing laboratory site software to ease release of new imaging diagnostic functionalities (Apps/modules/packages). This may be achieved by standardizing the interface between new Apps/modules/packages with existing/previous imaging software systems. In this regard, the Apps/modules/packages may be add-ons to system software such as a system software for an automated imaging system (e.g., in a processing system that controls any one or more of a plate/sample conveyor, incubator, camera, picking machine and/or related robotics, for moving such samples/plates within such automated laboratory cell/equipment etc.). Certain specimens determined to be negative for pathogens can be spiked with known pathogens and the subsequent image of the incubated, spiked specimen characterized as described herein. The images of the spiked specimens can then be used as a training set for Apps that can be used to evaluate and process less frequently occurring pathogens.

(14) This approach can provide maximal flexibility in prioritization and the cadence of App development. It would also make it possible to provide early App performance metrics to help better understand the added value of an App, the synergies at the solution level, and enable earlier recognition. The Data Lake may be generated such as with one or more hospital systems that meet a list of Corporate Clinical Development (CCD) criteria (i.e. technical and medical information). Implementation provides the ability to store and classify the data, query, and audit the database; lab protocols that define the program and processes; training, monitoring, compliance, quality metrics etc. as is typical for a clinical trial-and may be done as part of the development of the Data Lake, rather than at the end of a typical product development process. The approach may implement dedicated clinical lab resources to determine certain plate results outside standard protocols, and to link images with analysis results to the Data Lake. Certain specimens, plate types or image acquisition time points may need to be run specifically to develop the Data Lake in special processes that may be independent of normal/typical lab practice. Thus, in some versions, the Data Lake database may comprise images of clinical specimens with linked manual/standard analysis of truth (e.g., quantitation, IDs, result interpretation) and metadata associated with classifications for the image (e.g., select patient demographics, imaging time and conditions, media types, etc.).

(15) Establishment of both the algorithms and Data Lake may include a certain level of standardization. This will also define what any given App is validated for, and the analysis/instruction/output that may ultimately be obtained. Given the diversity of media types, vendors of media, and incubation times in use across labs globally, a Best Practices approach may be used to initiate this effort. Additional conditions may be added in the future by stocking the Data Lake with appropriate data. An example of a Media x Specimen matrix is summarized in FIG. 3. The table in FIG. 3 includes 12 media types. Note that Blood Agar, tryptic soy broth (TSA) and Columbia are grouped together as one media type. XLD is Xylose Lysine Deoxycholate Agar, SS agar is Salmonella, Shigella agar, CNA is Columbia Naladixic Acid Agar (CNA) and CLED is Cystine-Lactose-Electrolyte-Deficient agar. The listed media are well known to those skilled in the art as are the microorganisms known to be identifiable on the listed media. Thus, the standardization may include lab protocols, media types, imaging time, streak patterns, quantitation scoring, etc. that are classifications applied to the data for the database development and image analysis with reference to the data base/Data Lake/historical image information. This focuses validation efforts and minimizes time for development and allows lab to lab metrics, data sharing, etc. The use of the Data Lake as training data, validation data, clinical submission data, etc. for developing Apps is illustrated in FIG. 5.

(16) To ensure database accuracy, population of the data into the database may involve independent human image analysis of images performed by several individuals, or plate analysis done manually by human technologists. Human readings may be compared with clinical laboratory reports. In some cases, a further image review may be involved for discrepant readings. Database input may involve de-identification of patient information from image related data. Review of images for data entry may involve human scoring of the growth on plates for pure, predominant, complex and no growth; quadrant quantitation.

(17) Example Software Modules for Toolbox

(18) In a typical example, the Data Lake may contain media plate images, linked to lab-determined quantitation (e.g., no growth, +, ++, +++). It may also contain identification (ID) of organisms determined to be of significance for the kind or type of specimen (such as ones that may be significant a trained clinical microbiologist). The Data Lake may also contain image-based metadata, and in some cases, patient demographic information. Organisms not routinely identified as pathogens (i.e., normal flora) may also be requested to be identified to facilitate algorithm development. Once populated, a portion of images will be leveraged to train and test appropriate algorithms toward App development.

(19) At a high level, the imaging analysis tools can be deployed to enable a collection of different results that provide substantial value to the lab and/or actionable clinical results. These Modules (Mods) are comprised of one or more image analysis algorithms, as well as a set of rules that provide information on how to apply the Modules on specific media types and specific specimens. Some Modules, such as Screening-MRSA, may be implemented as an App. Other modules may more often be packaged with other Modules to provide a synergistic capability (for example, quadrant quantitation and purity will often be packaged as an App.) Some Apps can also have associated Expert Systems, which overlay an additional set of rules, typically regulatory/clinical guidance, that inform recommendations for actioning and interpreting the result (e.g. KB Zone described herein). Based on technical and clinical considerations, one or more Apps may also be bundled as components of a Launch Package depending on different clinical lab needs. It is also anticipated that some Apps will have versions (e.g., UCA V1.8 with FDA approval will become UCA 2.0).

(20) Some examples of the algorithms and Modules (collections of synergistic algorithms) are generic (generally working across specimens, pathogens, media) and are summarized below and may be considered in relation to the processes of the illustration of FIG. 2 (FIG. 2A and FIG. 2B). As previously mentioned, the classification of functionality of the various applications helps to provide development of detection applications for various species and media.

(21) 1. Growth App/Module 1010A

(22) Growth App 1010 (See FIG. 2A) may be directed to answering a simple question: is there anything growing that can be detected on this plate at this particular incubation time? The answer to this question will be a growth probability ranging from 0 to 1. Growth can be a module targeting any media, independent from dispense volume or streaking pattern. Growth may be detected as early as possible from pre-set imaging points. Rules specify whether to issue an alert based on specimen type and/or media. In some versions, gram stain results could also be integrated with App/Module where appropriate. In some cases, this may be implemented an Early Detection or Early Growth App/Module. Growth may be detected as early as possible (e.g., a detection window of 4 to 14 hours or more) from pre-set imaging points. Rules specify whether to issue an alert based on specimen type and/or media. In some versions, gram stain results could also be integrated here where appropriate.

(23) 2. Key ID App/Module 1020

(24) A Key Id Module 1020 (See, FIG. 2A) may be aimed at identifying a species potentially growing on a given media. For each and every requested Key ID organism, the module may provide a list (with probabilities) of colony locations per Key ID ordered by decreasing probability. These colonies can then be picked manually or by an automated picking system.

(25) In some versions, the system may include a Screening and Critical Pathogens module(s) developed on a pathogen basis. These modules may provide detection of specific pathogens, groups of pathogens or those with specific properties. The Screening Apps may be implemented to enable identification of specific pathogens on CHROMagar, and can be used for patient management as well as pathogen characterization e.g., MRSA, ESBL, CPE, VRE etc. CHROMagar may enable a collection of pathogens to be presumptively identified i.e., CHROMagar orientation for both Gram Negative (GN) and Gram Positive (GP) bacteria. Pathogen specific media may be used for specimenssuch as SS media for Salmonella, Shigella from stool. Certain organisms may be presumptively identified or flagged on more generic media based on, for example, Hemolytic properties on blood agar, or unique morphologic properties on specific media. A potential collection of modules with pathogen x media capabilities is summarized in the Table of FIG. 4.

(26) 3. Quadrant Quantitation App/Module 1030

(27) Based on a streaking pattern, e.g., a BD Kiestra InoqulA quadrant streaking pattern, this Quadrant Quantitation module 1030 (See, FIG. 2B), in case of detected growth, will provide a growth level being light, moderate or heavy. BD Kiestra InoqulA automates the processing of both liquid and non-liquid bacteriology specimens to help streamline workflow, enable standardized processes and ensure consistent and high-quality streaking for inoculation of solid growth media. The growth level will be returned as a vector of three probabilities (light, moderate or heavy) ranging on [0,1] and summing up to 1. For example, the Module may evaluate all plates to assess whether there is no growth or different amounts of growth (e.g., +, ++, +++) and whether any growth is pure, predominate or complex. A no Growth determination can result optionally in auto release, or batch release. In some cases, growth quantification (e.g., +, ++, +++) may be determined by three or more colonies in any particular quadrant.

(28) With respect to growth type, the images/plates may be characterized as pure, predominant and complex. This may be based on a minimal number of isolated colonies of each type. For example, predominant growth may be greater than (or equal two) two colony types, where one type is greater than a factor (e.g., 10 times) the other(s). Complex may be greater than two colony types with no predominate isolate or if the isolate is not identifiable within a presumptive ID table such as the example of FIG. 3. Complex plates may be automatically flagged/passed to manual interpretation. Pure and predominant plate types could be automatically passed for further automated processing such as for having representatives of each colony type indicated, with rules that drive further workup (e.g., picking, ID and/or AST).

(29) For example, a colony forming unit (CFU)/mL Quantitation App/Module 1040 (See, FIG. 2B) may be implemented. Based on InoqulA streaking pattern #4 (MonoPlates) or #6 (BiPlates), this module may provide a growth level being in <1, 1 to 9, 10 to 99, 100 to 999, 1000 CFU/media on plate. The growth level may be returned as a vector of 5 probabilities ranging on [0,1] and summing up to 1. To get equivalent CFU/ml buckets units, the dispense volume may need to be considered.

(30) In general, the Quantification module may, in some versions, determine if growth is due to a single growing organism, a predominant organism or a mixture of (multiple) organisms. A pure organism may be deemed to be an organism responsible for 99% of the observable/imageable growth. A predominant organism may be an organism responsible for (90%, 99%) of the observable/imageable growth. A purity level may be returned as a vector of probabilities (e.g., 3 probabilities as pure, predominant, complex) ranging on [0,1] and summing up to 1. In the case of pure or predominant growth up to five colony locations for the main organism will be given in decreasing probabilities.

(31) Examples of responses to determined quantification are as follows: 1) An image of a specimen on a plated media is determined to have greater than a predetermined threshold (100,000 CFU/mL) of mixed flora. The response of the App to this determination is to flag this plate as a complex plate, because the plate has over the threshold amount of mixed flora and recommend manual review of the plate. Such a determination is not made in the context of media or taxa so this is an App of broad applicability and not limited in deployment to specific media or taxa on the plate. 2) An image is evaluated and determined to exhibit no growth for 24 hrs. If the specimen is classified as a critical specimen the App issues a preliminary report of no growth and either recommends or controls the re-incubation of the plate for another 24 hours. If a subsequent image detects no growth in 48 hours, the App sends a final report to the user of no growth after 48 hours. The App either recommends or controls discarding the plate 3) An image of a specimen on a plated media is determined to have greater than a predetermined threshold (100,000 CFU/mL) of a pure growth and a colony size that exceeds 0.5 mm. The response of the App is to issue an instruction or control the pick of colony for ID and AST testing. The App will flag the specimen for review by the technician and will send the greater than 100,000 CFU/mL report to the physician associated with the specimen. 4) An image of a specimen is determined to reveal the presence of MRSA by the App. The App sends a report that MRSA is detected and adds the specimen to a positive MRSA work list. If the App determines that the size of the MRSA colony exceeds a threshold (e.g. greater than 0.5 mm) the App will issue an instruction or control sending the sample for ID and AST testing. As noted elsewhere herein, ID and AST have their own specimen work up and evaluation. As such, ID and AST systems and apparatus are typically downstream of the incubation/imaging apparatus (e.g. Kiestra ReadA compact). 5) An image of a specimen classified as ESBL is determined to show no growth. The App will issue a final report that no ESBL isolate was detected and will issue an instruction to discard or control discard of the plate. 6) An image of specimen classified as sputum is identified as having mixed flora (therefore a complex plate) that exceeds the threshold amount. When the App determines that the plate is complex it issues an instruction for a technician to review the plate. Note that different thresholds for mixed flora that trigger the requirement for manual review might be deployed depending on specimen classification. 7) An image of a specimen classified as critical on blood agar is determined to reveal growth. In such an instance, the critical specimen App would fire and send an alert to the physician associated with the specimen and cause the specimen to be added to the critical sample work list. If the App determines that a colony greater than a threshold size (i.e. greater than 0.5 mm) and specimen is classified as being disposed on MacConkey agar, then the App causes the specimen to be sent to auto pick for MALDI and for Gram negative AST. The App will also send a report indicating that a Gram-negative specimen has been isolated. 8) An image of a specimen reveals a colony number that is greater than 100,000 CFU/mL and classifies the image as pure and having a colony size greater than a predetermined threshold (e.g. greater than 0.5 mm). In response, the App causes the specimen to be auto picked for AST, causes the specimen to be added to the positive review list by a technician and causes a report to be sent to the physician associated with the sample that indicates that more than 100,000 CFU/mL of a colony was detected from the sample. Further, if the AST results reveal that the picked colony is resistant to Carbapenem, then the App causes a molecular confirmatory test to be performed. 9) An image of a specimen is determined to possess heavy predominant growth. In response, the App causes the growth to be auto picked and a suspension prepared for performing MALDI on the picked sample. If the MALDI identifies the colony as E. coli, then the App causes the sample to be further evaluated for Gram Negative AST (using either the MALDI suspension or a new pick of the colony).

(32) In some versions, growth detection may be implemented as two modules where one evaluates plates to determine growth/no growth and another module evaluates growth quantity (+, ++, +++) in 3 or more colonies in any particular quadrant. For example, a first module evaluates an image of a critical, normally sterile specimen. If that evaluation reveals growth at a predetermined time point, and determines that a colony size is greater that a predetermined threshold size, then the App identifies coordinates of representative colonies and issues instructions to the imaging apparatus (e.g. ReadA) that the plate is to be moved to an apparatus that will auto pick the identified colony. The picked colony is resuspended in a solution to a predetermined density for further testing in, e.g., a molecular diagnostic apparatus or test (e.g., PCR, Sequencing).

(33) 4. Presumptive ID App/Module 1050

(34) In case of pure or predominate growth, a Presumptive ID module 1050 (See, FIG. 2A) may identify the main organisms potentially growing on a given media using a set of identification algorithms based on the training with the Data Lake. This module may provide/output a name of a highest ranking (e.g., probability) organism (or organism groups) and up to five colony locations ranked from highest to lowest probability for that identification. These algorithms enable identification of specific species on specific media types where any number of colonies are present and considered clinically significant.

(35) For example, the media and colony identities may be those indicated in the Table of FIG. 3. Rules that action workup of these colonies can be included in the Module. As an example of a specific IDApp, a urine culture App (UCA) and Chrom ID App could enable presumptive ID on Orientation CHROMagar from urines for those organisms claimed by the media. Rules would apply the ability to auto report/auto release (or batch), and downstream workup (e.g., automatic picking, testing, etc.). In some cases, rules may determine High Value Positives to flag plates that should be rapidly reviewed and directed for further process by a work list or automated pick, etc.

(36) 4.1 Purity Plate App/Module

(37) In some versions, the system may implement a purity plate module. Pure, predominant and complex plates may require a minimal number of isolated colonies in each. Predominant growth has typically greater than two colony types, where one colony type is greater than ten times the other colony type. Complex type typically has greater than two colony types with no predominate isolate or if the isolate is not identifiable (presumptive ID table below). Complex plates are typically manually interpreted. Thus, the module may classify the plate image according to whether it is pure, predominant (which can be slightly mixed) and/or complex.

(38) 4.2 AutoSelect ID/AST Module

(39) Pure and predominant plates can have representatives of each colony type indicated by the Purity Plate Mod with associated Rules that drive further workup as set forth in the examples above. A pre-determined number of each colony type could be designated for automatic identification (ID) and AST workup on an ID/AST module that may involve an automatic picking system/robot.

(40) 5. Kirby-Bauer (KB) Zone Diameter Measurement App Module

(41) Some embodiments may utilize an optional measurement App. Such an App may leverage existing imaging capabilities and the AST Expert Systems to provide zone measurements. Optionally these measurements may be linked to expert systems to provide interpretations. Opportunities also exist for a version of this App for Early zone measurements for particular drug/organism combinations and for zones on media plated directly from positive blood culture. Such algorithms may be based on metadata and/or images of the Data Lake.

(42) In some versions, this App would leverage existing imaging capabilities and an AST Expert System to provide zone measurements and Abx Disk identification. Optionally these measurements could be linked to expert systems to provide guidance on an antibiotic susceptibility profile for a pathogen isolated from the patient and guidance on treatment/response. In some versions, implementations of this App may provide early zone measurements for particular drug/organism combinations and for zones on media plated directly from positive blood culture. Some Apps may include expert systems (interpretations) and could be facilitated significantly with the Data Lake being stocked, monitored, audited appropriately and with required metadata and images.

(43) Although a system 100 may include any one or more of the above imaging related modules/Apps, in some versions a particular segmentation of functionality of the modules may be implemented by the following set of discrete modules/Apps: (a) Quadrant-Quantitation; (b) Detection of No Growth; (c) Purity Plate: # Colony Types (e.g., pure, Predominant, Complex) (d) Screening (e.g., CHROMagar i.e. MRSA); (e) Critical pathogens; (f) Early Growth Detection; (g) Auto Select ID/AST and (h) Zone measurement Kirby-Bauer.

(44) Imaging Apps and Launch Packages

(45) The system 100, providing the combination of algorithms/modules/rules, the Data Lake, and the ability to extract predetermined subsets of data, can enable an on-demand ability to rapidly mature algorithms and develop Apps. As an example, Apps may be developed based on specimen type and may be implemented with a collection of Apps. The strategy to implement a collection of Apps is influenced by many factors: supporting software launch cadence; value of individual Apps vs Apps being together; the availability of certain algorithms or specimen/plate/organism types in the Data Lake, etc.

(46) An example may be considered in relation to the following table:

(47) TABLE-US-00001 TABLE 1 Exemplary Modules and Their Functions Module Number Module Name Function 1 Urine Culture Quantitation into 5 buckets; apply Quantitation user threshold rules 2 Urine Culture Auto- Auto Release (in Europe) of No Negative Release Significant Growth Plates 3 Urine Culture Quantitation into 5 buckets on each Quantitation-BiPlate half with rules 4 Urine Culture Early Earliest Growth Detected on Plate Growth Detection Provides a Notification to a User 2.1 Urine Culture Batch- Batch Release (in the US) of No Negative Release Growth Plates 5 Urine Presumptive ID Orientation CHROMagar-based ID of claimed taxa

(48) In this specimen-based example, a series of 5 different Modules concerns one specimen type. Apps validated against a particular specimen type is one way to package functionality, however, an Exponential approach will also allow other options. For example, Surveillance Apps will allow launch by particular targeted organism (MRSA, Streptococcus, Shigella); Quantitation Apps could be packaged by media type (quantity of the sample on blood agar, independent of specimen), etc. In this sense, however, certain Apps will likely have minimal value for certain Specimens (i.e., quantity of the sample on nonselective media with sputum, given high normal flora levels). Note that the Apps can be delimited by region with specific rules and specific functions limited to the geographic region from which the specimen under evaluation was obtained. Table 1 identifies functions specific to clinical requirements particular to the United States (US) and Europe (EU).

(49) Thus, potential Apps may be segregated into two-high-level buckets. A first bucket collection may be considered screening and key identification Apps. Such Apps typically target specific organisms on CHROMagars, or for high-value pathogens on, for example, Blood agar. Each of these are discrete and can be prioritized for development with minimal impact to other Apps or specimen types and launched individually if desired. Similarly, the Kirby-Bauer zone App is generally independent of the next generation Apps (Next Gen App), which have associated algorithms that can be independently prioritized. Additionally, as new CHROMagars (i.e., vancomycin resistant enterococci (VRE)) become available, appropriate specimens can be run to populate the Data Lake and be added to this list. If targeted isolates are relative rare, the Data Lake may be supplemented with contrived (spiked specimen) samples. Additional example screening and Key ID Apps are illustrated in the following Table.

(50) TABLE-US-00002 TABLE 2 App Classification, Construction, Function and Output Data Required App Type of Output/Instruction to Train and App Category Classification App Function Specimen From App Run App Critical Group B Strep Auto Negative urine Auto-Neg (EU only) Training Data; Pathogen on CHROMagar Report for Batch - US and EU data lake Orientation Group B Strep Auto Neg - US FDA validation; initial batch clinical review and release submission auto negative release in the US with clinical submission Critical MRSA on Flag growth Nasopharyngeal Auto-Neg (EU only) Training Data; Pathogen CHROMagar for MRSA perirectal Batch - US and EU data lake MRSA II initial batch Auto Neg - US FDA validation; review and release; clinical auto negative submission release in the US with clinical submission Critical Group B Strep on Flag growth for Vaginal Auto-Neg (EU only) Training Data; Pathogen TSA Blood Agar Group B Strep Batch - US and EU data lake initial batch Auto Neg - US FDA validation; review and release clinical auto negative submission release in US with clinical submission Critical Group A Strep Flag growth for Nasopharyngeal Auto-Neg (EU only) Training Data; Pathogen on Selective Group A Strep perirectal Batch - US and EU data lake Strep Agar and initial batch Auto Neg - US FDA validation; Blood Agar review and release clinical auto negative submission release in US with clinical submission Critical CHROMagar Flag growth for CPE Nasopharyngeal Auto-Neg (EU only) Training Data; Pathogen Carbapenems initial batch perirectal Batch - US and EU data lake review and release Auto Neg - US FDA validation; auto negative clinical release in US submission with clinical submission Critical CHROMagar Flag growth Nasopharyngeal Auto-Neg (EU only) Training Data; Pathogen ESBL for ESBL perirectal Batch - US and EU data lake initial batch Auto Neg - US FDA validation; review and release clinical auto negative submission release in US with clinical submission Critical Salmonella and Flag growth Stool Auto-Neg (EU only) Training Data; Pathogen Shigella on for Salmonella Batch - US and EU data lake Hecktoen, and Shigella Auto Neg - US FDA validation; XLD and SS agar initial batch clinical review and release submission auto negative release in US with clinical submission Critical N. gonorrhea Flag growth urogenital Auto-Neg (EU only) Training Data; Pathogen on Thayer for N gonorrhea Batch - US and EU data lake Martin media initial batch Auto Neg - US FDA validation; review and release clinical auto negative submission release in US with clinical submission Critical Haemophilus Flag growth Respiratory Auto-Neg (EU only) Training Data; Pathogen on Chocolate for Haemophilus Batch - US and EU data lake Agar initial batch Auto Neg - US FDA validation; review and release clinical auto negative submission release in US with clinical submission KB zone Zone measurement Zone measurement, Colony to KB EU Training Data; linked to Abx code user review and release US data lake validation. KB zone SIR Zone measurement Zone measurement, Colony to KB EU Training Data; linked to Abx user review and release US data lake code and simple validation. look up table for SIR KB zone Zone + Abx Zone measurement Colony to KB EU Training Data; expert code linked to linked to EuCAST US data lake an Expert Expert System may validation, System defined require new rules organism x organism x drug for certain organism x drug (EU); drug, Abx ID; For US add Zone measurement clinical software linked to FDA/CLSI submission Expert System KB early zone Zone + Abx Zone measurement, Colony to KB EU Training data, code at 10-12 user review and release US digital lake hours validation, organism x drug (EU); for US add manual v. digital submission. KB early zone Zone measurement; Zone measurement Colony to KB EU Training data, expert Expert System linked to EuCAST US digital lake defined Expert System may validation, organism x drug require new rules organism x for certain organism x drug (EU); for drug, Abx ID; US add clinical Zone measurement submission linked to FDA/CLSI software. Expert System KB zone- Zone measurement; Zone measurement Colony to KB EU Training data, positive blood Expert System linked to novel US database validation, culture defined expert system expert rules (EU) organism x drug for US add FDA submission and organism x drug

(51) It can be observed from Table 2 that Apps can be quite specific and that the outputs can depend on specimen classification (i.e. type of specimen, US region or EU region, etc.). Table 2 also illustrates at a high level, the type of data used to train the App.

(52) A second bucket collection of Apps use more generic algorithms and can be prioritized and grouped for launch by several criteria. A summary of examples of these Apps is provided in Table 3 below. In essence, each cell in Table 3 represents an App. Cells in the Table below that share the same number are appropriate capabilities for that specimen type where the shared cell numbers are reasonably packaged together in a common module. With this model, there are 8 additional launch packages/modules.

(53) TABLE-US-00003 TABLE 3 Sterile Fluids and Superficial Gastro- Capability Urine Tissue Wounds Respiratory Throat Intestinal Urogenital No Growth 1 1 1 2 2 2 2 Quadrant Growth 2 3 6 7 7 8 9 Score (+, ++, +++) Plate purity 5 3 6 7 7 8 9 (pure, predominant, complex Presumptive ID on 5 3 6 7 7 8 9 pure, predominant plates Early Growth 5 4 4 2 2 2 2 Detection Auto picking 5 3 6 7 7 8 9 pure, predominant
Example imaging modules may be considered with reference to the following table:

(54) TABLE-US-00004 TABLE 4 List of Exemplary Modules Imaging Module Mod 1. UCA 1.0 Urine Culture Semi-Quantitation Mod 2. UCA 1.0 Urine Culture Auto-Negative Release Eu Mod 3. UCA 1.5 Urine Culture Semi-Quantitation-BiPlates Mod 4. UCA 1.5 Urine Culture Early Growth Detect Mod 2.1 UCA 1.5 Urine Culture batch-Negative Release (US) Mod 5. UCA 1.8 Presumptive ID BBL Orientation CHROMagar for 6 groups; Group B Strep flag Mod 6. UCA 1.8 Pure/Predominant/Complex on BBL Orientation CHROMagar Mod 6.2 UCA 2.0: UCA 1.8-autorelease US. 510K approval. 25 mp camera launch Mod 7. Surveillance: MRSA on BBL CHROMagar. Batch Release EU/US Mod 8. Surveillance: Carbepenem producing Enterobacteriaciae (CPE) on BBL CHROMagar Mod 9. Surveillance: ESBL on BBL CHROMagar Mod 9. Critical Specimens: Early Growth (sterile fluids, superficial wound, tissue, urine) Mod 10. Critical Specimens: Presumptive ID Mod 11. Critical Specimens: Quadrant quantitation, Mod 12 Critical Specimens: Pure/Predominant/Complex Mod 13. Critical Specimens: No Growth Detection (user defined time pts). Batch/Auto Mod 14. Critical Specimens: Auto Select ID/AST (Wagtail) Mod 15. KB Zone (disk ID, Zone size) manual interpretation Mod 16 KB Zone SIR (susceptible, intermediate, resistant) manual release Mod # Critical Pathogen: Group A Strep on Selective Strep Agar, and Blood Agar Mod # Respiratory: Quadrant quantitation; Mod # Respiratory: Pure/predominant/complex Mod # Respiratory: Presumptive ID Mod # Respiratory: Auto Select ID/AST Mod # Respiratory: Critical Pathogen: Haemophilus on chocolate agar Mod # Stool: Negative growth batch report from enrichment culture on selective agar Mod # Stool: Negative growth batch reporting from primary culture Mod # Stool: Quadrant quantitation Mod # Stool: Pure/Predominant/Complex Mod # Stool: Critical Pathogen: Salmonella and Shigella on Hecktoen, XLD, and SS media Mod # Stool: Auto Pick (Wagtail) Mod # KB-Zone-expert System. CLSI (Clinical and Laboratory Standards Institute)/EUCAST guidelines. SIR

(55) EUCAST is the European Committee on Antimicrobial Susceptibility Testing. In one example of a process integrated with one or more Apps for specimen evaluation and process control, a specimen is inoculated onto a plated media using BD Kiestra InoqulA. The specimen is streaked onto the media using a predetermined pattern which is tracked as part of metadata via the bar code. The streaked sample is conveyed into BD Kiestra ReadA compact where it is incubated and imaged at times determined by the App. The images obtained by ReadA at the appointed time is analyzed by the App to determine if the specimen on the plate is pure. Further work up of the specification is performed based on the results of the determination. The image is evaluated to identify the coordinates of the selected colonies. The App can convey those coordinates to an apparatus (or technologist). The App can issue instructions to convey the specimen to the apparatus in which the colony will be picked. The App can further coordinate or control the picking of the colonies and conveying the picked colonies to another platform where ID of the pathogen is performed. In one example ID is performed by MALDI. As described elsewhere herein, a sample is evaluated by MALDI by placing the picked sample into suspension and inoculating a MALDI plate with the suspension. The App can also coordinate or control transfer of the colony suspension to BD Kiestra InoqulA. There the suspension can be inoculated onto another type of culture media (e.g., Mueller Hinton) using a spread pattern and then moved to an AST testing apparatus where predetermined antibiotic disks (e.g. BD BBL Sensi-DiscsT) are place on the culture. The plate carrying the inoculated specimen and the antibiotic disks is then conveyed to a ReadA compact under coordination and control of the App. The ReadA obtains images and provides those images to the App, the results of which are conveyed from the App to an Expert System for analysis of the resulting antibiotic disk zones and interpretation of the results. The Expert System then conveys the results of the analysis to the clinical lab staff

(56) Although the invention herein has been described with reference to particular embodiments, it is to be understood that these embodiments are merely illustrative of the principles and applications of the present invention. It is therefore to be understood that numerous modifications may be made to the illustrative embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims.