Automatic Counting System

Abstract

The present disclosure provides systems and methods for detecting the presence of colonies on agar plates. The systems and methods are preferably useful for assessing the presence of contaminants (e.g., bacteria, fungi, etc.) in samples, such as raw materials or components of pharmaceutical compositions, by analyzing whether the sample forms colonies on an agar plate. The systems use lighting arrangements to enhance the accuracy of the analysis and, in embodiments, include the use of artificial intelligence as part of the automated process.

Description

[Technical Field]

[0001]

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 380,665, filed October 24, 2022, and U.S. Provisional Patent Application No. 63 / 382,128, filed November 3, 2022, both of which are entitled "Automated Enumeration System" and list Archibald Delorme and Zichao BIAN as inventors. The contents of each of these applications are incorporated herein by reference in their entirety.

[0002]

[0002] The present disclosure provides systems and methods for detecting the presence of colonies on agar plates and quantifying those colonies. The systems and methods are preferably useful for assessing the presence of contaminants (e.g., bacteria, fungi, etc.) in samples, such as raw materials or components of consumer or pharmaceutical compositions, by analyzing whether the sample forms colonies on an agar plate. The systems use lighting arrangements and cameras to enhance the accuracy of the analysis, and in embodiments, include the use of artificial intelligence as part of the automated process. [Background technology]

[0003]

[0003] Consumer and pharmaceutical quality control (QC) departments often require microbial detection and enumeration systems to capture data from samples, including raw materials, water (purified, water for injection), intermediate materials, bulk drug substances, and environmental monitoring. Most QC relies on highly subjective and time-consuming manual counting methods. New regulations are challenging the accuracy of these methods, prompting the use of a four-eyes concept, where one person counts colonies on a plate and another person verifies the count for data integrity.

[0004]

[0004] US Pat. No. 9,290,382 provides an instrument for detecting microorganisms in target cells, but images the colonies through the lid of the Petri dish, which can cause errors in the detection.

[0005]

[0005] What is needed is an automated system and method for colony counting that is highly reproducible and requires less input from an operator. The present invention meets these needs. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] U.S. Patent No. 9,290,382 Summary of the Invention

[0007]

[0006] In an embodiment, there is provided herein a system for detecting colonies growing on an agar plate, the system comprising a photoelectric array detector having associated optics for detecting a detection area on the surface of the agar plate, one or more illumination sources for illuminating the detection area, the one or more illumination sources being positioned to illuminate the agar plate, a focusing lens system positioned between the photoelectric array detector and the agar plate, and a computer programmed to receive data collected by the photoelectric array detector, the data being a digital representation of the detection area, colonies growing on the agar in the plate being detected through the bottom surface of the agar plate and the agar, and the computer analyzing the data to quantify the number of colonies in the detection area.

[0008]

[0007] In a further embodiment, there is provided herein a system for detecting colonies growing on an agar plate, the system comprising a photoelectric array detector having associated optics for detecting a detection area on the underside of the agar plate, a plurality of illumination sources surrounding the sides of the agar plate for illuminating the detection area, a focusing lens system positioned between the photoelectric array detector and the agar plate, and a computer programmed to receive data collected by the photoelectric array detector, the data being a digital representation of the detection area, colonies growing on the agar in the plate being detected through the bottom surface of the agar plate and the agar, and the computer analyzing the data to quantify the number of colonies in the detection area.

[0009]

[0008] In a further embodiment, there is provided herein a method for quantifying the number of colonies growing on an agar plate, the method comprising the steps of illuminating a detection area on the underside of the agar plate with one or more illumination sources, the illumination sources being positioned to illuminate the front side of the agar plate; detecting the detection area with a photoelectric array detector having associated optics, the detecting being performed via a focusing lens system positioned between the photoelectric array detector and the agar plate; receiving data collected by the photoelectric array detector as a digital representation of the detection area; and analyzing the collected data to quantify the number of colonies in the detection area.

[0010]

[0009] There is further provided herein a method for training a neural network, the method comprising the steps of acquiring test images of colonies growing on an agar plate, illuminating a detection area on the surface of the agar plate with one or more illumination sources, the illumination sources being positioned to illuminate the agar plate; detecting the detection area with a photoelectric array detector having associated optics, the detecting being performed via a focusing lens system positioned between the photoelectric array detector and the agar plate; receiving data collected by the photoelectric array detector as a digital representation of the detection area; and generating an image of cells within the detection area; providing a training image set having a fixed number of colonies growing on the agar plate; comparing the test images with the training image set by a computing system to generate an indication of the number of colonies in the test images; and repeating the acquiring, providing and comparing steps with respect to a plurality of further test images to train the neural network.

[0011]

[0010] In a further embodiment, there is provided herein a method for quantifying the number of colonies growing on an agar plate, the method comprising the steps of: illuminating a detection area on the surface of the agar plate with one or more illumination sources, the illumination sources positioned to illuminate the surface of the agar plate; detecting the detection area with a photoelectric array detector having associated optics, the detecting being performed via a focusing lens system positioned between the photoelectric array detector and the agar plate; receiving data collected by the photoelectric array detector as a digital representation of the detection area; generating images of colonies in the detection area; applying, by a computer system, a trained neural network to the images of colonies to result in a quantitative determination of the number of colonies in the detection area, the trained neural network having been trained with a training image set generated from images of the detection area where colonies growing on the agar in the plate are detected on the surface of the agar plate and through the agar; and displaying the quantitative determination of the number of colonies in the detection area. [Brief explanation of the drawings]

[0012] [Figure 1A] FIG. 1 shows an exemplary agar plate. [Figure 1B] FIG. 1 shows an exemplary agar plate. [Figure 1C] FIG. 1 is a side view of an agar plate. [Figure 1D] FIG. 1 shows an exemplary system for detecting colonies growing on an agar plate, as described herein. [Figure 2A] FIG. 1 shows a further exemplary system for detecting colonies growing on an agar plate, as described herein. [Figure 2B] FIG. 1 illustrates an exemplary illumination source according to embodiments described herein. [Figure 3]FIG. 1 shows yet a further exemplary system for detecting colonies growing on an agar plate, as described herein. [Figure 4] FIG. 1 shows a system for detecting colonies growing on an agar plate, including a stage and a robotic arm. [Figure 5] FIG. 1 is a flow diagram illustrating the overall flow of a method and operations for training a neural network for automatic colony detection implemented by a computer system. [Figure 6] FIG. 1 shows exemplary agar plates used in a training set for training a neural network. [Figure 7A] FIG. 1 illustrates a computing system according to embodiments herein. [Figure 7B] FIG. 1 illustrates a non-transitory computer-readable medium according to embodiments herein. [Figure 8A] FIG. 1 shows an image of an agar plate taken through the top cover. [Figure 8B] FIG. 1 shows an image of an agar plate taken through the top cover. [Figure 9A] FIG. 1 shows an image taken through the bottom surface of an agar plate according to embodiments herein. [Figure 9B] FIG. 1 shows an image taken through the top surface of an agar plate according to embodiments herein. [Figure 10A] FIG. 1 shows an image collected using a system described herein with an illumination source illuminating the side of the agar plate and the agar. [Figure 10B] FIG. 1 shows an image collected using a system described herein with an illumination source illuminating the side of the agar plate and the agar. [Figure 11] FIG. 1 shows an example of the results of an algorithm applied using images from the system described herein. DETAILED DESCRIPTION OF THE INVENTION

[0013]

[0026] The use of the words "a" or "an" when used in conjunction with the term "comprising" in the claims and / or specification can mean "one," but is also consistent with the meanings of "one or more," "at least one," and "one or more than one."

[0014]

[0027] Throughout this application, the term "about" is used to indicate that a value includes the variation of error inherent in the method / device being used to determine the value. Typically, the term "about" is meant to encompass a variability of approximately 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 16%, 17%, 18%, 19%, or 20% or less, depending on the context.

[0015]

[0028] Although the use of the term "or" in the claims is used to mean "and / or" unless expressly specified to refer to alternatives only or unless those alternatives are mutually exclusive, the present disclosure supports a definition that refers to alternatives only and "and / or."

[0016]

[0029] As used in this specification and claims, the words "comprising" (and any form of comprising, such as "comprise" and "comprises"), "having" (and any form of having, such as "have" and "has"), "including" (and any form of including, such as "includes" and "include"), or "containing" (and any form of containing, such as "contains" and "contain") are inclusive or open-ended and do not exclude additional, unrecited elements or method steps.

[0017]

[0030] The systems, methods, processes, and devices described herein are useful for detecting colonies (e.g., bacteria, fungi, etc.) growing on agar plates and ultimately performing qualitative or quantitative evaluation of those colonies. As described herein, there is a need for such automated counting systems in consumer and pharmaceutical QC to capture data from samples, including raw materials, water (purified, water for injection), intermediate materials, bulk drug substances, and environmental monitoring. It is also desirable to be able to obtain accurate endpoint counts using non-proprietary consumables (i.e., plates, media, or filters). However, the use of traditional Petri dishes (also referred to herein as agar plates) in such counting systems and methods can pose challenges because condensation can often occur when colonies are viewed through the lids of such plates to avoid contamination caused by spore-forming molds and other contaminants, making it difficult to obtain clear images for analysis by automated systems.

[0018] System for quantification of colonies

[0031] In an embodiment, a system for detecting colonies growing on an agar plate is provided herein. FIG. 1A shows an agar plate 106 containing multiple colonies 103 growing on agar 108 within the plate 106. An agar plate, culture plate, culture dish, or flat, also called a Petri dish, refers to a shallow, transparent, lidded dish made of plastic (glass plates, including borosilicate glass, can also be used) that is generally cylindrical in shape and holds a growth medium (agar) in which cells can be cultured. The cells and colonies are preferably bacteria, fungi, or yeast. As used herein, a "colony" refers to a group or cluster of cells of an organism (preferably bacteria, fungi, or yeast) derived from the same mother cell, regardless of whether they are visible to the naked eye. Colonies constitute clones of the mother cell that are all genetically similar. As shown in FIG. 1A, multiple individual colonies 103 are visible to the naked eye in a photograph of the agar plate. Colonies can grow any suitable agar medium composition, including, for example, heterologous polysaccharides based on galactose. Agar can be composed of agarose and agaropectin polymers. A typical agar composition is 70% agarose and 30% agaropectin. Additional agar compositions are known in the art, including, for example, soy agar, Reasoner's 2A agar, blood agar, antibiotic-treated agar, Sabouraud agar, nutrient agar, and the like. Colonies can also be grown on absorbent pads or membrane surfaces in the agar within the agar plate, including mixed cellulose ester (MCE) filter membranes, and the like.

[0019]

[0032] 1B shows another exemplary agar plate 106 with condensation 105 on the plate lid. During testing and analysis, it is desirable to keep the lid on the agar plate to avoid contamination. However, as is readily apparent, the condensation obscures the visibility of any colonies 103 growing on the agar within the plate, and therefore, an accurate assessment of the number of colonies growing on the agar plate cannot be obtained. The present systems, methods, and devices enable the detection and counting of colonies growing on the agar plate, even in the presence of condensation 105.

[0020]

[0033] 1D, 2A, and 3 illustrate exemplary systems described herein for detecting colonies growing on an agar plate, in which colonies growing on an agar plate are detected through the surface of the agar plate, preferably through the bottom surface 113 of the agar plate 106.

[0021]

[0034] The system 100 shown in FIG. 1D preferably includes a photoelectric array detector 102 with associated optics for detecting a detection region of an agar plate 106. FIG. 1A shows an exemplary detection region 111 on a bottom surface 113 of an agar plate 106. As shown, in embodiments, the detection region 111 may preferably be the entire bottom surface 113 of the agar plate 106 (i.e., the entire circular area indicated by the thick black line). In other embodiments, the detection region 111 may be a section or portion of the entire bottom surface 113. In further embodiments, the detection region 111 may be any section or portion of the agar 108. FIG. 1C shows a side view of an agar plate showing the bottom surface 113, agar 108, agar top surface 117, and top cover 110 (also called a lid) of the agar plate.

[0022]

[0035] The systems described herein for detecting colonies growing on an agar plate preferably include one or more illumination sources for illuminating the detection area 111. As described herein, the illumination sources are preferably positioned to illuminate the agar plate. As used herein, "illumination source" refers to a light, laser, combination of light sources, or other element that provides light to enable data from the detection area to be collected by a photoelectric array detector.

[0023]

[0036] In system 100, illumination source 118 illuminates top agar surface 117 and agar 108 of agar plate 106. As shown in Figure 1D, illumination source 118 is preferably positioned so that light shines through top surface 117 of agar plate 106, into the agar 108 itself, illuminates detection zone 111, and exits bottom surface 113 of the agar plate.

[0024]

[0037] The light then passes through a collection lens system 104 positioned between the photosensitive array detector 102 and the agar plate 106. The system 100 also includes a computer 120 programmed to receive data collected by the photosensitive array detector 102. The data is preferably a digital representation of the detection area 111. Colonies 103 growing on the agar 108 in the agar plate 106 (preferably at the agar top surface 117) are detected through the surface (preferably the bottom surface 113) of the agar plate and the agar, and the computer analyzes the data to quantify the number of colonies 103 in the detection area 111. In embodiments, the collection lens system 104 is preferably a fixed focal length lens with a large field of view (e.g., a lens with a working distance of approximately 100-200 mm, including a 16 mm fixed focal length lens). In other embodiments, the collection lens system 104 can be a telecentric lens, a microscope objective lens, a liquid lens, a magnifying lens, or any other lens system capable of collecting light output from the sample plate and focusing it onto the detection surface of the photosensitive array detector. The distance between the photosensitive array detector and the detection area can vary and can range between 0-5 centimeters, 5-10 centimeters, 10-15 centimeters, 10-20 centimeters, 20-25 centimeters, 25-30 centimeters, or more, depending on the detector setup and specifications.

[0025]

[0038] In such an embodiment, system 100 provides a device and method for detecting and counting colonies growing on agar by acquiring images through the agar plate's bottom surface 113. Imaging colonies through the agar plate's bottom surface 113 reduces or eliminates imaging problems associated with condensation on the agar plate lid, making imaging difficult, allowing for accurate determination of the number of colonies growing on the plate even when the plate lid remains on.

[0026]

[0039] In an exemplary embodiment, system 100 further includes a light diffuser 112 positioned between illumination source 118 and agar plate 106. Exemplary light diffusers are known in the art and include, for example, various plastic, glass, or other material elements that act as a semi-transparent material positioned between the light source and agar plate 106 to spread, diffuse, or disperse light as it passes through the material. This material does not block or cut light, but rather redirects the light as it passes through due to the diffused light being spread onto the agar plate.

[0027]

[0040] 2A represents an embodiment in which one of the illumination sources 202 used to illuminate the detection region illuminates the side 220 of the agar plate 106 and the agar 108, since the side 220 forms a perimeter surrounding the bottom surface 113 of the agar plate 106. Preferably, the illumination source 202 includes multiple illumination sources surrounding the side 220 of the agar plate 106. As used herein, a plurality of illumination sources refers to two or more, preferably three or more, four or more, five or more, six or more, seven or more, eight or more, nine or more, ten or more, eleven or more, twelve or more, thirteen or more, fourteen or more, fifteen or more, sixteen or more, seventeen or more, eighteen or more, nineteen or more, twenty or more, twenty-one or more, twenty-two or more, twenty-three or more, twenty-four or more, twenty-five or more, twenty-six or more, twenty-seven or more, twenty-eight or more, twenty-nine or more, thirty or more, or between about 10-40, between about 20-40, or between about 20-30 illumination sources 202.

[0028]

[0041] In an embodiment, the illumination source 202 is a series 240 of individual illumination sources 202 (see FIG. 2B ). The multiple illumination sources preferably surround the side 220 of the agar plate 106 at a location between the top cover 110 of the agar plate 106 and the top agar surface 117 of the agar plate 106. In other embodiments, the multiple illumination sources can surround the side 220 of the agar plate 106 at a location between the bottom surface 113 of the agar plate and the top agar surface 117 of the agar plate 106. In other embodiments, the illumination sources can surround the side 220 of the agar plate 106 up to the lid 110 of the agar plate 106. FIG. 2A shows an area 204 within which the illumination sources are preferably positioned relative to the agar plate 106 so that the agar plate is illuminated. In an embodiment, the series 240 of illumination sources 202 can be connected together and surround the side 220 of the agar plate 106, as shown in FIG. 2B. FIG. 2B further shows a support 260 on which the agar plate 106 can be placed such that the side 220 of the agar plate 106 is illuminated. However, in other embodiments, the separate individual illumination sources 202 can simply be arranged such that they surround the side 220 of the agar plate 106. As used herein, "surrounding" includes instances where the illumination sources illuminate the entire side 220 (and therefore the circumference) of the agar plate 106 in a circular or near-circular pattern of the illumination sources, as well as partially illuminating the circumference of the agar plate 106.

[0029]

[0042] 3 illustrates an embodiment that includes two illumination sources for illuminating the detection region, including a first illumination source 202 positioned such that it can illuminate the side 220 of the agar plate 106 and the agar 108, along with a second illumination source 118 that illuminates the top surface 117 of the agar plate 106 and the agar 108. In one embodiment, the illumination source 202 can be positioned such that a mirror can be used to direct light onto the agar plate to illuminate the detection region.

[0030]

[0043] Illumination sources for use in the various systems and devices described herein are preferably light emitting diodes (LEDs), although in other embodiments the illumination source may include one or more lasers (e.g., 532 nm frequency doubled diode lasers) or may include a combination of light sources or other elements that provide light to enable data from the detection region to be collected.

[0031]

[0044] The photoelectric array detector 102 is a device or apparatus that converts optical signals into electrical signals, and in embodiments includes a charge-coupled device (CCD) detector, a photomultiplier tube detector, a complementary metal-oxide semiconductor (CMOS) detector, or a photodiode detector, preferably a CMOS camera.

[0032]

[0045] 4 illustrates a further embodiment in which the systems described herein also include a stage 402 for positioning the agar plate 106 relative to the illumination sources 202 and / or 118 and the photosensitive array detector 102. In an embodiment, the systems described herein may further include a robotic arm 404 for placing and removing the agar plate on the stage 402. The robotic arm 404 is shown for illustrative purposes only and should not be considered limiting of the types of robotic arms that may be used in conjunction with the systems described herein.

[0033]

[0046] As explained throughout, it has surprisingly been found that the systems provided herein allow for illumination and detection of colonies in the detection area even in the presence of the top cover 110. Although the top cover may contain condensation (e.g., due to respiration of the colonies or residual moisture in the agar), the systems and methods described herein allow for the detection of colonies growing on the agar and a reliable determination of the number of colonies.

[0034]

[0047] In a preferred embodiment, provided herein is a system for detecting colonies 103 growing on an agar plate 106, the system comprising a photoelectric array detector 102 with associated optics for detecting a detection area 111 on the underside 113 of the agar plate 106. As described herein, the system preferably includes a plurality of illumination sources 202 surrounding a side 220 of the agar plate 106 for illuminating the detection area 111. The system also preferably includes a collection lens system 104 positioned between the photoelectric array detector 102 and the agar plate 106. As described herein, the system further includes a computer 120 programmed to receive data collected by the photoelectric array detector 102. As described herein, the data is a digital representation of the detection area 111, where colonies 103 growing on the agar 108 in the plate 106 are detected through the bottom surface 113 of the agar plate 106 and the agar 108, and the computer 120 analyzes the data to quantify the number of colonies 103 in the detection area 111.

[0035]

[0048] As described herein, the system preferably further includes a second illumination source 118 that illuminates the agar top surface 117 and the agar 108 of the agar plate 106, and a light diffuser 112 preferably positioned between the second illumination source and the agar plate. The plurality of illumination sources 202, for example in the form of a series 240 lights (such as LED lights or laser light sources), preferably surround the sides of the agar plate at a location between the bottom surface of the agar plate and the top of the agar plate.

[0036]

[0049] Exemplary photoelectric array detectors are described herein, including, for example, charge-coupled device (CCD) detectors, photomultiplier tube detectors, complementary metal-oxide semiconductor (CMOS) detectors (including CMOS cameras), or photodiode detectors.

[0037]

[0050] In a preferred embodiment, the system may further include a stage 402 for positioning the agar plate 106 relative to the illumination source 202 and the photoelectric array detector 102, and a robotic arm 404 for placing and removing the agar plate on the stage.

[0038] Automated method for quantification of colonies

[0051] In a further embodiment, there is provided herein a method for quantifying the number of colonies growing on an agar plate, the method preferably comprising illuminating a detection area on the underside of the agar plate with one or more illumination sources, the illumination sources being positioned to illuminate the front side of the agar plate.

[0039]

[0052] The method further includes detecting the detection area with a photoelectric array detector having associated optics, the detecting being performed by a collection lens system disposed between the photoelectric array detector and the agar plate. Data collected by the photoelectric array detector is received as a digital representation of the detection area. The method further includes analyzing the collected data to quantify the number of colonies in the detection area.

[0040]

[0053] The method is preferably automated in that no user input is required to position the agar plates, illuminate the plates, and / or collect and analyze data; instead, these elements are performed by machine-driven processes. This automation reduces the likelihood of contamination of the agar plates with foreign bacteria, etc., and allows high-throughput methods to analyze many agar plates (and therefore samples) quickly and easily.

[0041]

[0054] As described herein, preferably, two illumination sources are used to illuminate the detection region, with a first illumination source illuminating the sides of the agar plate and the agar. The first illumination source preferably includes multiple illumination sources surrounding the sides of the agar plate, for example, between the bottom surface of the agar plate and the top surface of the agar plate. In embodiments, a second illumination source illuminates the top surface of the agar plate and the agar, and preferably the second illumination source illuminates the agar plate through a light diffuser positioned between the second illumination source and the agar plate.

[0042]

[0055] Exemplary illumination sources, lens systems and photosensitive array detectors are described herein.

[0043] Methods for training neural networks and methods for quantification using artificial intelligence

[0056] In a further aspect, provided herein is a method of training a neural network, the method including acquiring test images of colonies growing on an agar plate, the acquiring step including illuminating a detection area on the surface of the agar plate with one or more illumination sources positioned to illuminate the agar plate; detecting the detection area with a photosensitive array detector having associated optics, the detecting being performed via a collection lens system positioned between the photosensitive array detector and the agar plate; receiving data collected by the photosensitive array detector as a digital representation of the detection area; and generating an image of cells in the detection area; providing a training image set having a fixed number of colonies growing on the agar plate; comparing the test images with the training image set by a computing system to generate an indication of the number of colonies in the test images; and repeating the acquiring, providing, and comparing steps with a plurality of additional test images to train the neural network.

[0044]

[0057] Further provided herein is a method for quantifying the number of colonies growing on an agar plate, the method comprising: illuminating a detection area on the surface of the agar plate with one or more illumination sources, the illumination sources positioned to illuminate the surface of the agar plate; detecting the detection area with a photoelectric array detector having associated optics, the detecting being performed through a focusing lens system positioned between the photoelectric array detector and the agar plate; receiving data collected by the photoelectric array detector as a digital representation of the detection area; generating images of the colonies in the detection area; applying, by a computer system, one or more trained neural networks or AI algorithms to the images of the colonies to provide a quantitative determination of the number of colonies in the detection area, the trained neural networks having been trained with a training image set generated from images of the detection area where colonies growing on the agar in the plate are detected on the surface of the agar plate and through the agar; and displaying the quantitative determination of the number of colonies in the detection area.

[0045]

[0058] One aspect of the embodiments herein relates to a computing system for processing endpoint images of colonies using a trained neural network (or other form of machine learning) to count and analyze colonies and further enhance the accuracy and precision of the colony counting system. The computing system (e.g., computer 120) can include at least one processing circuit, a non-transitory computer-readable medium, a display device, and a communication interface. The endpoint images of the colonies are obtained from a photoelectric array detector having associated optics for detecting the detection region. The endpoint images can also represent test images of colonies growing on an agar plate 106.

[0046]

[0059] 5 provides a flow diagram illustrating a method and overall flow of operations for training a neural network for automated colony detection performed by a computer system. In one embodiment, method 500 includes operations 502-514, including receiving (or providing) an image set (502), dividing the image set into two portions, one for training and one for validation (504), initializing network weights (506), forward propagating (508) the neural network, using a cost function to optimize the network's parameters (510), backpropagating (512) through an optimization technique, and validating the training using a validation database (514), and then repeating these steps (508) with a plurality of additional images to detect and quantify the number of colonies growing on a plurality of agar plates, as depicted in a plurality of additional test images.

[0047]

[0060] In one embodiment, operation 502 includes receiving an image set including image data representing colonies 103 growing on an agar plate 106. The colonies 103 may include various bacteria, yeast, and mold. The image set can be generated by a system described herein. For example, generating a training image set can include illuminating a detection area on the surface of the agar plate with one or more illumination sources, the illumination sources positioned to illuminate the agar plate. The detection area can be detected by a photoelectric array detector with associated optics. Detection can be via a collection lens system positioned between the photoelectric array detectors as a digital representation of the detection area. Images of the colonies within the detection area can be generated as generated data for use in the training set(s).

[0048]

[0061] After receiving the image set, method 500 may include operation 504, in which the computer system divides the image set into two portions that can be used to train and validate a neural network to learn how to automatically detect and quantify the number of colonies growing on an agar plate. For example, the training image set may include a first image set representing an agar plate containing colonies 103 (i.e., bacteria, yeast, and / or mold) and a second image set representing an agar plate without growing colonies. In one embodiment, at least one image in the training image set may include an agar plate containing multiple overlapping or connected colonies. Annotations may be associated with the images that make up the image set to indicate the presence of objects of interest in the image. For example, an image that makes up the image set used for training may have several objects described by the annotation. In one embodiment, the annotation may indicate a number or count for each of the training image set or for each of the validation image set. In one embodiment, the annotation may indicate or identify the type of microorganism(s) present. Annotations may be created manually or may be applied to each image in an image set. For example, one image from a training image set may include multiple growths, or colonies 103, present on an agar plate 106. In that case, a bounding box 602 may be applied to each colony 103 present in the image along with an associated training score, as illustrated in FIG. 6.

[0049]

[0062] Method 500 may include operation 506, in which the weights of the network are initialized, allowing each node of the network to be associated with a particular value. For example, for training the algorithm, the weights may be initialized with random weights or may be initialized with weights from networks trained with similar types of data.

[0050]

[0063] Method 500 may include operation 508, in which the computer system receives training images having a fixed number of colonies growing on an agar plate. Generating the training images may include multiple steps. For example, generating the test images may include illuminating a detection area on the surface of the agar plate with one or more illumination sources, the illumination sources positioned to illuminate the agar plate. The detection area may be detected by a photoelectric array detector with associated optics. Detection may occur through a focusing lens system positioned between the photoelectric array detectors as a digital representation of the detection area. Images of the colonies in the detection area may be generated as generated data for use in the training set(s). In embodiments, several annotations may be added to the images to identify objects of interest present in the images.

[0051]

[0064] Method 500 may include operation 508, in which an image, which may be generated as described in the previous paragraph, is sent to a neural network (i.e., forward propagated through the neural network). The purpose of this step may be to regenerate a feature map corresponding to the image and extract information of interest from the feature map. In one embodiment, the neural network may be an auto-neural network, a convolutional neural network, a corner neural network, or any other type of neural network. For example, the neural network may be a convolutional neural network having a series of convolutional layers, which may include a first convolutional layer configured to apply a convolutional filter to an input test image to generate an activation map, and may further include additional convolutional layers configured to apply each convolutional filter in turn to a respective activation map or other output of a previous convolutional filter. Each of the series of convolutional layers may include a respective activation layer, which may apply an activation function (e.g., a rectified linear unit (ReLU) function) to the output of the convolutional filter in the convolutional layer. In one embodiment, training a neural network may involve a computing system adjusting weights or other parameter values ​​of convolutional filters or other components of a series of convolutional layers to cause the series of convolutional layers to transform a set of training images into output values ​​that correspond to the annotations used for training, or at least deviate by a small amount from the annotations used for training.

[0052]

[0065] Method 500 may include operation 510, in which a cost function is defined that can be used to describe how well the network is suited for each of the nodes of the network. In one embodiment, the cost function may be a regression cost function, a cost function for binary classification, or a cost function for multi-class classification. An implementation cost function applied to a neural network may be capable of detecting how well the neural network fits the data.

[0053]

[0066] Method 500 may include operation 512, in which the computer system performs backpropagation in the neural network. During this step, the weights of the network may be adjusted so that the final predicted results match expectations.

[0054]

[0067] Method 500 may include operation 514, in which a neural network is applied to a second portion of the image set, the validation data set. From this step, it may be possible to confirm that the network has properly learned from the first portion of the image set used to train the neural network. This may be done to compare annotations generated by the network in response to the validation image set as input, and to compare the generated annotations to the validation image set.

[0055]

[0068] Operations 506-508 can then be repeated and applied to multiple image sets to automatically and efficiently identify colonies using the trained neural network. When subjecting a computer system to this method, detection / counting of colonies present on an agar plate can be achieved with a minimum of 80% precision, more preferably 90% precision or 95% precision, with maximum recall to avoid any undetected colonies.

[0056]

[0069] The above embodiments may be implemented separately or may be combined. For example, a combination may include using one or more neural networks to extract image portions from a test image that represent colonies on an agar plate, to determine the growth level of colonies on the agar plate, to determine the total number of colonies present on the agar plate, and to determine the presence of moisture / water in or on the agar plate. Such a combination may use a single neural network or multiple neural networks.

[0057]

[0070] FIG. 7A illustrates a computing system 700 configured to train a neural network and output information, as described above. In one embodiment, the computing system includes a processing circuit 702 configured to execute a neural network training module 710 configured to train the neural network based on a training image set 706 and output training scores 708. The neural network training module 710 may rely on machine learning platforms and frameworks such as TensorFlow, PyTorch, Keras, OpenCV, etc. In one embodiment, the neural network may be trained with a large training image set 706 to reflect a wide range of colony growth values, colony sizes, instances of colony overlap, agar plate sizes, presence of moisture, and microbial species. FIG. 7B illustrates elements of a non-transitory computer-readable medium 720, which may include an image processing module 722, a neural network module 724, and test images 726. [Example]

[0058] Configuration of a system for detecting colonies on agar plates

[0071] A system for detecting colonies on agar plates, as described herein, was constructed.

[0059]

[0072] Initial tests were performed by imaging the agar plates with the top cover in place to minimize contamination of the plates. However, the resulting images were of poor quality due to condensation on the lid. See Figures 8A-8B. As noted, the condensation droplets 802 obscured the colonies 103 growing underneath. Use of an automated imaging and counting system to quantify these colonies was unsuccessful, as many of the colonies were not counted or the condensation droplets were counted as colonies.

[0060]

[0073] Instead of imaging the plate from the top cover, the system described herein was designed to image the plate from the bottom (by flipping the plate over). Placing a diffuser near the plate lid (ideally in direct contact with or very close to the agar plate lid) is preferred to allow for evenly distributed illumination of the sample and to help eliminate condensation effects. A 3 mm thick white diffuser with 45% transmittance was selected, although other diffusers can also be used. The photoelectric array detector 102 (fixed focal length lens) was positioned approximately 100 mm to approximately 180 mm (preferably approximately 160 mm) from the agar plate 106 (see 114 in Figure 1D). The illumination source 118 was positioned approximately 15 mm to approximately 20 mm (preferably approximately 16 mm) below the agar plate 106 (see 116 in Figure 1D). Images collected using this system through the bottom surface of an agar plate (see Figure 9A) showed significantly reduced obscuring effects from condensation relative to images taken through the top surface of the plate with condensation (see Figure 9B).

[0061]

[0074] A further embodiment was designed in which illumination sources illuminate the sides of the agar plate and the agar. Multiple illumination sources were designed to surround the sides 220 of the agar plate 106, between the top cover 110 of the agar plate 106 and the agar surface 117 of the agar plate 106. The multiple illumination sources were preferably in the form of a series 240 of LED lights. A photoelectric array detector 102 (fixed focal length lens) was positioned approximately 100 mm to approximately 180 mm (preferably approximately 160 mm) from the agar plate 106 (see 206 in FIG. 2A). The illumination sources were preferably positioned such that the angle of the light source relative to the horizontal plane of the agar plate lid was within approximately 0°, 1°, 2°, 3°, 4°, 5°, 6°, or 7°, or even 10°, and were preferably positioned within approximately 3 mm of the vertical orientation of the top surface of the agar plate or within 9 mm of the bottom surface of the agar plate (see 208 in FIG. 2A). In some embodiments, the angle of the light relative to the horizontal plane of the lid is 2° or 5°. Figures 10A-10B show various images collected using the system described herein with the illumination source illuminating the side of the agar plate and the agar. As can be seen, with this system design, colonies growing on a membrane within the agar plate can also be clearly seen.

[0062]

[0075] To further improve image contrast for better colony identification, apply a contrast-limited adaptive histogram equalization (CLAHE) algorithm to the images. Feasibility studies and machine learning methods

[0063]

[0076] To support the initial feasibility study, microbial culture plates were prepared for imaging. Water bioburden samples, primarily containing Gram-negative bacteria, were filtered through a 0.45 μm white gridded (MCE) filter membrane, cultured in R2A medium, and incubated for 3–5 days.

[0064]

[0077] For this initial feasibility study, the following five organisms were selected:

[0065]

[0078] Burkholderia cepacia ATCC 25416

[0066]

[0079] Pseudomonas aeruginosa ATCC 9027

[0067]

[0080] Bacillus subtilis ATCC 6633

[0068]

[0081] Aspergillus brasiliensis ATCC 16404

[0069]

[0082] Staphylococcus aureus ATCC 6538 was selected.

[0070]

[0083] Culture plates included tryptone soy agar (TSA), Reasoner's 2A (R2A), and Sabouraud dextrose agar (SDA). Filter membranes included MicroFunnel Filter Funnels (MCE) with 0.45 μm white gridded membrane Supor, and MicroFunnel filter funnels with 0.2 μm white gridded and plain white (non-gridded) filter membrane Supor.

[0071]

[0084] All stock microorganisms were from EZ-Accu Shot™. Lyophilized cultures (Microbiologics) were prepared according to the manufacturer's instructions. Based on labeling and colony-forming unit (CFU) counts per microorganism, the volume added to each sample was varied to demonstrate a range of colony growth between 0 and 50 CFU per PRD requirement. Negative control plates using filter membranes were prepared using sterile water only for baseline calculations for the machine learning algorithm.

[0072]

[0085] All samples were incubated at 30°C, and plates were re-examined after 48 hours, 72 hours, and 5 days of incubation. Multiple days of imaging were used to generate larger sample sets for machine learning evaluation. This provided an opportunity to capture images of microorganisms with various colony sizes and morphologies. A total of 213 images were taken using the system described herein. Samples contained varying amounts of condensation, from no condensation.

[0073]

[0086] Once the images were collected, they were annotated using Make Sense AI (makesence.ai) so that the data generated for each image could be input for machine learning deployment.

[0074]

[0087] Images were imported and bounding boxes 602 were manually created around each colony 103 (region of interest), see Figure 6. Labels for each microorganism tested were created as a further step to improve the detection algorithm. Once annotation was complete, the data was exported as an XML file.

[0075]

[0088] Manual counting of colonies from images compared to compendial (standard) methods demonstrated that colonies can be accurately counted using the system described herein. [Table 1]

[0076] Overview of machine learning concepts and initial results

[0089] A machine learning algorithm was deployed to detect / count the microbial colonies present on the agar plates by achieving a minimum precision of 95% with maximum recall to avoid non-detection of any bacterial colonies. Hardware and algorithms used

[0077]

[0090] The computer used to process the algorithm was an Intel i7-10700 CPU with 32 GB of RAM and equipped with an RTX 3070 Nvidia GPU with 8 GB of memory. We chose a general-purpose artificial intelligence architecture: an instance segmentation architecture configured for use with custom datasets.

[0078] Results of deploying machine learning algorithms Databases using the systems described herein

[0091] Figure 11 shows an example of the results of the algorithm applied using an image from the system described herein. Table 2 below shows the results of the algorithm application. [Table 2]

[0079]

[0092] The results obtained with the system described herein support the feasibility and demonstrate the high image quality and contrast achieved by the imaging system.

[0080]

[0093] Embodiment 1 is a system for detecting colonies growing on an agar plate, the system comprising: a photoelectric array detector having associated optics for detecting a detection area on the surface of the agar plate; one or more illumination sources for illuminating the detection area, the illumination sources being positioned to illuminate the agar plate; a focusing lens system positioned between the photoelectric array detector and the agar plate; and a computer programmed to receive data collected by the photoelectric array detector, the data being a digital representation of the detection area, wherein colonies growing on the agar in the plate are detected through the bottom surface of the agar plate and the agar, and the computer analyzes the data to quantify the number of colonies in the detection area.

[0081]

[0094] Embodiment 2 includes the system of embodiment 1, in which two illumination sources are used to illuminate the detection area, with the first illumination source illuminating the side of the agar plate and the agar.

[0082]

[0095] Embodiment 3 includes the system of embodiment 2, wherein the second illumination source illuminates the top surface of the agar.

[0083]

[0096] Embodiment 4 includes the system of embodiment 3, further comprising a light diffuser disposed between the second illumination source and the agar plate.

[0084]

[0097] Example 5 includes the system of any one of Examples 2-4, wherein the first illumination source includes multiple illumination sources surrounding the sides of the agar plate.

[0085]

[0098] Embodiment 6 includes the system of embodiment 5, wherein the plurality of illumination sources surround the sides of the agar plate at locations between the top cover of the agar plate and the top surface of the agar.

[0086]

[0099] Embodiment 7 includes the system of any one of embodiments 1-6, wherein the colonies are grown on an absorbent pad or membrane surface in the agar within the agar plate.

[0087]

[0100] Example 8 includes the system of any one of Examples 1-7, wherein the illumination source includes one or more lasers.

[0088]

[0101] Example 9 includes the system of any one of Examples 1-7, wherein the illumination source includes one or more light emitting diodes.

[0089]

[0102] Example 10 includes the system of any one of Examples 1-9, wherein the photoelectric array detector includes a charge-coupled device (CCD) detector, a photomultiplier tube detector, a complementary metal-oxide semiconductor (CMOS) detector, or a photodiode detector.

[0090]

[0103] Example 11 includes the system of Example 10, wherein the CMOS detector is a CMOS camera.

[0091]

[0104] Example 12 includes the system of any one of Examples 1-11, further comprising a stage for positioning the agar plate relative to the illumination source and the photosensitive array detector.

[0092]

[0105] Embodiment 13 includes the system of embodiment 12, further comprising a robotic arm for placing the agar plate on the stage and removing the agar plate from the stage.

[0093]

[0106] Embodiment 14 includes the system of any one of embodiments 1-13, wherein the colonies are colonies of bacterial cells or colonies of fungal cells.

[0094]

[0107] Example 15 includes the system of any one of Examples 1-14, wherein a top cover of the agar plate is present during illumination and detection.

[0095]

[0108] Example 16 includes the system of any one of Examples 1 to 15, wherein the detection area is the bottom surface of the agar plate.

[0096]

[0109] Embodiment 17 is a system for detecting colonies growing on an agar plate, the system comprising: a photoelectric array detector having associated optics for detecting a detection area on the underside of the agar plate; a plurality of illumination sources surrounding the sides of the agar plate for illuminating the detection area; a focusing lens system positioned between the photoelectric array detector and the agar plate; and a computer programmed to receive data collected by the photoelectric array detector, the data being a digital representation of the detection area; colonies growing on the agar in the plate being detected through the bottom surface of the agar plate and the agar; and the computer analyzing the data to quantify the number of colonies in the detection area.

[0097]

[0110] Embodiment 18 includes the system of embodiment 17, further comprising a second illumination source that illuminates the top surface of the agar plate and the agar.

[0098]

[0111] Embodiment 19 includes the system of embodiment 18, further comprising a light diffuser disposed between the second illumination source and the agar plate.

[0099]

[0112] Embodiment 20 includes the system of any one of embodiments 17-19, wherein the multiple illumination sources surround the sides of the agar plate at locations between the top cover of the agar plate and the top surface of the agar.

[0100]

[0113] Embodiment 21 includes the system of any one of embodiments 17-20, wherein the colonies are grown on an absorbent pad or membrane surface in the agar within the agar plate.

[0101]

[0114] Example 22 includes the system of any one of Examples 17-21, wherein the illumination source includes one or more lasers.

[0102]

[0115] Example 23 includes the system of any one of Examples 17-21, wherein the illumination source includes one or more light emitting diodes.

[0103]

[0116] Example 24 includes the system of any one of Examples 17 to 23, wherein the photoelectric array detector includes a charge-coupled device (CCD) detector, a photomultiplier tube detector, a complementary metal-oxide semiconductor (CMOS) detector, or a photodiode detector.

[0104]

[0117] Embodiment 25 includes the system of embodiment 24, wherein the CMOS detector is a CMOS camera.

[0105]

[0118] Example 26 includes the system of any one of Examples 17 to 25, further comprising a stage for positioning the agar plate relative to the illumination source and the photosensitive array detector.

[0106]

[0119] Embodiment 27 includes the system of embodiment 26, further comprising a robotic arm for placing the agar plate on the stage and removing the agar plate from the stage.

[0107]

[0120] Embodiment 28 includes the system of any one of embodiments 17-27, wherein the colonies are colonies of bacterial cells or colonies of fungal cells.

[0108]

[0121] Embodiment 29 includes the system of any one of embodiments 17-28, wherein the top cover of the agar plate is present during illumination and detection.

[0109]

[0122] Embodiment 30 includes the system of any one of embodiments 17-29, wherein the detection area is the underside of the agar plate.

[0110]

[0123] Embodiment 31 is a method for quantifying the number of colonies growing on an agar plate, the method comprising the steps of illuminating a detection area on the underside of the agar plate with one or more illumination sources, the illumination sources being positioned to illuminate the surface of the agar plate; detecting the detection area with a photoelectric array detector having associated optics, the detecting being performed through a focusing lens system positioned between the photoelectric array detector and the agar plate; receiving data collected by the photoelectric array detector as a digital representation of the detection area; and analyzing the collected data to quantify the number of colonies in the detection area.

[0111]

[0124] Example 32 includes the method of example 31, in which two illumination sources are used to illuminate the detection area, with a first illumination source illuminating the side of the agar plate and the agar.

[0112]

[0125] Embodiment 33 includes the method of embodiment 32, wherein a second illumination source illuminates the top surface of the agar plate and the agar.

[0113]

[0126] Example 34 includes the method of any one of Examples 31 to 33, further comprising a light diffuser disposed between the second illumination source and the agar plate.

[0114]

[0127] Example 35 includes the method of any one of Examples 31 to 34, wherein the first illumination source includes multiple illumination sources surrounding the sides of the agar plate.

[0115]

[0128] Example 36 includes the method of example 35, wherein the multiple illumination sources surround the sides of the agar plate at locations between the top cover of the agar plate and the top surface of the agar.

[0116]

[0129] Embodiment 37 includes the method of any one of embodiments 31 to 36, wherein the colonies are grown on an absorbent pad or membrane surface in the agar in the agar plate.

[0117]

[0130] Example 38 includes the method of any one of Examples 31 to 37, wherein the illumination source includes one or more lasers.

[0118]

[0131] Example 39 includes the method of any one of Examples 31 to 37, wherein the illumination source includes one or more light emitting diodes.

[0119]

[0132] Embodiment 40 includes the method of any one of embodiments 31 to 39, wherein the photoelectric array detector includes a charge-coupled device (CCD) detector, a photomultiplier tube detector, a complementary metal-oxide semiconductor (CMOS) detector, or a photodiode detector.

[0120]

[0133] Embodiment 41 includes the method of embodiment 40, in which the CMOS detector is a CMOS camera.

[0121]

[0134] Example 42 includes the method of any one of Examples 31 to 41, further including placing the agar plate on a stage relative to the illumination source and the photosensitive array detector.

[0122]

[0135] Embodiment 43 includes the method of embodiment 42, further including placing the agar plate on the stage and removing the agar plate from the stage by a robotic arm.

[0123]

[0136] Embodiment 44 includes the method of any one of embodiments 31 to 43, wherein the colonies are colonies of bacterial cells or colonies of fungal cells.

[0124]

[0137] Embodiment 45 includes the method of any one of embodiments 31 to 44, wherein the top cover of the agar plate is present during illumination and detection.

[0125]

[0138] Embodiment 46 includes the method of any one of embodiments 34 to 45, wherein the detection area is the underside of the agar plate.

[0126]

[0139] Embodiment 47 is a method of training a neural network, the method including the steps of acquiring test images of colonies growing on an agar plate, the acquiring step including illuminating a detection area on the surface of the agar plate with one or more illumination sources, the illumination sources being positioned to illuminate the agar plate; detecting the detection area with a photosensitive array detector having associated optics, the detecting being performed via a focusing lens system positioned between the photosensitive array detector and the agar plate; receiving data collected by the photosensitive array detector as a digital representation of the detection area; and generating an image of cells within the detection area; providing a training image set having a fixed number of colonies growing on the agar plate; comparing the test images with the training image set by a computing system to generate an indication of the number of colonies in the test images; and repeating the acquiring, providing, and comparing steps with a plurality of additional test images to train the neural network.

[0127]

[0140] Example 48 includes the method of example 47, in which the neural network is a convolutional neural network or a corner neural network.

[0128]

[0141] Embodiment 49 is a method for quantifying the number of colonies growing on an agar plate, the method comprising: illuminating a detection area on the surface of the agar plate with one or more illumination sources, the illumination sources positioned to illuminate the surface of the agar plate; detecting the detection area with a photosensitive array detector having associated optics, the detecting being performed through a focusing lens system positioned between the photosensitive array detector and the agar plate; receiving data collected by the photosensitive array detector as a digital representation of the detection area; generating images of colonies in the detection area; applying, by a computer system, a trained neural network to the images of colonies to provide a quantitative determination of the number of colonies in the detection area, the trained neural network having been trained with a training image set generated from images of the surface of the agar plate and the detection area detected through the agar of the plate; and displaying the quantitative determination of the number of colonies in the detection area.

[0129]

[0142] Example 50 includes the method of example 49, in which the neural network is a convolutional neural network or a corner neural network.

[0130]

[0143] Embodiment 51 includes the method of embodiment 49 or embodiment 50, in which two illumination sources are used to illuminate the detection area, with the first illumination source illuminating the side of the agar plate and the agar.

[0131]

[0144] Embodiment 52 includes the method of embodiment 51, wherein a second illumination source illuminates the top surface of the agar plate and the agar.

[0132]

[0145] Embodiment 53 includes the method of embodiment 52, further comprising a light diffuser disposed between the second illumination source and the agar plate.

[0133]

[0146] Embodiment 54 includes the method of any one of embodiments 49 to 53, wherein the first illumination source includes multiple illumination sources surrounding the sides of the agar plate.

[0134]

[0147] Embodiment 55 includes the method of embodiment 54, wherein the multiple illumination sources surround the sides of the agar plate at locations between the top cover of the agar plate and the top surface of the agar.

[0135]

[0148] Embodiment 56 includes the method of any one of embodiments 49 to 55, wherein the colonies are grown on an absorbent pad or membrane surface in the agar in the agar plate.

[0136]

[0149] Embodiment 57 includes the method of any one of embodiments 49 to 56, wherein the illumination source includes one or more lasers.

[0137]

[0150] Embodiment 58 includes the method of any one of embodiments 49 to 56, wherein the illumination source includes one or more light emitting diodes.

[0138]

[0151] Embodiment 59 includes the method of any one of embodiments 49 to 58, wherein the photoelectric array detector includes a charge-coupled device (CCD) detector, a photomultiplier tube detector, a complementary metal-oxide semiconductor (CMOS) detector, or a photodiode detector.

[0139]

[0152] Embodiment 60 includes the method of embodiment 59, in which the CMOS detector is a CMOS camera.

[0140]

[0153] Example 61 includes the method of any one of Examples 49 to 60, further including placing the agar plate on a stage relative to the illumination source and the photosensitive array detector.

[0141]

[0154] Embodiment 62 includes the method of embodiment 61, further including placing the agar plate on the stage and removing the agar plate from the stage by a robotic arm.

[0142]

[0155] Embodiment 63 includes the method of any one of embodiments 49 to 62, wherein the colonies are colonies of bacterial cells or colonies of fungal cells.

[0143]

[0156] Embodiment 64 includes the method of any one of embodiments 49 to 63, wherein the top cover of the agar plate is present during illumination and detection.

[0144]

[0157] Embodiment 65 includes the method of any one of embodiments 49 to 64, wherein the detection area is the underside of the agar plate.

[0145]

[0158] Although specific embodiments have been shown and described herein, it is to be understood that the claims are not limited to the specific forms or arrangements of parts described and shown. Although exemplary embodiments have been disclosed herein and specific terms are employed, they are used in a generic and descriptive sense only and not for purposes of limitation. Modifications and variations of the embodiments are possible in light of the above teachings. It is therefore to be understood that the embodiments may be practiced otherwise than as specifically described.

[0146]

[0159] All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference.

Claims

1. 1. A system for detecting colonies growing on an agar plate, the system comprising: (a) a photoelectric array detector having associated optics for detecting a detection area on the surface of said agar plate; (b) one or more illumination sources for illuminating the detection region, the one or more illumination sources being positioned to illuminate the agar plate; (c) a focusing lens system disposed between the photoelectric array detector and the agar plate; (d) a computer programmed to receive data collected by the photosensitive array detector; and Equipped with The data is a digital representation of the detection area, the colonies growing on the agar in the plate are detected through the bottom of the agar plate and the agar, and the computer analyzes the data to quantify the number of colonies in the detection area.

2. The system of claim 1 , wherein two illumination sources are used to illuminate the detection area, a first illumination source illuminating the side of the agar plate and the agar.

3. The system of claim 2 , wherein a second illumination source illuminates the top surface of the agar.

4. The system of claim 3 , further comprising a light diffuser disposed between the second illumination source and the agar plate.

5. The system of any one of claims 2 to 4, wherein the first illumination source comprises a plurality of illumination sources surrounding the sides of the agar plate.

6. The system of claim 5 , wherein the plurality of illumination sources surround the sides of the agar plate at locations between a top cover of the agar plate and a top surface of the agar.

7. The system of any one of claims 1 to 6, wherein the colonies are growing on an absorbent pad or membrane surface in the agar within the agar plate.

8. The system of any one of claims 1 to 7, wherein the illumination source comprises one or more lasers.

9. The system of any one of claims 1 to 7, wherein the illumination source comprises one or more light emitting diodes.

10. 10. The system of any one of claims 1 to 9, wherein the photosensitive array detector comprises a charge-coupled device (CCD) detector, a photomultiplier tube detector, a complementary metal-oxide semiconductor (CMOS) detector, or a photodiode detector.

11. The system of claim 10 , wherein the CMOS detector is a CMOS camera.

12. a stage for positioning the agar plate relative to the illumination source and the photosensitive array detector; The system of any one of claims 1 to 11, further comprising:

13. a robotic arm for placing the agar plate on the stage and removing the agar plate from the stage; The system of claim 12 further comprising:

14. The system according to any one of claims 1 to 13, wherein the colonies are colonies of bacterial cells or colonies of fungal cells.

15. The system of any one of claims 1 to 14, wherein a top cover of the agar plate is present during said illumination and detection.

16. The system according to any one of claims 1 to 15, wherein the detection area is the bottom surface of the agar plate.

17. 1. A system for detecting colonies growing on an agar plate, the system comprising: (a) a photoelectric array detector having associated optics for detecting a detection area on the underside of said agar plate; (b) a plurality of illumination sources surrounding the sides of the agar plate for illuminating the detection zone; (c) a focusing lens system disposed between the photoelectric array detector and the agar plate; (d) a computer programmed to receive data collected by the photosensitive array detector; and Equipped with The data is a digital representation of the detection area, the colonies growing on the agar in the plate are detected through the bottom of the agar plate and the agar, and the computer analyzes the data to quantify the number of colonies in the detection area.

18. a second illumination source for illuminating the top surface of the agar plate and the agar; 20. The system of claim 17, further comprising:

19. a light diffuser disposed between the second illumination source and the agar plate; 20. The system of claim 18, further comprising:

20. 20. The system of any one of claims 17 to 19, wherein the plurality of illumination sources surround the sides of the agar plate at a location between a top cover of the agar plate and the top surface of the agar.

21. 21. The system of any one of claims 17 to 20, wherein the colonies are growing on an absorbent pad or membrane surface in the agar within the agar plate.

22. The system of any one of claims 17 to 21, wherein the illumination source comprises one or more lasers.

23. The system of any one of claims 17 to 21, wherein the illumination source comprises one or more light emitting diodes.

24. 24. The system of any one of claims 17 to 23, wherein the photosensitive array detector comprises a charge-coupled device (CCD) detector, a photomultiplier tube detector, a complementary metal-oxide semiconductor (CMOS) detector, or a photodiode detector.

25. 25. The system of claim 24, wherein the CMOS detector is a CMOS camera.

26. a stage for positioning the agar plate relative to the illumination source and the photosensitive array detector; The system of any one of claims 17 to 25, further comprising:

27. a robotic arm for placing the agar plate on the stage and removing the agar plate from the stage; 27. The system of claim 26, further comprising:

28. The system according to any one of claims 17 to 27, wherein the colonies are colonies of bacterial cells or colonies of fungal cells.

29. 29. The system of any one of claims 17 to 28, wherein a top cover of the agar plate is present during the illumination and detection.

30. The system of any one of claims 17 to 29, wherein the detection area is the backside of the agar plate.

31. 1. A method for quantifying the number of colonies growing on an agar plate, the method comprising: (a) illuminating a detection area on a backside of the agar plate with one or more illumination sources, the illumination sources being positioned to illuminate a front side of the agar plate; (b) detecting the detection area with a photoelectric array detector having associated optics through a collection lens system disposed between the photoelectric array detector and the agar plate; (c) receiving data collected by the photosensitive array detector as a digital representation of the detection area; (d) analyzing the collected data to quantify the number of colonies in the detection area; A method comprising:

32. 32. The method of claim 31, wherein two illumination sources are used to illuminate the detection area, a first illumination source illuminating the agar plate and the side of the agar.

33. 33. The method of claim 32, wherein a second illumination source illuminates the top surface of the agar plate and the agar.

34. 34. The method of any one of claims 31 to 33, further comprising a light diffuser positioned between the second illumination source and the agar plate.

35. 35. The method of any one of claims 31 to 34, wherein the first illumination source comprises a plurality of illumination sources surrounding the side of the agar plate.

36. 36. The method of claim 35, wherein the plurality of illumination sources surround the sides of the agar plate at a location between a top cover of the agar plate and a top surface of the agar.

37. 37. The method of any one of claims 31 to 36, wherein the colonies are grown on an absorbent pad or membrane surface in the agar within the agar plate.

38. The method of any one of claims 31 to 37, wherein the illumination source comprises one or more lasers.

39. The method of any one of claims 31 to 37, wherein the illumination source comprises one or more light emitting diodes.

40. 40. The method of any one of claims 31 to 39, wherein the photosensitive array detector comprises a charge coupled device (CCD) detector, a photomultiplier tube detector, a complementary metal oxide semiconductor (CMOS) detector, or a photodiode detector.

41. 41. The method of claim 40, wherein the CMOS detector is a CMOS camera.

42. placing the agar plate on a stage relative to the illumination source and the photosensitive array detector; The method of any one of claims 31 to 41, further comprising:

43. placing the agar plate on the stage and removing the agar plate from the stage by a robotic arm; 43. The method of claim 42, further comprising:

44. 44. The method of any one of claims 31 to 43, wherein the colonies are colonies of bacterial cells or colonies of fungal cells.

45. 45. The method of any one of claims 31 to 44, wherein a top cover of the agar plate is present during said illumination and detection.

46. 46. ​​The method of any one of claims 34 to 45, wherein the detection area is the backside of the agar plate.

47. 1. A method for training a neural network, the method comprising: (a) acquiring test images of colonies growing on an agar plate, i. illuminating a detection area on a surface of the agar plate with one or more illumination sources, the illumination sources positioned to illuminate the agar plate; ii. detecting the detection area with a photoelectric array detector having associated optics through a collection lens system disposed between the photoelectric array detector and the agar plate; iii. receiving data collected by the photosensitive array detector as a digital representation of the detection area; iv. generating an image of the cells within the detection area; obtaining the data; (b) providing a training image set having a quantified number of colonies growing on said agar plates; (c) comparing, with a computing system, the test image with the set of training images to generate an indication of the number of colonies in the test image; (d) repeating the acquiring, providing, and comparing steps with a plurality of additional test images to train the neural network; A method comprising:

48. 48. The method of claim 47, wherein the neural network is a convolutional neural network or a corner neural network.

49. 1. A method for quantifying the number of colonies growing on an agar plate, the method comprising: (a) illuminating a detection area on a surface of the agar plate with one or more illumination sources, the illumination sources being positioned to illuminate the surface of the agar plate; (b) detecting the detection area with a photoelectric array detector having associated optics through a collection lens system disposed between the photoelectric array detector and the agar plate; (c) receiving data collected by the photosensitive array detector as a digital representation of the detection area; (d) generating an image of the colonies within the detection area; (e) applying, by a computer system, a trained neural network to the images of the colonies to provide a quantitative determination of the number of colonies in the detection area, the trained neural network having been trained with a training image set generated from images of the agar plate and the detection area in which colonies growing on agar in the plate are detected through the surface of the agar; (f) displaying a quantitative determination of the number of colonies within the detection area; A method comprising:

50. 50. The method of claim 49, wherein the neural network is a convolutional neural network or a corner neural network.

51. 51. The method of claim 49 or claim 50, wherein two illumination sources are used to illuminate the detection area, a first illumination source illuminating the side of the agar plate and the agar.

52. 52. The method of claim 51, wherein a second illumination source illuminates the top surface of the agar plate and the agar.

53. 53. The method of claim 52, further comprising a light diffuser disposed between the second illumination source and the agar plate.

54. 54. The method of any one of claims 49 to 53, wherein the first illumination source comprises a plurality of illumination sources surrounding the side of the agar plate.

55. 55. The method of claim 54, wherein the plurality of illumination sources surround the sides of the agar plate at a location between a top cover of the agar plate and the top surface of the agar.

56. 56. The method of any one of claims 49 to 55, wherein the colonies are grown on an absorbent pad or membrane surface in the agar within the agar plate.

57. 57. The method of any one of claims 49 to 56, wherein the illumination source comprises one or more lasers.

58. 57. The method of any one of claims 49 to 56, wherein the illumination source comprises one or more light emitting diodes.

59. 59. The method of any one of claims 49 to 58, wherein the photosensitive array detector comprises a charge coupled device (CCD) detector, a photomultiplier tube detector, a complementary metal oxide semiconductor (CMOS) detector, or a photodiode detector.

60. 60. The method of claim 59, wherein the CMOS detector is a CMOS camera.

61. placing the agar plate on a stage relative to the illumination source and the photosensitive array detector; 61. The method of any one of claims 49 to 60, further comprising:

62. placing the agar plate on the stage and removing the agar plate from the stage by a robotic arm; 62. The method of claim 61, further comprising:

63. 63. The method of any one of claims 49 to 62, wherein the colonies are colonies of bacterial cells or colonies of fungal cells.

64. 64. The method of any one of claims 49 to 63, wherein a top cover of the agar plate is present during said illuminating and detecting.

65. 65. The method of any one of claims 49 to 64, wherein the detection area is the backside of the agar plate.

Images