JP2020054247A - Colony detection device, computer program, colony detection method, identifier generation method and identifier - Google Patents

Abstract

To provide a colony detection device, a computer program, a colony detection method, an identifier generation method and an identifier which are capable of improving detection accuracy of colonies.SOLUTION: A colony detection device comprises: a culture medium image acquiring part which acquires culture medium image data of culture medium on which microorganisms are cultured; a specifying part which inputs acquired culture medium image data to an identifier generated by using learning data including culture medium images and culture medium types, and specifies a detector for detecting colonies on the basis of the culture medium types identified by the identifier; and a detecting part which detects colonies based on the acquired culture medium image data by using the specified detector.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a colony detection device, a computer program, a colony detection method, a classifier generation method, and a classifier.

  In recent years, with the growth of the fast food and other food service industries and the increase of easily processed foods, opportunities for others to eat cooked foods have been significantly increased. On the other hand, if dangerous substances such as toxic substances or harmful microorganisms are mixed in foods such as processed foods, they directly affect the human body. Therefore, when handling processed foods, not only the safety of raw materials but also the hygiene of the manufacturing process is strictly controlled, and the safety of foods throughout the production and processing of foods is required.

  Patent Document 1 discloses a colony detection system for culturing microorganisms collected from food on a Petri dish or film and counting colonies.

JP 2016-15944 A

  However, microorganisms contained in the food, for example, general bacteria, Escherichia coli, Escherichia coli, various bacteria such as Staphylococcus aureus, and also, the type of sample or the type or dilution rate of the diluent, culture conditions, etc. Complicated operations such as selection of a detection logic for detecting a colony and specification of a detection processing parameter are required in accordance with the various conditions. In particular, the specification of the detection processing parameter is an important operation for improving the detection accuracy of the colony.

  The present invention has been made in view of such circumstances, and has as its object to provide a colony detection device, a computer program, a colony detection method, a discriminator generation method, and a discriminator capable of improving the detection accuracy of colonies. And

  A colony detection device according to an embodiment of the present invention includes a medium image acquiring unit that acquires medium image data of a medium in which microorganisms are cultured, and a medium image data acquired by the medium image acquiring unit. Input to a discriminator generated using learning data including, a specifying unit for specifying a detector for detecting a colony based on a culture medium type identified by the discriminator, and a detector specified by the specifying unit And a detecting unit for detecting a colony based on the culture medium image data acquired by the culture medium image acquiring unit using the method.

  The colony detection device according to the embodiment of the present invention is a medium image acquisition unit that acquires medium image data of a medium in which microorganisms are cultured, and medium image data acquired by the medium image acquisition unit, a medium image, a colony. Input to the discriminator generated by using the learning data including the detector data of the detector for detection and the detection processing parameter data, and generated based on the detector information and the detection processing parameter information specified by the discriminator. And a detection unit for detecting a colony based on the culture medium image data acquired by the culture medium image acquisition unit using a detector.

  A computer program according to an embodiment of the present invention uses a computer to perform processing for acquiring medium image data of a medium in which microorganisms are cultured, and to use the acquired medium image data, learning data including medium image data and medium type. Input to the generated learning model, processing for specifying a detection algorithm for detecting a colony based on the medium type identified by the learning model, and a colony based on the medium image data obtained using the specified detection algorithm And a process for detecting

  The colony detection method according to the embodiment of the present invention obtains medium image data of a medium in which microorganisms are cultured, and generates the obtained medium image data using learning data including the medium image data and the medium type. The detector is input to the identified discriminator, and a detector for detecting a colony is specified based on the medium type identified by the discriminator, and a colony based on the culture medium image data acquired using the specified detector is detected.

  The discriminator generation method according to the embodiment of the present invention obtains medium image data of a medium in which microorganisms are cultured, obtains a medium type of the medium image data, and includes the obtained medium image data and medium type. A classifier is generated using the learning data.

  The discriminator according to the embodiment of the present invention is generated using the culture medium image data of the culture medium in which the microorganism is cultured and the learning data including the culture medium type of the culture medium image data.

  ADVANTAGE OF THE INVENTION According to this invention, the detection accuracy of a colony can be improved.

It is a block diagram showing an example of composition of a colony detection device of this embodiment. It is a schematic diagram which shows an example of a film type culture medium. It is a schematic diagram which shows an example of a structure of a learning model. It is a schematic diagram which shows the process performed in a convolutional layer. It is a schematic diagram which shows the process performed in a pooling layer. It is a schematic diagram which shows the 1st example of the learning method of the learning model by the learning processing part. It is a schematic diagram which shows the 1st example of the colony detection method by a processing part. It is a schematic diagram which shows an example of a colony detection result. It is a schematic diagram which shows the 2nd example of the learning method of the learning model by a learning processing part. It is a schematic diagram which shows the 2nd example of the colony detection method by a processing part. It is a schematic diagram which shows the 3rd example of the colony detection method by a processing part. It is a schematic diagram which shows an example of the learning method of the learning model by the learning server. It is a flowchart which shows an example of the learning processing procedure of the learning model by the learning processing part. It is a flowchart which shows an example of the colony detection processing procedure by a colony detection apparatus.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram illustrating an example of the configuration of the colony detection device 50 according to the present embodiment. One or more (one in the example of FIG. 1) scanner 10 as an optical reader is connected to the colony detection device 50 via a network 1 such as the Internet. The scanner 10 can read a film-type culture medium described below from a sanitary management system 100 such as a company or a company, and transmit culture medium image data obtained by the reading to the colony detection device 50.

  The sanitary management system 100, for example, in the manufacturing process or the inspection process to inspect the food during or after processing the food, to extract the food during or after processing as a sample, the sample liquid in the film-type medium described below (Specimen) is added dropwise, an appropriate sterile diluent is added, and the sample solution is diluted and cultured under required culture conditions. The film-type medium after culturing can be read by the scanner 10 as described above.

  FIG. 2 is a schematic diagram illustrating an example of the film-type medium 20. The film-type medium 20 includes a base sheet 21 formed of a film, a circular frame 22 formed in the center of the base sheet 21, a culture layer 23 for culturing bacteria provided in the frame 22, and a culture layer 23. And a cover sheet 24 that covers the base sheet 21 including the base sheet. FIG. 2 shows a film-type medium after culturing, and colonies can be visually observed. A colony refers to a colony of bacteria that are obtained by culturing bacteria that are so small as to be invisible and increasing the bacteria until they are visible. Colonies are counted to determine the number of bacteria.

  Although not shown, supplementary information can be described in a margin of the film type medium 20 excluding the culture layer 23. The incidental information can include, for example, the type of the specimen, the type or dilution ratio of the diluent, the culture conditions, and the like. The supplementary information may be printed characters or symbols, or may be handwritten characters or the like. The accompanying information can be read by the scanner 10.

  In addition, the medium type information can be described in a blank of the film type medium 20 excluding the culture layer 23. The medium type information includes, for example, the type of medium (medium type), and may be printed characters or symbols, or may be handwritten characters or the like. The medium type information can be read by the scanner 10.

  As shown in FIG. 1, the colony detection device 50 includes a control unit 51 that controls the entire device, a communication unit 52, a storage unit 53, an adjustment unit 54, a determination unit 55, an output processing unit 56, and a processing unit 60. The processing unit 60 includes a learning model 61 as a classifier, a learning processing unit 62, and a detection unit 63. The detection unit 63 includes a detection logic unit 631 as a detector and a detection processing parameter unit 632. When the learning model 61 has been learned, the learning processing unit 62 may not be provided, and is not an essential configuration.

  The control unit 51 can be configured by a CPU, a ROM, a RAM, and the like.

  The communication unit 52 has a function of communicating with the scanner 10 via the network 1, and can receive (acquire) information read by the scanner 10.

  More specifically, the communication unit 52 has a function as a medium image acquiring unit, and can acquire, from the scanner 10, medium image data of a medium in which microorganisms (bacteria) are cultured.

  The communication unit 52 has a function as an additional information acquisition unit, and can acquire additional information related to the acquired culture image data from the scanner 10. In addition, the communication unit 52 has a function as a type information obtaining unit, and can obtain the culture medium type information of the obtained culture medium image data from the scanner 10.

  The processing unit 60 includes, for example, hardware such as a CPU (for example, a multi-processor having a plurality of processor cores), a GPU (Graphics Processing Units), a DSP (Digital Signal Processors), and an FPGA (Field-Programmable Gate Arrays). Can be combined. Also, a quantum processor can be combined.

  The learning model 61 has a function as a discriminator, and can use a multilayer neural network (deep learning). For example, a convolutional neural network (Convolutional Neural Network) can be used. May be used.

  FIG. 3 is a schematic diagram illustrating an example of the configuration of the learning model 61. The learning model 61 shown in FIG. 3 is a convolutional neural network, in which an input layer 611, a convolutional layer 612, a pooling layer 613, a convolutional layer 614, a pooling layer 615, a fully connected layer 616, and an output layer 617 are connected in this order. ing. Note that the numbers of the convolutional layers, the pooling layers, and the total connection layers are for convenience and are not limited to the numbers shown in FIG. For convenience, the layer of the activation function is omitted. Medium image data is input to the input layer 611. The learning model 61 can output an identification result (a medium type, a detection processing parameter, and the like to be described later) for each medium image data obtained by reading one film-type medium 20 with the scanner 10.

  FIG. 4 is a schematic diagram showing a process performed in the convolutional layer. The input / output data of the convolutional layer is also called a feature map, the input data of the convolutional layer is called an input feature map, and the output data of the convolutional layer is called an output feature map. The input feature map of the first convolutional layer is an input frame-based image. The processing performed in the convolution layer (also referred to as “convolution operation”) is a filter operation using a convolution filter (also referred to as “filter”).

  As shown in FIG. 4, for convenience, the input feature map is 8 × 8 pixels. The size of the filter is 3 × 3 pixels. In the convolution operation, the filter element is multiplied by the corresponding element of the input feature map while sliding the window of the filter at a fixed interval with respect to the input feature map, and the sum is obtained. In the corresponding pixel of. In the example of FIG. 4, the calculation result of the area S corresponding to the filter F1 of the input feature map is stored in the pixel S1 of the output feature map. Further, the calculation result of the area S corresponding to the filter F2 of the input feature map is stored in the pixel S2 of the output feature map. Similarly, the calculation result of the area S corresponding to the filter F3 of the input feature map is stored in the pixel S3 of the output feature map. By performing the same operation by moving the filter one pixel at a time, the output feature map has a size of 6 × 6 pixels. Here, three output feature maps are obtained by using the three filters F1, F2, and F3.

  In the learning of the learning model 61, as parameters relating to the filter, for example, the value of the element of the filter, the number of filters (3 in the example of FIG. 4), the size of the filter (3 × 3 in the example of FIG. 4), the filter 4 (also referred to as a “slide”, one pixel in the example of FIG. 4), padding that fills the periphery (edge region) of the input feature map with 0, and the like. With the convolutional layer, spatial features of the image can be extracted.

  FIG. 5 is a schematic diagram illustrating a process performed in the pooling layer. The pooling layer performs a process of reducing the size of the two-dimensional feature map output from the convolution layer. More specifically, a process of consolidating the local region of the image into one element is performed. For example, as shown in FIG. 5, in a 6 × 6 pixel feature map (output feature map), 2 × 2 local regions (windows W) are aggregated into “4” which is the maximum value of each element. ing. Note that the sliding of the window W is equal to the size of the window W. In the example of FIG. 5, since the sliding is performed by two pixels, the feature map of 6 × 6 pixels is reduced to 3 × 3 pixels. By the pooling layer, for example, even if the characteristic portion is slightly deformed or displaced in the image, the difference due to the deformation or displacement can be absorbed to extract the characteristic portion.

  The learning processing unit 62 learns / updates (generates) the learning model 61 using the learning data. Details of the learning processing unit 62 will be described later.

  The detection unit 63 uses an optimum detection logic specified from a plurality of detection logics (also referred to as a detection algorithm and a detector) in the detection logic unit 631 according to the type of the culture medium, and a detection processing parameter unit 632. The colony is detected by performing image processing on the medium image data using the optimum detection processing parameter specified from among the detection processing parameters in the above. The detection of a colony includes, for example, colony position information (X coordinate and Y coordinate on an image), reliability of colony detection, and the like. The reliability means detection accuracy. For example, the reliability is higher as the difference or ratio between the color feature value for detecting a colony and the threshold value of the color feature value is larger.

  The detection processing parameters include, for example, a feature amount (color feature amount) related to the color when detecting a colony from the culture medium image, a threshold value of the color feature amount, a pass / fail threshold for performing the medium determination, a mask radius, and a detection logic. Including various parameters.

  Next, a method of generating (learning) the learning model 61 will be described.

  FIG. 6 is a schematic diagram illustrating a first example of a learning method of the learning model 61 by the learning processing unit 62. As illustrated in FIG. 6, the learning processing unit 62 gives learning input data and learning output data to the learning model 61 as learning data. The learning input data includes medium image data. The learning output data includes the type of culture medium (medium type), the detection processing parameters used by the detection unit 63, the position information of the colony output by the detection unit 63 based on the culture medium image data, the reliability of colony detection, and the like. Can be included. Note that the position information of the colony is not essential.

  The learning processing unit 62 can learn and update (generate) the learning model 61 using the medium image data and the learning data including the type of the medium. Examples of the type of medium include a medium for measuring the number of general viable bacteria (also referred to as AC), a medium for measuring the number of coliforms (also referred to as CC), a medium for measuring the number of coliforms and coliforms (also referred to as EC), and yellow. A medium for measuring the number of staphylococci (also referred to as SA) is included. Thereby, the learning model 61 can identify (output) the type of the culture medium based on the culture medium image data.

  In addition, the learning processing unit 62 can learn and update (generate) the learning model 61 using the culture image data, the detection processing parameters for detecting the colony, and the learning data including the colony detection result information. The detection result information can include colony position information and reliability of colony detection (for example, the probability of how correct the colony detection result is). For example, each time the medium image data is provided as learning data to the learning model 61 and the detection processing parameter is updated, the updated detection processing parameter and the colony detection result information obtained as a result of the update are provided to the learning model 61 as learning data. The learning model 61 can be learned. Accordingly, the learning model 61 can identify (output) a detection processing parameter necessary for setting the reliability of colony detection to a predetermined threshold or more.

  Next, a method for detecting a colony will be described.

  FIG. 7 is a schematic diagram illustrating a first example of a colony detection method by the processing unit 60. Since the learning model 61 is learned using the culture medium image data and the learning data including the type of the culture medium, for example, when the culture medium image data whose culture medium type is unknown is input to the learning model 61, the learning model 61 The type of medium can be identified based on the obtained medium image data. As the type of culture medium output (identified) by the learning model 61, for example, the probability can be output for each of the aforementioned AC, CC, EC, and SA. The type of medium with the highest probability can be used as the identification result.

  The detection unit 63 has a function as a specifying unit, and based on the type of the culture medium identified by the learning model 61, the type of the culture medium identified from among the detection logics A, B, C, and D included in the detection logic unit 631. The most suitable detection logic can be specified. For example, the detection logics A, B, C, and D can be specified according to the type of the medium, AC, CC, EC, or SA. Thereby, an optimal detection logic can be specified according to the type of the culture medium. Although FIG. 7 illustrates the detection logics A to D, the number of the detection logics is not limited to four. Even if the detection logic that combines at least two of the detection logics A, B, C, and D is specified according to the magnitude of the probability for each of AC, CC, EC, and SA that is output from the learning model 61. Good.

  The detection unit 63 can perform image processing on the medium image data using the specified detection logic to detect a colony.

  With the above-described configuration, simply by inputting the culture medium image data to the processing unit 60 (the learning model 61), the type of the culture medium is identified, the detection logic optimal for the identified culture medium type is specified, and the colony is detected. Therefore, complicated work such as selection of detection logic is not required, and the accuracy of colony detection can be improved.

  Further, since the learning model 61 is learned using learning data including detection processing parameters for detecting a colony and colony detection result information, for example, when medium image data is input to the learning model 61, the learning model 61 The reference numeral 61 can specify a detection processing parameter (detection processing parameter information) that is optimal for a detection logic for performing image processing on the culture medium image data.

  The detection unit 63 can detect a colony by performing image processing on the culture medium image data using the specified detection logic and a detection processing parameter optimal for the detection logic.

  With the above-described configuration, only by inputting the culture medium image data to the processing unit 60 (learning model 61), the optimum detection processing parameter can be specified for the detection logic and the colony can be detected. A complicated operation such as identification is not required, and the accuracy of colony detection can be improved.

  The output processing unit 56 outputs a colony detection result based on the type of medium (medium type) identified by the learning model 61, the position information of the colony output by the detection unit 63, the reliability of colony detection, and the like.

  FIG. 8 is a schematic diagram illustrating an example of a colony detection result. As shown in FIG. 8, in the culture layer, a rectangular mark indicated by a symbol M1 indicates a colony counted as one. If the colonies are multicolored, the number of colonies can be counted for each color. The mark indicated by the symbol M2 indicates a state where a plurality of colonies are connected. In the example of FIG. 8, the number of colonies is four. The mark indicated by the symbol M3 indicates the boundary of the colony. A boundary line is displayed at the boundary between the colony and the medium. Thereby, the boundary of the colony can be clearly determined. The mark indicated by the symbol M4 is a marker display for coloring which may be erroneously recognized as a colony, or a warning display for a sample in which erroneous detection is likely to occur. Thus, it is possible to warn the user not to misidentify the colony.

  The adjustment unit 54 can adjust the detection processing parameters specified by the learning model 61 based on the supplementary information acquired by the communication unit 52. The acquired culture medium image data is image data obtained as a result of culturing based on the type of the specimen and the type or dilution rate of the diluent, culture conditions, and the like. Can be considered to be implicitly included. Therefore, when the learning model 61 learns the medium image data as the learning data, it is considered that the learning data implicitly includes conditions such as the type of the sample, the type or dilution ratio of the diluent, and the culture condition. Can be

  However, by adjusting the detection processing parameters specified by the learning model 61 based on the incidental information, more optimal detection processing parameters can be obtained, and the detection accuracy of colonies can be further improved.

  The output processing unit 56 has a function as an output unit, and outputs a predetermined notification when the reliability of the type of culture medium identified by the learning model 61 is equal to or less than a predetermined threshold. The reliability is an index indicating how reliable the identification result of the learning model 61 is, and the threshold can be set to an appropriate value such as, for example, 98% or 95%. The predetermined notification includes, for example, a notification that the type of the culture medium identified by the learning model 61 is incorrect or that the detection of a colony is impossible. Thereby, if the colony cannot be detected, the fact is notified, so that erroneous colony detection can be prevented.

  The detection unit 63 has a function as a setting unit, and can set necessary detection logic for detecting a colony when the output processing unit 56 outputs a predetermined notification. For example, even when the type of the culture medium cannot be identified and the detection logic cannot be specified, the required detection logic can be set, so that the colony can be detected.

  When the detection unit 63 cannot detect a colony based on the acquired culture image data, the determination unit 55 determines the culture medium of the culture image data based on the culture type information (culture type) acquired by the communication unit 52. The type can be determined. Even when the learning model 61 cannot identify the type of the culture medium, the type of the culture medium can be determined based on the culture medium type information, so that the learning effort of the learning model 61 can be reduced.

  The control unit 51 can collect the culture medium image data input to the learning model 61 and the type of culture medium identified by the learning model 61 in the storage unit 53 as learning data. The learning processing unit 62 can update the learning model 61 using the collected learning data. The learning data to be collected can include data when the identification by the learning model 61 is wrong. Thereby, the learning model 61 can be re-learned, and the discriminating ability of the learning model 61 can be enhanced.

  FIG. 9 is a schematic diagram illustrating a second example of the learning method of the learning model 61 by the learning processing unit 62. The difference from the first example illustrated in FIG. 6 is that the detection logic is added as learning data. Specifically, detection logic information (detector information) for specifying each of a plurality of detection logics and detection processing parameter information for specifying each of various detection processing parameters can be used as learning data. Although the example of FIG. 9 shows a configuration in which the culture medium type is used as the learning data, the learning data need not include the culture medium type.

  The learning model 61 is generated using learning data including a culture medium image, detection logic information of a detector for detecting a colony, and detection processing parameter information. Thus, when the culture medium image data is input to the learning model 61, the learning model 61 can specify the detection logic information and the detection processing parameter information based on the input culture medium image data.

  FIG. 10 is a schematic diagram showing a second example of the colony detection method by the processing unit 60. As shown in FIG. 9 described above, the learning model 61 is generated using learning data including a culture medium image, detection logic information of a detector for detecting a colony, and detection processing parameter information. The information and the detection processing parameter information can be specified. For example, assuming that the detection logics are detection logics A, B, C, and D, the detection logic information indicates the optimality of each of the detection logics A, B, C, and D (for example, how suitable is the detection logic. ) Can be represented by a numerical value or a ratio. In addition, the detection processing parameter information can indicate, for example, the optimality in the case of using each of various detection processing parameters by a numerical value or a ratio.

  The detection unit 63 can generate a new detection logic 633 by using the detection logic information specified by the learning model 61 and combining the detection logics A, B, C, and D. As a result, it is not necessary to specify the culture medium type, and it is possible to collectively specify the detection logic and the detection processing parameters, so that it is possible to flexibly detect a colony even for a new type in the future. Become.

  FIG. 11 is a schematic diagram showing a third example of the colony detection method by the processing unit 60. For example, when culture medium image data of a culture medium in which an unknown specimen such as a specimen that has not been tested in the past is cultured is input to the learning model 61, the culture medium type identified by the learning model 61 and the culture medium type optimal for the culture medium type are input. It is assumed that the detection unit 63 cannot detect a colony even with a simple detection logic. In the example of FIG. 11, a state in which the learning model 61 cannot provide the detection unit 63 with, for example, optimal detection processing parameter information is schematically indicated by a symbol X.

  In such a case, the detection logic selection unit 634 can select (generate) a new detection logic based on the supplementary information related to the medium type and the medium image data. Thus, even when an unknown sample is tested, the detection logic is automatically adjusted, so that a colony can be detected.

  FIG. 12 is a schematic diagram illustrating an example of a learning model learning method performed by the learning server 200. The learning server 200 has the same configuration as the colony detection device 50 illustrated in FIG. The learning server 200 acquires learning information from the hygiene management system 100h of the H company, acquires learning information from the sanitation management system 100j of the J company, and acquires learning information from the sanitation management system 100k of the K company. . Although the example of FIG. 12 illustrates the sanitation management system of three companies, the acquisition source of the learning information is not limited to three companies.

  The culture medium information DB 210 records information such as a detection logic and a detection parameter. Further, in the culture medium information DB 210, acquired learning information (such as culture medium image data) can be recorded.

  The learning server 200 learns / updates (generates) the learning model using the learning information acquired from the sanitary management system 100h of the H company, and uses the learned / updated learning model as a learning model for the H company for the H company. Can be introduced into the colony detection device 50h. Similarly, the learning server 200 learns / updates (generates) a learning model using the learning information acquired from the sanitary management system 100j of J company, and uses the learned / updated learning model as a learning model for J company. It can be introduced into the colony detection device 50j for Company J. The learning server 200 learns / updates (generates) a learning model using the learning information acquired from the sanitary management system 100k of K company, and uses the learned / updated learning model as a learning model for K company. It can be introduced into a company-specific colony detection device 50k.

  As described above, the acquired culture medium image data is classified for each acquisition source. The acquisition source can be, for example, a company or a company. Different companies handle different samples and culture conditions. A learning model is generated for each acquisition source using learning data including the classified medium image data for each acquisition source. As a result, a learning model optimized for each company or company can be generated, and the discriminating ability of the learning model can be enhanced.

  FIG. 13 is a flowchart illustrating an example of a learning processing procedure of the learning model 61 by the learning processing unit 62. In the following, the processing subject will be described as a learning processing unit 62 for convenience. The learning processing unit 62 acquires the medium image data (S11), and acquires the type of medium (medium type) (S12). The learning processing unit 62 acquires the position information of the colony and the reliability of the colony detection from the detection unit 63 (S13), and acquires the detection processing parameters used by the detection unit 63 (S14).

  The learning processing unit 62 generates the acquired culture medium image data, the culture medium type, the detection processing parameters, the colony position information, and the teacher data based on the reliability of the colony detection (S15), and performs learning and updating of the learning model 61. (S16).

  The learning processing unit 62 determines whether or not to update the detection processing parameter (S17), and if so (YES in S17), updates the detection processing parameter (S18), and continues the processing from step S13. Thus, the learning model 61 can be learned by gradually changing the detection processing parameters.

  If the detection processing parameters are not updated (NO in S17), the learning processing unit 62 determines whether there is another culture medium image data for learning (S19), and if there is another culture medium image data (YES in S19), The processing after step S11 is continued. When there is no other culture medium image data (NO in S19), the learning processing unit 62 ends the processing.

  FIG. 14 is a flowchart illustrating an example of a colony detection processing procedure performed by the colony detection device 50. In the following, the main part of the processing will be described as the control unit 51 for convenience. The control unit 51 acquires medium image data (S31), and determines the type of medium (medium type) (S32). The control unit 51 specifies a detection processing parameter (S33), and specifies a detection logic (S34). The detection processing parameters are specified by the learning model 61, and the detection logic can be specified according to the type of the culture medium identified by the learning model 61.

  The control unit 51 performs image processing on the acquired culture medium image data using the specified detection logic and detection processing parameters to detect a colony (S35). The detection of the colony includes, for example, detecting the position information of the colony and obtaining the reliability of the colony detection.

  The control unit 51 determines whether or not the reliability of the colony detection is equal to or lower than a predetermined threshold (S36). If the reliability is not equal to or lower than the predetermined threshold (NO in S36), the colony detection result information (for example, FIG. 8 is generated and output (S37), and the process ends. When the reliability is equal to or less than the predetermined threshold (YES in S36), the control unit 51 outputs a notification that colony detection is impossible (S38), and ends the process.

  When the notification that the colony detection is impossible is output in step S38, the control unit 51 (the detection unit 63) may set a necessary detection logic for detecting the colony. This makes it possible to detect colonies

  The control unit 51 and the processing unit 60 of the present embodiment can also be realized using a general-purpose computer including a CPU (processor), a GPU, a RAM (memory), and the like. That is, as shown in FIGS. 13 and 14, a computer program defining the procedure of each processing is loaded into a RAM (memory) provided in the computer, and the computer program is executed by a CPU (processor). Thus, the control unit 51 and the processing unit 60 can be realized. The computer program may be recorded on a recording medium and distributed. The learning model 61 trained by the learning server 200 and the computer program based on the learning model 61 may be distributed to the colony detecting device 50 via the network 1 and installed.

  In the above-described embodiment, the configuration has been described in which the culture medium image data is acquired by reading the film-type culture medium 20 with a scanner. However, the present invention is not limited to this. The medium image data may be obtained by imaging with.

  In the above embodiment, when there is no colony, for example, the following processing can be performed. In the first process, when the culture medium type is not included in the supplementary information, for example, the user can manually input AC as 0 and EC as 0. In the second process, the medium type is recorded only when there is a colony, for example, 100 SAs can be recorded, and 0 can be recorded for other SAs.

  The colony detection device of the present embodiment includes a culture medium image acquisition unit that acquires culture medium image data of a culture medium in which microorganisms are cultured, culture medium image data acquired by the culture image acquisition unit, and learning data including a culture image and a culture medium type. Input to the discriminator generated using, a specifying unit that specifies a detector for detecting a colony based on the medium type identified by the discriminator, and the culture medium using the detector specified by the specifying unit A detection unit that detects a colony based on the culture medium image data acquired by the image acquisition unit.

  The computer program according to the present embodiment includes a computer that performs a process of acquiring culture medium image data of a culture medium in which a microorganism is cultured, and learns the acquired culture medium image data using the learning data including the culture medium image data and the culture medium type. Input to the model, processing to specify a detection algorithm for detecting colonies based on the culture medium type identified by the learning model, and processing to detect colonies based on culture medium image data obtained using the specified detection algorithm Is executed.

  The colony detection method of the present embodiment acquires the culture medium image data of the culture medium in which the microorganism is cultured, and inputs the acquired culture medium image data to the discriminator generated using the learning data including the culture medium image data and the culture medium type. Then, a detector for detecting a colony is specified based on the culture medium type identified by the discriminator, and a colony is detected based on the culture medium image data obtained by using the specified detector.

  The discriminator generation method of the present embodiment obtains culture medium image data of a culture medium in which a microorganism is cultured, obtains a culture medium type of the culture medium image data, and uses the acquired culture medium image data and learning data including the culture medium type. Generate a classifier.

  The discriminator of the present embodiment is generated by using the culture medium image data of the culture medium in which the microorganism has been cultured and the learning data including the culture medium type of the culture medium image data.

  The medium image acquiring unit acquires medium image data of a medium in which the microorganism has been cultured. Specifically, for example, an appropriate sterilizing diluent is added to a sample (specimen) of a film-type medium, the sample solution is diluted, and a film-type medium cultured under required culture conditions is read by an optical reader (eg, a scanner or the like). ) Can be obtained.

  The discriminator is generated using the culture medium image data and the learning data including the culture medium type. As the discriminator, a multilayer neural network (deep learning) can be used. For example, a convolutional neural network (Convolutional Neural Network) can be used, but other machine learning may be used. The medium type indicates the type of medium, for example, a medium for measuring the number of general viable bacteria (also referred to as AC), a medium for measuring the number of coliforms (also referred to as CC), a medium for measuring the number of coliforms / coliforms (EC). ), And a medium for measuring the number of Staphylococcus aureus (also referred to as SA).

  Since the discriminator is generated (learned) using the culture medium image data and the learning data including the culture medium type, for example, when the culture medium image data whose culture medium type is unknown is input to the discriminator, the discriminator receives the input. The medium type can be identified based on the medium image data.

  The specifying unit inputs the medium image data acquired by the medium image acquiring unit to the discriminator, and specifies a detector for detecting a colony based on the medium type identified by the discriminator. The detector is a detection logic or a detection algorithm. Thereby, an optimal detector can be specified according to the type of medium (medium type).

  The detection unit detects a colony based on the culture medium image data acquired by the culture image acquisition unit using the detector identified by the identification unit. A colony refers to a colony of bacteria that are obtained by culturing bacteria that are so small as to be invisible and increasing the bacteria until they are visible. Colonies are counted to determine the number of bacteria.

  According to the above configuration, the culture medium type is identified only by inputting the culture medium image data to the discriminator, and the optimal detector for the identified culture medium type can be identified and the colony can be detected. This eliminates the need for complicated operations such as the above, and improves the accuracy of colony detection.

  In the colony detection device of this embodiment, the detection unit generates the culture medium image data acquired by the culture medium image acquisition unit using learning data including a detection processing parameter for detecting a colony and colony detection result information. The colony is detected based on the detection processing parameters identified by the discriminator and the detector.

  The discriminator generation method of the present embodiment acquires detection result information of a colony corresponding to the obtained culture medium image data, and is a detection processing parameter for detecting a colony, and a detection processing parameter corresponding to the detection result information. Is obtained, and the classifier is generated using learning data including the obtained detection result information and the detection processing parameter.

  The discriminator is generated (learned) using learning data including a detection processing parameter for detecting a colony and colony detection result information. The detection result information can include colony position information and reliability of colony detection (for example, the probability of how correct the colony detection result is). For example, each time the medium image data is provided to the classifier as learning data and the detection processing parameter is updated, the updated detection processing parameter and the detection result information of the colony obtained as a result of the update are provided to the classifier as learning data, Can be learned. Thereby, the discriminator can discriminate a detection processing parameter necessary for setting the reliability of colony detection to a predetermined threshold or more.

  The detection unit inputs the culture medium image data acquired by the culture medium image acquisition unit to the discriminator, and detects a colony based on the detection processing parameter identified by the discriminator and the detector.

  With the above-described configuration, it is possible to identify the optimum detection processing parameter for the detector and detect a colony simply by inputting the culture medium image data to the discriminator. Labor is not required, and the accuracy of colony detection can be improved.

  The colony detection device according to the embodiment includes an additional information acquisition unit that acquires additional information related to the culture medium image data acquired by the culture image acquisition unit, and a detection processing parameter specified by the identifier is acquired by the additional information acquisition unit. And an adjusting unit for adjusting based on the additional information.

  The supplementary information acquisition unit acquires supplementary information related to the acquired medium image data. For example, by providing an area in which characters or symbols can be written in the margin excluding the culture layer of the film-type medium, and by writing incidental information in the area, by reading the film-type medium with an optical reader , Additional information can be obtained. The incidental information can include, for example, the type of the specimen, the type or dilution ratio of the diluent, the culture conditions, and the like.

  The adjustment unit adjusts the detection processing parameter specified by the classifier based on the incidental information acquired by the incidental information acquisition unit. The culture medium image data acquired by the culture medium image acquisition unit is image data obtained as a result of culturing based on the type of the specimen and the type or dilution rate of the diluent, culture conditions, and the like. And the conditions such as the type or dilution rate of the diluent and the culture conditions can be considered to be implicitly included. Therefore, when the discriminator learns the medium image data as learning data, it is considered that the learning data implicitly also includes conditions such as the type of the specimen, the type or dilution rate of the diluent, and the culture conditions. .

  However, by adjusting the detection processing parameters specified by the discriminator based on the supplementary information, more optimal detection processing parameters can be obtained, and the detection accuracy of colonies can be further improved.

  The colony detection device of this embodiment includes an output unit that outputs a predetermined notification when the reliability of the culture medium type identified by the discriminator is equal to or less than a predetermined threshold.

  The output unit outputs a predetermined notification when the reliability of the culture medium type identified by the classifier is equal to or less than a predetermined threshold. The reliability is an index indicating how reliable the classification result of the classifier is, and the threshold can be set to an appropriate value such as 98% or 95%. The predetermined notification includes, for example, a notification that the medium type identified by the discriminator is incorrect or that the detection of a colony is impossible. Thereby, if the colony cannot be detected, the fact is notified, so that erroneous colony detection can be prevented.

  The colony detection device of this embodiment includes a setting unit that sets a required detector for detecting a colony when the output unit outputs the notification.

  The setting unit sets a required detector for detecting a colony when the output unit outputs the notification. For example, even when the type of medium cannot be identified and the detector cannot be specified, the required detector can be set, so that the colony can be detected.

  The colony detection device of this embodiment has a type information acquisition unit that acquires the culture medium type information of the culture medium image data acquired by the culture medium image acquisition unit, and a case in which a colony based on the culture medium image data acquired by the culture image acquisition unit is not detected. A determination unit that determines a culture medium type of the culture medium image data based on the culture medium type information acquired by the type information acquisition unit.

  The type information acquisition unit acquires the culture medium type information of the culture medium image data acquired by the culture medium image acquisition unit. For example, the medium type information can be described in a margin excluding the culture layer of the film-type medium. The medium type information includes the medium type, and can be described by, for example, characters or symbols.

  When a colony based on the medium image data acquired by the medium image acquiring unit is not detected, the determining unit determines the medium type of the medium image data based on the medium type information acquired by the type information acquiring unit. Even when the discriminator cannot identify the culture medium type, the type of culture medium can be determined based on the culture medium type information, so that the learning effort of the classifier can be reduced.

  The colony detection device of the present embodiment detects a colony based on the culture medium type identified by the discriminator and the supplementary information related to the culture medium image data acquired by the culture medium image acquisition unit when the detection unit cannot detect the colony. And a colony based on the culture medium image data acquired by the culture medium image acquisition unit using the detector produced by the production unit.

  For example, when medium image data of a medium in which an unknown sample such as a sample that has not been tested in the past is cultured is input to the classifier, the type of the medium identified by the classifier and the optimal detection for the type of medium are determined. It is assumed that the detection unit cannot detect a colony even when using a detector. Therefore, if the detection unit cannot detect a colony, a new detector is generated based on the supplementary information related to the culture medium type and the culture medium image data. Since the adjustment is performed automatically, the colony can be detected.

  The colony detection device according to the present embodiment includes a medium image acquiring unit that acquires medium image data of a medium in which microorganisms are cultured, and a medium image data acquired by the medium image acquiring unit, a medium image, detection for detecting a colony. Input to the discriminator generated using the learning data including the detector data and the detection processing parameter data, and using the detector generated based on the detector information and the detection processing parameter information specified by the discriminator. And a detecting unit for detecting a colony based on the medium image data acquired by the medium image acquiring unit.

  The medium image acquiring unit acquires medium image data of a medium in which the microorganism has been cultured. Specifically, for example, an appropriate sterilizing diluent is added to a sample (specimen) of a film-type medium, the sample solution is diluted, and a film-type medium cultured under required culture conditions is read by an optical reader (eg, a scanner or the like). ) Can be obtained.

  The discriminator is generated using learning data including a culture medium image, detector information of a detector for detecting a colony, and detection processing parameter information. Thus, when the culture medium image data is input to the discriminator, the discriminator can specify the detector information and the detection processing parameter information based on the input culture medium image data.

  The detection unit can detect a colony based on the culture medium image data acquired by the culture medium image acquisition unit using the detector information and the detection processing parameter information specified by the classifier. As a result, it is not necessary to specify the culture medium type, and it is possible to collectively specify the detector and the detection processing parameter, so that it is possible to flexibly detect a colony even in the future new type. Become.

  The colony detection method of the present embodiment reads a culture medium in which microorganisms are cultured in a culture layer formed on a film-like base sheet with an optical reader, and obtains culture medium image data obtained by reading with the optical reader. I do.

  The culture medium image data can be obtained by reading a culture medium in which microorganisms are cultured in a culture layer formed on a film-like base sheet with an optical reader.

  In the discriminator generation method of this embodiment, the culture medium image data input to the discriminator and the culture medium type identified by the discriminator are collected as learning data, and the discriminator is updated using the collected learning data.

  The culture medium image data input to the classifier and the culture medium type identified by the classifier are collected as learning data, and the classifier is updated using the collected learning data. The learning data to be collected can include data when the identification by the classifier is incorrect. Thereby, the classifier can be re-learned, and the classifying ability of the classifier can be enhanced.

  In the discriminator generation method of the present embodiment, the acquired culture medium image data is classified for each acquisition source, and the classifier is generated for each acquisition source using learning data including the classified culture medium image data for each acquisition source.

  The acquired medium image data is classified for each acquisition source. The acquisition source can be, for example, a company or a company. Different companies handle different samples and culture conditions. A classifier is generated for each acquisition source using learning data including the classified medium image data for each acquisition source. As a result, a classifier optimized for each company or company can be generated, and the classifier can have higher discrimination ability.

Reference Signs List 20 film type medium 21 base sheet 22 frame 23 culture layer 24 cover sheet 50 colony detection device 51 control unit 52 communication unit 53 storage unit 54 adjustment unit 55 determination unit 56 output processing unit 60 processing unit 61 learning model 62 learning processing unit 63 Detection unit 631 Detection logic unit 632 Detection processing parameter unit

Claims (16)

A culture medium image acquisition unit that acquires culture medium image data of a culture medium in which microorganisms have been cultured,
The medium image data acquired by the medium image acquiring unit is input to a discriminator generated using learning data including a medium image and a medium type, and the colony is detected based on the medium type identified by the discriminator. A specifying unit for specifying a detector of
A detection unit for detecting a colony based on the culture medium image data acquired by the culture medium image acquisition unit using the detector identified by the identification unit. The detection unit,
The culture medium image data acquired by the culture medium image acquisition unit is further input to the classifier generated using learning data including a detection processing parameter for detecting a colony and colony detection result information, and the classifier is 2. The colony detection device according to claim 1, wherein a colony is detected based on a detection processing parameter to be identified and the detector. Supplementary information acquisition unit for acquiring supplementary information related to the culture medium image data acquired by the culture medium image acquisition unit,
The colony detection device according to claim 2, further comprising: an adjustment unit configured to adjust a detection processing parameter specified by the classifier based on the incidental information acquired by the incidental information acquisition unit.   The colony detection device according to any one of claims 1 to 3, further comprising: an output unit that outputs a predetermined notification when the reliability of the culture medium type identified by the classifier is equal to or less than a predetermined threshold.   The colony detection device according to claim 4, further comprising: a setting unit configured to set a required detector for detecting a colony when the output unit outputs the notification. Type information acquisition unit for acquiring the culture medium type information of the culture medium image data acquired by the culture medium image acquisition unit,
A determination unit that determines a culture medium type of the culture medium image data based on the culture medium type information acquired by the type information acquisition unit when a colony based on the culture medium image data acquired by the culture image acquisition unit is not detected. The colony detection device according to any one of claims 1 to 4. If the detection unit cannot detect a colony, a detector for detecting a colony is generated based on the culture medium type identified by the discriminator and incidental information related to the culture medium image data acquired by the culture medium image acquisition unit. Including a generating unit,
The colony detection device according to any one of claims 1 to 6, wherein a colony based on the culture medium image data acquired by the culture medium image acquisition unit is detected using a detector generated by the generation unit. A culture medium image acquisition unit that acquires culture medium image data of a culture medium in which microorganisms have been cultured,
The culture medium image data acquired by the culture medium image acquisition unit, a culture medium image, input to the discriminator generated using learning data including detector information and detection processing parameter information of a detector for detecting colonies, the A colony detection device comprising: a detection unit that detects a colony based on the culture medium image data acquired by the culture image acquisition unit using a detector generated based on the detector information and the detection processing parameter information specified by the classifier. On the computer,
A process of acquiring medium image data of the medium in which the microorganism has been cultured,
The acquired culture medium image data is input to a learning model generated using learning data including the culture medium image data and the culture medium type, and a detection algorithm for detecting a colony based on the culture medium type identified by the learning model is specified. Processing,
A process of detecting a colony based on the culture medium image data obtained using the specified detection algorithm. Obtain medium image data of the medium in which the microorganism was cultured,
The acquired medium image data is input to the discriminator generated using the learning data including the medium image data and the medium type, and a detector for detecting a colony based on the medium type identified by the discriminator is used. Identify
A colony detection method for detecting a colony based on culture medium image data obtained by using a specified detector. The culture medium in which the microorganisms are cultured in the culture layer formed on the film-like base sheet is read by an optical reader,
The colony detection method according to claim 10, wherein culture medium image data obtained by reading with the optical reading device is obtained. Obtain medium image data of the medium in which the microorganism was cultured,
Obtain the medium type of the medium image data,
A classifier generation method for generating a classifier using acquired medium image data and learning data including a medium type. Acquire colony detection result information corresponding to the acquired medium image data,
A detection processing parameter for detecting a colony, to obtain a detection processing parameter corresponding to the detection result information,
The classifier generation method according to claim 12, wherein the classifier is generated using learning data including the acquired detection result information and the detection processing parameter. Collect the culture medium image data and the culture medium type identified by the classifier as learning data input to the classifier,
14. The discriminator generation method according to claim 12, wherein the discriminator is updated using the collected learning data. Classify the acquired culture image data for each acquisition source,
The discriminator generation method according to any one of claims 12 to 14, wherein the discriminator is generated for each acquisition source using learning data including the classified medium image data for each acquisition source. A discriminator that is generated using medium image data of a medium in which microorganisms are cultured and learning data including a medium type of the medium image data.

Images