• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    A method for determining the growth of microorganisms in a biological sample capable of containing microorganisms, said biological sample being contained in an analysis vessel such as a petri dish, said analysis vessel being subjected to a incubation of a determined duration, said method comprising the following steps: - acquisition of a first plurality of initial images of the analysis container at a first acquisition time T1, before or during the incubation; acquiring a second plurality of final images of the analysis container at a second acquisition time T2, during or after the incubation; - resetting each initial image of the first plurality of acquired initial images, with each corresponding final image of the second plurality of acquired final images; locating at least one potential microorganism growth zone within at least one image of the second plurality of acquired images; - assessment of the content of the potential microorganism growth zone identified to determine the presence of microorganisms.
    公开号:FR3028866A1
    申请号:FR1461533
    申请日:2014-11-26
    公开日:2016-05-27
    发明作者:Marthe Lagarrigue-Charbonnier；Lorene Allano；Marine Depecker
    申请人:Biomerieux SA；Commissariat a lEnergie Atomique et aux Energies Alternatives CEA；
    IPC主号:
    专利说明:

    [0001] TECHNICAL FIELD The invention relates to the field of computer-generated image processing, and more specifically, to a method, a method, a method, a computer program, a computer program and a computer program. system and a computer program product for determining the growth of microorganisms for detecting and identifying microorganisms within an object such as a petri dish, after the incubation of said object. STATE OF THE ART In the field of microbiology, it is known to use images of a Petri dish acquired at different incubation times in order to detect and identify possible microorganisms. The Petri dish contains a culture medium such as a specific agar in which a sample adapted to said agar is inoculated. In known manner, a first image of the Petri dish is acquired at a time TO, prior to the incubation of the Petri dish and by means of a suitable imaging system. Thus, the first acquired image does not include any colony of microorganisms. The Petri dish is then incubated for a determined period of time, up to a time Ti. A second image of the Petri dish is then acquired at time Ti.
    [0002] Then, the first image acquired at the time TO is compared with the image acquired at time T1, in order to determine whether colonies of microorganisms appeared during the incubation period, that is to say if the Inoculated sample generated growth of microorganism colonies within the Petri dish.
    [0003] However, the comparison of the images of the Petri dish does not reliably detect a possible presence of colonies of microorganisms. In fact, many parameters can lead to an erroneous result of the comparison of the images of the Petri dish. With respect to the Petri dish itself, various parameters may result in erroneous detection of microorganism colonies. A first parameter relates to the material of the Petri dish. Thus, the Petri dish is made of a material, such as glass or plastic, which may have defects such as scratches. Depending on the conditions of acquisition of the images of the Petri dish, the stripes may appear on the image of the Petri dish as a separate element. Thus a scratch can be interpreted, erroneously, as representative of a growth of microorganisms. A second parameter relates to elements associated with the Petri dish. Thus, the Petri dish generally includes a screen printed inscription, for example a number to identify the origin of the production of the Petri dish. In a manner similar to the stripes, according to the conditions of acquisition of the images of the Petri dish, the silkscreened inscription may appear on the image of the Petri dish as a separate element. Thus, a silk screen inscription may be erroneously interpreted as representative of growth of microorganisms. A third parameter relates to the culture medium disposed within the Petri dish, such as agar. Indeed, the agar necessary for culturing the sample may contain structural defects. Depending on the acquisition conditions of the Petri dishes, these structural defects such as bubbles or agar shrinkage may also appear as distinct elements on the Petri dish image. Thus, structural defects can also be misinterpreted as representative of microorganism growth.
    [0004] 3028866 P129079.EN.01 3 A fourth parameter concerns the presence of condensation. Indeed, during the incubation period, fogging water particles may appear on the walls of the Petri dish. However, during the analysis of the Petri dish, it undergoes a reversal which can cause the drop of water particles on the agar. Thus, depending on the conditions of acquisition of an image of the Petri dish, said water particles may appear as distinct elements on said image of the Petri dish. Thus, the presence of mist may also lead to a misinterpretation of growth of microorganisms.
    [0005] Finally, a fifth parameter concerns possible displacements of the Petri dish in translation and / or in rotation between the acquisition of the first image at TO and the acquisition of the second image at Ti. These displacements generate a shift of the position of the object within the first acquired image with respect to the position of said object within the second acquired image. However, a shift between the images acquired at time T0 and T1 does not allow to compare them by superposition, in order to detect the presence of colonies of microorganisms. In the prior art, EP796319 discloses a method of identifying colonies of microorganisms comprising acquiring images of a culture device such as a Petri dish or a PetrilfilmTM film. The method comprises steps of resetting the images acquired by using as a reference position the center of the culture device. The registration relates solely to a translational movement of the culture device between a first image acquired at a time TO and a second image acquired at a time Ti. Consequently, patent EP796319 does not describe a resetting step making it possible to consider any possible rotational movement of the culture device between the two acquisition times TO and T1 of the images. In addition, the image acquisition conditions are such that the illumination conditions of an image acquired at time TO are identical to those of an image acquired at time T1. Thus, the patent EP796319 describes image comparison steps comprising the subtraction of the pixels of the image acquired at TO with the pixels of the image acquired at T1 in order to obtain a comparison image comprising P129079.FR.01 4 resulting pixels whose intensity is possibly representative of a growth of microorganisms. Therefore, patent EP796319 does not describe image matching steps adapted to images acquired under different illumination conditions.
    [0006] Therefore, it is necessary to improve the growth detection of microorganisms within an object such as a Petri dish by considering the image registration related to a translation and / or rotation. of the Petri dish during the acquisition of the images of the Petri dish at different times.
    [0007] OBJECT OF THE INVENTION The present invention aims to overcome at least partially the problems mentioned above.
    [0008] Thus, a first object of the invention is to provide a method for determining the growth of microorganisms in a biological sample capable of containing microorganisms, said biological sample being contained in an analytical container such as a Petri dish, said analysis container being subjected to an incubation of a determined duration, said method comprising the following steps: - acquiring a first plurality of initial images of the analysis container at a first acquisition time T1 , before or during incubation; acquiring a second plurality of final images of the analysis container at a second acquisition time T2, during or after the incubation; - resetting each initial image of the first plurality of acquired initial images, with each corresponding final image of the second plurality of acquired final images; locating at least one potential microorganism growth zone within at least one image of the second plurality of acquired images; - Evaluate the content of the identified potential microorganism growth area to determine the presence of microorganisms.
    [0009] Advantageously, the registration step comprises a primary registration step associated with an identifier located on the analysis container.
    [0010] Advantageously, the registration step comprises a secondary registration step associated with the contents of the analysis container. Advantageously, the step of locating at least one potential microorganism growth zone comprises a step of detecting a high density of microorganism colonies, a step of detecting microorganism colonies. non-circular and a step of assigning a location element to said potential area of growth of microorganisms. Advantageously, the step of evaluating the content of the potential growth zone identified comprises a step of determining the values of two evolution parameters. Advantageously, the two evolution parameters comprise a correlation parameter and a contrast parameter.
    [0011] Advantageously, according to the invention the secondary registration step comprises the creation of a mesh of sub-images for each initial image and each final image.
    [0012] Thus, a second object of the invention is to provide a system for determining the growth of microorganisms in a biological sample capable of containing microorganisms, said biological sample being contained in an analytical container such as a Petri dish, said analysis container being subjected to an incubation of a determined duration, said system comprising: an image capture device in order to acquire a plurality of images of the object to be analyzed at a first T1 acquisition time, before or during incubation, and at a second T2 acquisition time, during or after incubation; an illumination device comprising one or more light sources in order to illuminate the analysis container; A control device for controlling the application of the method for determining the growth of microorganisms according to the invention in order to determine the presence of microorganisms. Thus, a third object of the invention is to provide a computer program product comprising software instructions for carrying out a method according to the invention when said program is executed by a data processor.
    [0013] BRIEF DESCRIPTION OF THE DRAWINGS The invention, its functionality, its applications as well as its advantages will be better understood on reading the present description, with reference to the figures, in which: FIG. 1 shows a schematic view of a system image analysis apparatus 20 according to one embodiment of the invention; FIG. 2 shows a diagram of the steps of the overall method for determining the growth of microorganisms according to one embodiment of the invention; FIG. 3 shows a diagram of the steps of the primary registration method according to one embodiment of the invention; FIG. 4 shows an image of a Petri dish acquired at time T1; FIG. 5 shows an image of the Petri dish according to FIG. 4 with a targeted representation on the pixels distant from the center of the Petri dish according to an R-10 pixel radius; FIG. 6 shows an image of the Petri dish according to FIG. 4 with a targeted representation on the pixels distant from the center of the Petri dish in an R-50 pixel radius; FIG. 7 shows an image of the Petri dish according to FIG. 4 in application of a segmentation operation; FIGS. 8 and 9 show an image of the Petri dish according to FIG. 4 after the application of a morphological opening operation; FIG. 10 shows an image of the Petri dish in FIG. 4 after application of an erosion function; - Figure 11 shows in detail the contents of the rectangle shown in Figure 10; FIG. 12 shows an image of the Petri dish according to FIG. 9 after application of an opening function; Figure 13 shows in detail the contents of the rectangle shown in Figure 12; Fig. 14 shows a graphical representation of the theoretical side label; FIG. 15 shows a graphical representation of the theoretical side label in broken lines, the orientation angle θ in solid line and the equivalent ellipsoid in the form of dots; FIG. 16 shows a comparative graphical representation of the images of the lateral tag acquired at T1 and T2; Fig. 17 shows a graphical representation according to Fig. 16 based on the upper halves of said theoretical side labels; Fig. 18 shows a graphical representation of the tags according to Fig. 16 based on the lower halves of said theoretical side tags; FIG. 19 shows an image of a Petri dish acquired at a time T1; Figure 20 shows an image of the Petri dish acquired at time T2; FIG. 21 shows a diagram of the steps relating to the secondary registration method according to the embodiment of the invention; FIG. 22 shows the representation of a pixel within an initial cell; FIG. 23 shows in detail the contents of the square represented in FIG. 22; FIG. 24 shows the representation of a pixel within FIG. a final cell; FIG. 25 shows in detail the contents of the square shown in FIG. 24; FIG. 26 shows a diagram of the steps relating to the local registration process according to FIG. 21; FIG. FIG. 27 shows an image of the Petri dish acquired at a time T1 according to the "backlight" acquisition condition; FIG. 28 shows an image of the Petri dish according to FIG. 27 acquired at a time T2 according to the "backlight" acquisition condition; Figure 29 shows in detail the contents of the square shown in Figure 28; FIG. 30 shows a diagram of the steps relating to the method for determining the prior growth of microorganisms according to FIG. 2; FIG. 31 shows a diagram of the steps of the time analysis method; FIG. 32 shows a diagram of the steps relating to the image pre-processing method according to FIG. 31; FIG. 33 shows examples of sub-images acquired at T1 and T2 according to the "backlight", "bottom" and "median" acquisition conditions after the application of a contrast adjustment method; FIG. 34 shows examples of sub-images acquired at T1 and T2 according to the "backlight", "bottom" and "median" acquisition conditions after the application of a median filter; FIG. 35 shows a diagram of the steps relating to the evaluation of the first and second evolution parameters according to FIG. 31; FIG. 36 shows a diagram of the steps relating to the evaluation of the contrast parameter according to one embodiment of the invention; FIG. 37 shows a diagram of the steps relating to the evaluation of the correlation parameter according to one embodiment of the invention; Fig. 38 shows a diagram of the steps for determining the effective growth according to Fig. 31 according to one embodiment of the invention; Fig. 39 shows a graphical representation of the growth of microorganisms as detected, as a function of the values of the contrast and correlation parameters.
    [0014] DETAILED DESCRIPTION OF THE INVENTION The purpose of the following detailed description is to explain the invention in a sufficiently clear and complete manner, in particular by means of examples, but must in no way be regarded as limiting the invention. scope of protection to the particular embodiments and examples presented hereinafter. The present invention relates to the analysis of a sample. According to the present invention, the sample may be of various origins, for example of food, environmental, veterinary, clinical, pharmaceutical or cosmetic origin. Examples of food-based samples include, but are not limited to, a sample of dairy products (yogurts, cheeses, etc.), meat, fish, eggs, fruit, vegetables, water, drink (milk, fruit juice, soda, etc.). Of course, these food-based samples may also come from sauces or more elaborate dishes or unprocessed or partially processed raw materials. A food sample may also be derived from a feed intended for animals, such as cakes, animal meal.
    [0015] As previously indicated, the sample may be of environmental origin and may consist of, for example, surface sampling, water, air, etc. The sample may also consist of a sample of clinical origin, which may correspond to samples of biological fluid (urine, whole blood, or derivatives such as serum, saliva, pus, cerebrospinal fluid, etc.), stool (eg, cholera diarrhea), nasal, throat, skin, wound, organ, tissue or isolated cell specimens. This list is obviously not exhaustive. In general, the term "sample" refers to a portion or quantity, more particularly a small portion or a small amount, taken from one or more entities for analysis. This sample may possibly have undergone prior treatment, involving for example mixing, dilution or grinding steps, in particular if the starting entity is in the solid state.
    [0016] The sample taken is, in general, capable of - or suspected of - containing at least one target external microorganism such as a bacterium, a yeast, a fungus.
    [0017] According to the present invention, the analysis is carried out using an analysis vessel such as a Petri dish comprising a receptacle and a lid. The receptacle and / or the lid may comprise an identifier. The receptacle comprises a culture medium such as an agar culture medium. A fluid is seeded on the culture medium before the start of the analysis. The purpose of the analysis is to determine the presence of specific microorganisms and the concentration of microorganisms within the fluid on the culture medium after incubation of the Petri dish. In the present invention, the analysis vessel is a Petri dish which comprises as a identifier a lateral label disposed on the side wall of the Petri dish receptacle. According to the present invention, the conditions for acquiring the images of a Petri dish are identical. Indeed, the illumination and the angle of view of the Petri dish, when capturing images corresponding to two distinct acquisition times, are identical for each acquired image. According to the present invention, the objective of the image analysis of a Petri dish is to readjust an acquired image before incubation of the Petri dish with a corresponding image acquired after incubation of the Petri dish. Petri dish, or two images of the Petri dish acquired at different incubation times. Indeed, during the manipulation of the Petri dish, using an automated device or manually, the position of the Petri dish may vary between the moment of acquisition of the image before incubation and the instant of acquisition of the corresponding image after incubation. Thus, the comparison of the two images acquired at different times makes it possible to determine the presence or absence of colonies of microorganisms located within the Petri dish. Since the Petri dish is circular, the registration may require rotation and / or translation of the image of the Petri dish to obtain the correspondence between the two images. FIG. 1 shows an image analysis system 10 comprising an image capture device 12, an illumination device 14, an image analysis device 16 and a control device 18.
    [0018] The image capture device 12 comprises a digital device such as a digital camera for acquiring images of the object to be analyzed. The image capture device 12 acquires a first series of images at a time T1 and a second series of images at a time T2. The time T1 corresponds to an initial moment 20 that is to say before the seeding of microorganisms and before translation, or at an initial moment in the presence of microorganisms and after a first incubation period within the Petri dish. The time T2 corresponds to a time subsequent to the initial time T1, after incubation of the petri dish, that is to say, potentially, in the presence of microorganisms within the Petri dish. By way of example, the image capture device 12 comprises a CCD-type monochrome image sensor consuming between 12 and 24 volts and using scanning technology with a maximum acquisition speed of 17 frames per second.
    [0019] The illumination device 14 comprises for example one or more light sources such as light emitting diodes (LEDs) for illuminating the object to be analyzed according to conditions determined by the user. The light sources are disposed either above or below the Petri dish to be analyzed. The illumination device 14 makes it possible to generate acquisition conditions relating to different possibilities of illumination of the Petri dish such as a red, green and / or blue illumination (RGB) combined with the presence 5 or the absence of a cover located below the Petri dish, ie below the receptacle or below the lid of the Petri dish. Thus, the "backlight" type of illumination corresponds to the acquisition of images in the presence of a rear illumination of the Petri dish, in the absence of a cache located below the Petri dish. The illumination of "bottom annular" or "bottom" type 10 corresponds to the acquisition of images in the presence of an illumination located above the Petri dish, and in four directions with respect to the Petri dish, that is, to the left, to the right, to the upper part, and to the lower part of said Petri dish. The "left bottom" illumination corresponds to obtaining images based on an illumination located above the "bottom" Petri dish and on the left side of said Petri dish. The "median" type of illumination corresponds to the obtaining of a median image based on the four distinct images relating to each direction, left, right, lower part, and upper part of the Petri dish, obtained by means of a bottom illumination. Other acquisition conditions can be envisaged by using a combination of light sources arranged in a specific direction with respect to the Petri dish in the presence or absence of a cache located below the Petri dish in order to obtain a visible object without artefact depending on the nature of the sample and the micro-organisms to be identified.
    [0020] The illumination device 14 can also produce vertical or annular illuminations with respect to the Petri dish to be analyzed. The image analysis device 16 comprises in particular an object support and means for holding the object such as a clamp in order to hold an object such as a Petri dish.
    [0021] The control device 18 notably comprises a microprocessor (not shown), a display screen (not shown), a storage memory (not shown), in order to execute processing algorithms. images. The control device 18 makes it possible to define the operating parameters of the image capture device 12, the illumination device 14 and the image analysis device 16 in order to control the operation of the image capture device. 12, the illumination device 14 and the image analysis device 16. The image analysis system 10 operates in conjunction with a device 10 for incubation 20. The incubation device 20 is used to incubate at least one Petri dish according to specific incubation conditions concerning the duration and the temperature, in order to promote the growth of microorganisms within the Petri dishes.
    [0022] Thus, according to a conventional operation, and considering a single Petri dish by way of example, the Petri dish seeded with a sample is conveyed via a known means of transport according to the state of the art, such as a treadmill, within the incubator 20.
    [0023] At the end of the incubation period, the Petri dish is again conveyed within the image analysis system 10. After processing the images obtained by means of the image analysis system 10, an identification system (not shown) makes it possible to determine, according to a manual or automated method applied to the acquired images, the nature and the number of colonies of microorganisms which have possibly developed within the Petri dish. The present invention relates to the analysis of images available within the image analysis system 10, after incubation of a Petri dish and acquiring images of said Petri dish.
    [0024] In the present description, by way of example, the object under consideration is a Petri dish. The present invention applies to a plurality of images or series of images relating to one or more Petri dishes. A first series of images It is acquired at a time T = T1 and a second series of images 12 is acquired at a time T = T2. The present description, by way of example, refers to an image I1 resulting from the first series of images I1 acquired at T1 and to an image 12 resulting from the second series of images 12 acquired at T2. As shown in FIG. 2, the overall method according to the present invention 10 comprises a first acquisition step 22 of the first series of images I1, at a time T1, a second acquisition step 24 of the second series of images. 12, at a time T2, and a step 26 relating to a first primary registration method, a step 28 relating to a second secondary registration method, a step 30 relating to a third method of prior detection of micro-15 organizations. In step 22, the images are acquired according to a method described in the international application published under the number W02012 / 152769. The first, second and third methods are distinct and their respective algorithms are stored in the storage memory of the control device 18.
    [0025] The first method relates to the primary registration or coarse registration shown in FIG. 3 and relating to the positioning of the Petri dish between the first image II acquired at T1 and the second image 12 acquired at T2.
    [0026] The primary registration method is based on the detection of the side label of the Petri dish. Thus, the primary registration method comprises, as shown in FIG. 3, a method for detecting the side label 1001 and a method for determining the primary registration parameters 1002 associated with the characteristics of the side label.
    [0027] The method of detecting the side label 1001 comprises the use of at least two images of the Petri dish, with at least one acquired image using the "left bottom" illumination condition. And at least one acquired image using the backlight illumination condition. The method of detecting the side label 1001 includes a first step 100 relating to the detection of the Petri dish. The first step 100 comprises a detection of the edges of the Petri dish, according to a method known in the state of the art. The images used in the first step correspond to the images acquired at time T1 and at time T2 with the condition of illumination "backlight". Thus, the first step makes it possible to determine the coordinates of the center C and the length of the radius R of the Petri dish, for each acquired image. In the following steps, only the pixels within a disk characterized by the thus determined values of R and C are considered within the acquired images. The method of detecting the lateral tag comprises a second step 110 concerning the location of the lateral tag.
    [0028] The images used in the second step 110 correspond to the images acquired at time T1 as shown in FIG. 4, and at time T2 with the "left bottom" illumination condition. In fact, the lateral label located on the right side of the Petri dish is optimally visible according to this illumination condition. The second step 110 comprises the creation of a first disk of radius R1 = R-d1 and the creation of a second disk of radius R2 = R-d2, d1 and d2 being values determined in advance by the user, with d1 <d2, such that d1 = 10 pixels and d2 = 50 pixels. For a Petri dish of radius R = 950 pixels, the first disc, as shown in FIG. 5, corresponds to a radius R1 = 940 pixels and the second disc corresponds to a radius R2 = 900 pixels. The second step comprises the creation of a resulting ring containing the pixels located at a distance d between R1 and R2, ie between 900 and 940 pixels, as shown in FIG. 6. In the following steps only pixels within the resulting ring 5 are considered within the acquired images. The second step 110 then comprises the application of a segmentation operation or binarization on the acquired image of the Petri dish. Indeed, since the side label is distinguished by a clear ring portion of the remainder of the resulting ring, the segmentation operation effectively isolates the side label. The segmentation operation can be used by applying an algorithm such as the k-means or "k-means" algorithm of MatlabTM software based on a search for two groups or "two clusters" while considering, for a cluster, a centroid initialized to the minimum value of the RGB triplet to approach the value of the white and thus group the bright pixels and, for another cluster, a centroid initialized to the maximum value of the RGB triplet to approach the value of the black, and thus group the dark pixels. The segmentation operation makes it possible to locate different objects within the acquired images, as shown in FIG. 7. These objects comprise, for example, the lateral label and the reflection defects. The reflection defects are generated by the reflection of the light beams of the illumination device 14 on the edges of the Petri dish. The defects of light or "barbs" have a generally narrow and curved shape. The side label optionally includes illumination defects relating to overexposure during illumination of the Petri dish.
    [0029] In general, the side label is formed in one piece. However, the lateral label may sometimes consist of different fragments, voluntarily or involuntarily. Thus, depending on the type of conditioning of the Petri dishes or depending on incidents related for example to the quality of the transport of the Petri dishes, a fragment of the side label may be missing. A step of suppressing illumination defects does not always make it possible to distinguish the side label from the illumination defects when the size of the side label is of the same order of magnitude as the size of the illumination defects. Thus, the method of detecting the lateral label comprises a third step 120 concerning the suppression of the illumination defects of the lateral label. The third step 120 comprises the application of a function of morphological opening, according to the prior art, on the binarized image, shown in FIGS. 8 and 9. The opening function corresponds to the combination of a function erosion and dilation function. For the application of the opening function, the STREL function of the MatlabTM software is used to define a specific structuring element comprising a segment of a given length, for example 10 pixels, relating to a defined number of neighboring pixels of the pixel in question, for example. example 3 pixels, associated with an angle of rotation of said structuring element between -90 and 90 degrees. As shown in Figures 10 and 11, the value of the rotation angle is 0 °. As shown in FIGS. 12 and 13, the value of the angle of rotation is 45 °. The erosion function, with the values defined above, makes it possible to eliminate objects having a thickness of less than 10 pixels whatever the curvature of these objects. The expansion function makes it possible to obtain a correction of the curvature of the lateral label, in order to obtain an overall curvature of convex shape. The aperture function is applied for each color channel of the RGB triplet. Thus, at the end of the application of the opening function, the illumination defects of the lateral label are deleted on the images acquired at time T1 and time T2.
    [0030] Next, the method of detecting the lateral tag comprises the fourth step 130 relating to the determination of a theoretical lateral tag formed of a single fragment. The zone of the Petri dish associated with the theoretical side label corresponds to a crown circle arc located in the center of the Petri dish, of width 20 pixels and with a spreading angle O. The angle of spread 0 of the theoretical side label is determined by the maximum value Omax and the minimum value Emin of 0 in order to define the spreading surface covering all the fragments of the side label detected in the previous step. The fourth step 130 provides a theoretical side label, as shown in Fig. 14. The theoretical side label is located around the perimeter of the Petri dish, as shown in Fig. 5, and contains all the fragments of the side label. Thus, the method of determining the primary registration parameters 1002 can be applied as described below.
    [0031] The method for determining the primary registration parameters 1002 comprises a first step 140 concerning the determination of the orientations of the lateral labels and the determination of a range of values of the angles of rotation. To define a search range for the angle of rotation, the orientation of the theoretical side tags at time T1 and time T2 is used. In known manner, the "regionprops" function associated with the "Orientation" property of the MatlabTM software makes it possible to determine the orientation of objects within a binarized image. The orientation of the lateral theoretical label is determined by obtaining the angles O1 and O2. This orientation corresponds to the orientation of the ellipse having the same moment of inertia as that of the theoretical lateral label, as shown in FIG. Figure 15.
    [0032] To define the search range for the angle of rotation, the theoretical side label is used in three different ways. Indeed, according to a first aspect, shown in FIG. 16, the calculation of the orientation 5 of the theoretical lateral label is based on the totality of the theoretical lateral label in order to obtain the corresponding angles θt and θtot and the Angular difference AOtot. According to a second aspect, shown in FIG. 17, the calculation of the orientation of the theoretical side label is based on the upper half of the theoretical side label in order to obtain the corresponding angles θsup and θsup and the deviation angular corresponding AOsup. According to a third aspect, shown in Fig. 18, the calculation of the orientation of the theoretical side label is based on the lower half of the theoretical side label to obtain the corresponding angles 01inf and 02inf and the gap angular corresponding AOinf. For each of the three aspects, the deviations between the orientation angles at T1 and at T2 are calculated as indicated below: AOtot = 101tot-02tot1, AOsup = 101sup-02sup I, AOinf = 101inf-02infl, An interval possible values of the optimal orientation angle Oopt is defined below: Optimal interval: [min (00tot, AOsup, AOinf) - 0,2 '; max (00ot, AOsup, AOinf) + 0.2 °].
    [0033] The method for determining the primary registration parameters 1002 comprises a second optimal adjustment step 150 concerning the rotation and the translation of the Petri dish between the times T1 and T2, the center of the rotation being assimilated in the center of the Petri dish. The optimal registration step makes it possible to determine, from the image acquired at time T2, optimal primary registration parameters comprising a rotation angle and a translation vector in order to apply these parameters to the image acquired at the time. time T1 so that the two acquired images are superimposable. The determination of the optimal parameters is obtained using a 2D cross correlation function, as available in the MatlabTM software. To estimate the optimal parameters, all rotation angle values in the optimum interval are tested with a pitch of 0.1 ° of the image at T2 and the cross correlation between the T2-rotated image and the rotated T2 image is calculated. image at Ti. Each resulting value represents the optimal translation for each rotation angle value in the optimum range. The maximum of these resulting values makes it possible to determine the optimal rotation and its associated optimal translation. The second method of the overall method of the invention relates to the secondary registration or late registration 28. The second method may optionally be carried out first, without the first method relating to the primary registration being carried out beforehand. FIG. 19 shows a first image or initial image of a Petri dish, acquired by means of the image analysis system 10 at a time T1, before the incubation of the Petri dish, for example. Fig. 20 shows a second or final image of the Petri dish shown in Fig. 19 and acquired at time T2 after incubation of the Petri dish within incubator 20. As shown in Fig. 20 colonies of microorganisms are present within the Petri dish at time T2.
    [0034] As shown in FIG. 21, the secondary registration method comprises a step 260 relating to the creation of a mesh of sub-images within the initial image 11 acquired at time T1 and from the final image 12 acquired at time T2, a step 270 relating to a local registration process and a step 280 relating to a global registration method. For the step 260 of creating a mesh of sub-images, an initial image I1 is squared, as shown in FIG. 22, according to an initial mesh that comprises initial cells. Figure 23 shows an example of an initial cell. The final image 12 is squared as shown in Fig. 24 in a final mesh that includes final cells. Figure 25 shows an example of a final cell. The size of an initial cell is larger than the size of a final cell. As shown in FIG. 21, the secondary registration method comprises a local adjustment step 270 for each pair of sub-images obtained at the end of step 260 described above. The local registration step 270, shown in Fig. 26, includes a substep 272 for calculating the value of an optimal displacement between the two sub-images considered within the pair of sub-images. The optimal displacement is estimated by considering the displacement of the final cell on the corresponding initial cell. Insofar as the size of the final cell is smaller than the size of the initial cell, the displacement of the final cell is possible according to a margin of displacement parameterized by the user. For each move a value of a similarity parameter is calculated in a step 273. This similarity parameter may be, for example, the distance or mutual information measure. In the present invention, the similarity parameter corresponds to the correlation coefficient. For each displacement of the final subimage on the initial subimage, a value of the correlation coefficient is obtained. The maximum value of this correlation coefficient defines the optimal displacement for a pair of sub-images.
    [0035] Sub-step 274 makes it possible to calculate the value of a contrast parameter between the two sub-images considered within the pair of corrected sub-images according to the optimal displacement estimated in the substep 272. The second contrast parameter makes it possible to evaluate the amount of information present within each subimage of the pair of sub-images. In the present invention, the value of the contrast parameter is calculated as the standard deviation of the pixel-pixel calculated color distance according to the formula below: XE [X]) 2] = ./E[X2 ] - E [X] 2 The contrast parameter is used to determine the sub-images that are relevant for subsequent calculation in the overall registration process. Thus, step 270 makes it possible to obtain a set of translational local transformations between the initial image acquired at T1 shown in FIG. 27 and the final image acquired at T2 shown in FIG. 28 by vectors for each point of the mesh. Figure 29 shows in detail the contents of the square shown in Figure 38. As shown in Figures 28 and 29, these local transformations are represented by vectors.
    [0036] As shown in Fig. 21, the image registration method according to the invention comprises a step 280 of global registration. Step 280 makes it possible to use the results of step 270 in order to deduce, for each image pair II and 12, global registration parameters. The determination of a global transformation is carried out on the basis, on the one hand, of all local transformations, ie translations, and on the other hand, similarity and contrast parameters for all couples of sub-images.
    [0037] The values of the similarity and contrast parameters make it possible to select the most relevant pairs of sub-images. Thus, the more relevant local transformations can be deduced, for example, by applying a comparison step with a threshold value. Then, a global transformation is determined from the most relevant local transformations for example by considering the average, median or local transformation most frequently represented among the local transformations. As shown in FIG. 2, the overall method according to the invention comprises a step 30 of preliminary determination of growth of microorganisms on the image 12 acquired at time T = T2. Step 30 of Figure 2 comprises applying a method of prior growth determination to detect the possible growth of microorganisms within the Petri dish after incubating said Petri dish. The method of preliminary growth determination comprises a method for identifying potential growth areas and a method for temporally analyzing said growth zones. As shown in FIG. 30, the method of preliminary growth determination comprises a step 300 relating to obtaining one or more images of the Petri dish, at the end of the secondary registration process, according to different conditions of growth. relative acquisition including the illumination device 14. The method of preliminary growth determination also comprises a step 302 for creating a digitized image from the images obtained in step 300. The method of preliminary growth determination comprises a step 304 to detect the presence of a high density of microorganism colonies as well as colonies of non-circular microorganisms. Thus, step 304 makes it possible to detect a growth zone of microorganisms, the shape of said growth zone being variable. In order to facilitate the subsequent application of the calculations, said growth zone is defined by means of a rectangle called "enclosing rectangle". The "enclosing rectangle" is centered on the growth area and allows the area of the growth area to be enlarged to a predetermined size, for example 10 pixels around the growth zone. The method of preliminary growth determination then comprises a step 306 for associating a locating element such as a bounding rectangle, with each growth element determined by the prior growth determination method. Thus, the method for the preliminary determination of growth of microorganisms makes it possible to determine whether colonies of microorganisms are present within the Petri dish.
    [0038] In order to improve the results obtained from the preliminary determination method, a temporal analysis method 40 of the growth of microorganisms is then applied in order to validate each rectangle encompassing a growth zone obtained during the application of the method of Prior determination of growth of microorganisms and thus obtain the indication of the effective growth of microorganisms within the Petri dish. The temporal analysis method therefore considers the images obtained after the application of the optional primary registration method and the secondary registration method, and for which the application of the prior growth determination method has generated rectangles encompassing areas potential for growth within the corresponding images. According to the embodiment of the present invention, the initial images and the final images considered correspond to images acquired with the following three acquisition conditions: "backlight", "bottom" and "median".
    [0039] Indeed, each acquisition condition of the illumination device 14 allows a specific observation of the Petri dish. Thus, the "backlight" type of illumination makes it possible to observe by transparency the elements contained within the Petri dish.
    [0040] The acquisition conditions of the "bottom" and "median" type make it possible to observe the elements located on the surface of the agar culture medium within the Petri dish.
    [0041] As shown in Fig. 31, the time analysis method comprises a step 420 for applying an image preprocessing method to the initial images and the final images, a step 430 for evaluating evolution criteria based on the initial images. and the final images and a step 440 for determining the effective growth of microorganisms.
    [0042] Step 420, relating to the image preprocessing method and shown in Fig. 32, includes a substep 424 for enhancing the contrast within each bounding rectangle area of said area within the subpictures. Thus, sub-step 424 relates to the application of an image contrast adjustment method such as the histogram equalization known in the prior art. Thus, the images obtained after step 424 contain different levels of gray as shown in FIG. 33, depending on the different illumination conditions, ie "backlight", "bottom" and "median".
    [0043] Step 420 then comprises a substep 425 to reduce, within the image, the noise generated by the contrast enhancement. Thus, sub-step 425 relates to the application of a median filter as known in the prior art, for example a filter for a 3 × 3 window.
    [0044] Thus, at the conclusion of substep 425, the areas relating to the rectangles encompassing the potential growth areas contain transitions shown in FIG. 34 which are smoother and more homogeneous than the transitions of said zones shown in FIG. 33, at the end of step 424. As shown in FIG. 31, the time analysis method comprises a step 430 for determining, within the sub-images, the value of FIG. at least one evolution parameter. According to a preferred embodiment of the invention, the value of two evolution parameters must be determined for each set of sub-images.
    [0045] Step 430 comprises a substep 461 shown in FIG. 35 and relating to the first evolution parameter which makes it possible to characterize the color evolution within the sub-images between the acquisition times T1 and T2. According to one embodiment of the invention, the first evolution parameter corresponds to a contrast parameter which makes it possible to evaluate the overall stability of the subimage between the acquisition times T1 and T2. According to one embodiment of the invention, the first evolution parameter corresponds to a correlation parameter. Step 430 comprises a substep 462 shown in FIG. 35 and relating to the second evolution parameter. According to one embodiment of the invention, the contrast measurement method is based on the calculation of the standard deviation. for example by means of the "std" function of the MatlabTM software, concerning the measured distances, for example by means of the "pdist" function of the MatlabTM software, between two pixels, and for all the pixels located inside the enclosing rectangle: C = For the measure of the contrast between voxels in a color sub-image, the variable "array" corresponds to a vector column nx 3.
    [0046] For the measurement of the contrast between pixels in a grayscale subimage, the variable "array" corresponds to a vector column n.
    [0047] As shown in Fig. 36, the sub-step 461 for calculating the contrast parameter comprises a plurality of steps based on the contrast measurement defined with equation C above, according to a embodiment of the invention.
    [0048] The first step 463 corresponds to the contrast measurement for each subimage of an initial image and for each illumination condition "backlight", "bottom" and "median".
    [0049] The second step 464 corresponds to the contrast measurement for each subimage of a final image and for each illumination condition "backlight", "bottom" and "median". The third step 465 corresponds to calculating the ratio of the contrast measurements by considering two by two the contrast measurements of a subimage at the times T1 and T2, for each illumination condition. The fourth step 466 corresponds to the determination of the minimum value of the ratios resulting from step 465 to determine the value of the contrast parameter.
    [0050] Sub-step 461 corresponds to the application of the ContC equation below: ContC = min (C {TI backlight} C {T b ottom} C {Ti median} n) n C {T2 backhght} L According to another embodiment of the invention, another ContC_alt equation can be applied using the same contrast measurements: ContC alt min (C {T, backlight}, C { T acklight}) min (C {T ottom}, C {Tibottom}) min (C {Timedian}, C {Timedian}) = min (4 C {T2backlight}, C {T 2b acklight}) maC {T2bottom} C {T2bottom}) min (C {T2median}, C {T2median}) my 3028866 P129079.EN.01 28 The ContC_alt equation has advantages over the ContC equation. Indeed, the values obtained with the ContC_alt equation are always included in the interval [0; 1]. Indeed, for a stable element, that is to say an element that has not evolved over time, the value obtained with the ContC_alt equation is close to 1, according to a margin of appreciation that the user can fix. Thus, for a margin of appreciation of 0.2, the value obtained for a stable element is in the range [0.8; 1]. For a growth element, that is to say an element that has evolved over time, the value obtained with the ContC_alt equation is close to 0, according to a margin of appreciation that the user can set. Thus, for a margin of appreciation of 0.2, the value obtained for an evolved element is in the range [0; 0,2]. As shown in Fig. 37, substep 462 comprises a plurality of steps. Thus, the first step 450 corresponds to the selection of the most contrasting initial color subimage, among the identical sub-images acquired according to the three acquisition conditions "backlight", "bottom" and "median". The second step 451 corresponds to the selection of a corresponding preprocessed final subimage and as obtained at the end of steps 424 and 425.
    [0051] The selection of the most contrasted image within steps 450 and 451 is based on the contrast measurement method described in step 461. The third step 452 comprises the transformation of the colored subimage into sub-image in grayscale. The fourth step 453 comprises calculating the correlation coefficient between the acquired subimage at T1 and the acquired subimage at T2, both in gray levels, using the "corr2" function of the MatlabTM software.
    [0052] The fifth step 454 corresponds to the calculation of the value of the correlation parameter which corresponds to the maximum value between the two correlation coefficient values calculated in step 453, using equation c. below: CorrP = max [corr2 (Tlgris, T2gris), corr2 (T1 improved, T2 improved)] As shown in FIG. 31, the time analysis method comprises a step 440 for determining the effective growth of microorganisms within of the Petri dish, at the end of a determined period of incubation.
    [0053] The objective of step 440 is to check whether the element marked by a determined growth zone during the application of the method of preliminary growth determination within a sub-image, during step 30 , effectively corresponds to a growth element such as a colony or a set of colonies of microorganisms.
    [0054] As shown in FIG. 38, step 440 comprises a step 470 for listing the different values obtained for the two evolution parameters in step 430. These values are associated with each rectangle encompassing potential growth areas. of the Petri dish for the T2 acquisition time Step 440 then comprises a step 472 relating to the application of a logistic regression method. This method makes it possible to obtain a separation between two groups of data to be discriminated on the basis of several parameters by means of a defined regression line. Within the present invention, the logistic regression method used makes it possible to distinguish two groups of data. The first group of data concerns potential areas of growth that are proven growth areas, that is, positive areas. The second group of data concerns potential areas of growth that do not correspond to growth areas. These are negative areas. The logistic regression method within the present invention is based on two parameters, CorrP and ContC, according to the equation below: ## EQU1 ## As a result of the correlation parameters CorrP and ContC, the equation P below makes it possible to determine a probability of the presence of a growth element such as colonies of microorganisms from the value of CL defined above. above.
    [0055] 1 P = 1 +% - In order to classify the potential areas of growth, the user can set a decision criterion on the values of P obtained. According to an embodiment of the present invention shown in FIG. 39, the decision criterion P is set with the following conditions: if P <0.25 then the potential growth zone corresponds to a stable element in the time therefore the potential growth zone is not a growth zone, it is a so-called negative zone; if P> 0.6 then the potential growth zone corresponds to an element that changes over time and therefore the potential growth zone is a proven growth zone, it is a so-called positive zone; - if 0.25 <P <0.6 then it is not possible to determine whether the potential area of growth corresponds to a stable element or an element that evolves over time, that is, the distinction between between a so-called positive zone and a so-called negative zone is not possible.
    [0056] The time analysis method then makes it possible to compare, in a step 474, the value of the local annotation with the decision criterion P, and thus to find, for each potential growth zone of a final image, the correspondence with the corresponding initial image. Thus, in a step 476, the user can determine the absence or presence of microorganism colonies within the Petri dish analyzed.
    [0057] After calculating the correlation and contrast parameters, the temporal analysis method makes it possible to determine whether the growth zone corresponds to an element that has evolved or has not evolved over time.
    [0058] When the growth zone corresponds to an element that has evolved over time, the time analysis method makes it possible to confirm the validity of the growth zone as determined by the prior growth determination method. Thus, the growth zone effectively corresponds to growth of microorganisms. When the growth zone corresponds to an element that has not evolved over time, such as a defect related to the material of the Petri dish, the time analysis method makes it possible to obtain a correction of the growth zone. . Thus, the corrected growth area indicates that the element observed within said growth zone does not correspond to growth of microorganisms. After growth determination, the user has the opportunity to identify the nature of the microorganism colonies by means of an identification system (not shown).
    权利要求:
    Claims (9)
    [0001]
    REVENDICATIONS1. A method for determining the growth of microorganisms in a biological sample capable of containing microorganisms, said biological sample being contained in an analysis vessel such as a petri dish, said analysis vessel being subjected to a incubation of a determined duration, said method comprising the following steps: - acquisition of a first plurality of initial images of the analysis container at a first acquisition time T1, before or during the incubation; acquiring a second plurality of final images of the analysis container at a second acquisition time T2, during or after the incubation; - resetting each initial image of the first plurality of acquired initial images, with each corresponding final image of the second plurality of acquired final images; locating at least one potential microorganism growth zone within at least one image of the second plurality of acquired images; - assessment of the content of the potential microorganism growth zone identified to determine the presence of microorganisms.
    [0002]
    2. Method of determining the growth of microorganisms according to claim 1, the registration step comprising a primary registration step associated with an identifier located on the analysis container.
    [0003]
    A method of determining the growth of microorganisms according to claim 1 or 2, the registration step comprising a secondary registration step associated with the contents of the test container.
    [0004]
    4. Method for determining the growth of microorganisms according to one of the preceding claims, the step of locating at least one potential microorganism growth zone comprising a step of detecting a high density of colonies. of microorganisms, a step of detecting colonies of non-circular microorganisms and a step of assigning a locating element to said potential growth zone of microorganisms.
    [0005]
    5. Method for determining the growth of microorganisms according to one of the preceding claims, the step of evaluating the content of the potential growth zone identified comprising a step of determining the values of two evolution parameters.
    [0006]
    6. A method of determining the growth of microorganisms according to claim 5, the two evolution parameters comprising a correlation parameter and a contrast parameter.
    [0007]
    The method of determining the growth of microorganisms according to one of the preceding claims, wherein the secondary registration step comprises creating a sub-image mesh for each initial image and each final image.
    [0008]
    8. A system for determining the growth of microorganisms in a biological sample capable of containing microorganisms, said biological sample being contained in an analysis vessel such as a petri dish, said analysis vessel being incubated for a determined period of time, said system comprising: - an image capture device (12) for acquiring a plurality of images of the object to be analyzed at a first acquisition time T1, before or during incubation, and at a second T2 acquisition time, during or after incubation; an illumination device (14) comprising one or more light sources for illuminating the analysis container; a control device (18) for controlling the application of the method for determining the growth of microorganisms according to one of claims 1 to 7 in order to determine the presence of microorganisms. 3028866 P129079.EN.01 34
    [0009]
    9. Computer program product comprising software instructions for implementing a method according to one of claims 1 to 7 when said program is executed by a data processor.
    类似技术:
    公开号 | 公开日 | 专利标题
    EP3224372B1|2019-03-20|Method, system and computer program product for determining microorganism growth
    JP6062059B2|2017-01-18|Bioimaging method
    US10663712B2|2020-05-26|Methods and apparatus for detecting an entity in a bodily sample
    TWI499669B|2015-09-11|Method for detecting a microorganism, apparatus for detecting a microorganism, and program
    AU2013353154B2|2016-06-23|Cell monitoring device, cell monitoring method and program thereof.
    Xiao-bo et al.2010|In-line detection of apple defects using three color cameras system
    Blasco et al.2009|Recognition and classification of external skin damage in citrus fruits using multispectral data and morphological features
    Yoon et al.2010|Detection of Campylobacter colonies using hyperspectral imaging
    Li et al.2012|A machine vision system for identification of micro-crack in egg shell
    JP2018525746A5|2019-05-30|
    WO2012119191A1|2012-09-13|Method and software for analysing microbial growth
    Pandit et al.2014|Literature review on object counting using image processing techniques
    Li et al.2019|Detection of early decayed oranges based on multispectral principal component image combining both bi-dimensional empirical mode decomposition and watershed segmentation method
    KR20180068150A|2018-06-21|Classification system of dried sea cucumber using image analysis
    Bhardwaj et al.2015|An imaging approach for the automatic thresholding of photo defects
    Fang et al.2011|Detecting preserved eggshell crack using machine vision
    Lu et al.2015|Detecting defects on citrus surface based on circularity threshold segmentation
    FR3034103A1|2016-09-30|METHOD FOR MEASURING THE CONCENTRATION OF MICROORGANISMS IN A LIQUID
    Lu et al.2017|Detecting defects on citrus fruit surface by circularity gradient threshold segmentation
    Zakiyabarsi et al.2019|Crab Larvae Counter Using Image Processing
    Yang et al.2019|Automatic Recognition of Potato Germ Based on AdaBoost Algorithm
    CN111429361A|2020-07-17|Bacterial colony counting method based on machine vision
    FR3080211A1|2019-10-18|SYSTEM AND METHOD FOR ANALYZING TEST DEVICES
    WO2014048936A1|2014-04-03|Method for identifying cells in a biological tissue
    Belini et al.2019|Automatic cell segmentation from brightfield microscopy images of pseudohyphal cell-aggregates
    同族专利:
    公开号 | 公开日
    EP3224372B1|2019-03-20|
    EP3224372A1|2017-10-04|
    FR3028866B1|2018-03-09|
    US20170260564A1|2017-09-14|
    WO2016083744A1|2016-06-02|
    US10407708B2|2019-09-10|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US5510246A|1993-05-14|1996-04-23|Minnesota Mining And Manufacturing Company|Method for rapid quantification of microorganism growth|
    EP1061127A1|1998-02-03|2000-12-20|Hakuju Institute for Health Science Co., Ltd|Inspection method for microorganisms and the like, and unit therefor|
    WO2014059357A1|2012-10-11|2014-04-17|Institute For Systems Biology|Automated quantification of microorganism growth parameters through temporally resolved microscopic imaging|
    US5694478A|1994-12-15|1997-12-02|Minnesota Mining And Manufacturing Company|Method and apparatus for detecting and identifying microbial colonies|
    ES2430854T3|2001-09-06|2013-11-22|Rapid Micro Biosystems, Inc.|Rapid detection of cells in replication|
    EP2520923A1|2011-05-06|2012-11-07|bioMérieux|Bio-imaging method and system|AU2015212758B2|2014-01-30|2019-11-21|Bd Kiestra B.V.|A system and method for image acquisition using supervised high quality imaging|
    CA2985848A1|2015-04-23|2016-10-27|Bd Kiestra B.V.|Colony contrast gathering|
    US10696938B2|2015-04-23|2020-06-30|Bd Kiestra B. V.|Method and system for automated microbial colony counting from streaked sample on plated media|
    CN213537929U|2019-06-27|2021-06-25|Bd科斯特公司|System for locating bacterial colonies on a culture plate|
    法律状态:
    2015-10-30| PLFP| Fee payment|Year of fee payment: 2 |
    2016-05-27| PLSC| Publication of the preliminary search report|Effective date: 20160527 |
    2016-11-07| PLFP| Fee payment|Year of fee payment: 3 |
    2017-11-29| PLFP| Fee payment|Year of fee payment: 4 |
    2018-11-26| PLFP| Fee payment|Year of fee payment: 5 |
    2019-11-08| TP| Transmission of property|Owner name: BIOMERIEUX, FR Effective date: 20190930 |
    2019-11-27| PLFP| Fee payment|Year of fee payment: 6 |
    2020-11-26| PLFP| Fee payment|Year of fee payment: 7 |
    2021-11-22| PLFP| Fee payment|Year of fee payment: 8 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1461533A|FR3028866B1|2014-11-26|2014-11-26|METHOD, SYSTEM AND COMPUTER PROGRAM PRODUCT FOR DETERMINING THE GROWTH OF MICROORGANISMS|
    FR1461533|2014-11-26|FR1461533A| FR3028866B1|2014-11-26|2014-11-26|METHOD, SYSTEM AND COMPUTER PROGRAM PRODUCT FOR DETERMINING THE GROWTH OF MICROORGANISMS|
    PCT/FR2015/053221| WO2016083744A1|2014-11-26|2015-11-26|Method, system and computer program product for determining microorganism growth|
    US15/529,114| US10407708B2|2014-11-26|2015-11-26|Method and system for determining microorganism growth|
    EP15808742.9A| EP3224372B1|2014-11-26|2015-11-26|Method, system and computer program product for determining microorganism growth|
    [返回顶部]