Method and apparatus for determining the vertical position of a horizontally extending interface between a first component and a second component

Abstract

The present invention relates to a method and an apparatus for determining the vertical position (z_pv) of a horizontally extending interface (1) between a first component (2) and a second component (3), wherein the first component and the second component are contained in a laboratory sample container (4) in the form of layers vertically separated from each other, the method comprising the following steps: - generating first data, wherein the first data depend on the vertical position of the horizontally extending interface, - generating second data, wherein the second data is in the form of image data of the laboratory sample container, - determining a first probability distribution function (f1) in response to the first data, - determining a second probability distribution function (f2) in response to the second data, wherein the first probability distribution function and the second probability distribution function respectively assign a probability of the presence of the horizontally extending interface to the vertical position (z), and - determining the vertical position of the horizontally extending interface in dependence on the first probability distribution function and the second probability distribution function.

Description

Technical Field

[0001] The invention relates to a method and a device for determining the vertical position of a horizontally extending interface between a first component and a second component, wherein the first component and the second component are contained in a laboratory sample container in layers vertically separated from one another. Background Art

[0002] EP 2 770 317 A1 discloses a device for determining the vertical position of at least one interface between a first component and at least one second component, said components being comprised as different layers in a sample container or a sample tube. Summary of the invention

[0003] It is an object of the present invention to provide a method and a device for determining the vertical position of a horizontally extending interface between a first component and a second component, which method and device have improved properties compared to prior art solutions.

[0004] By means of the inventive method of the present invention, the vertical position of the interface between the first component and the second component can be determined. If the longitudinal axis of the laboratory sample container is perpendicular to the horizontal plane (i.e. when the laboratory sample container is arranged so that the laboratory sample contained in the laboratory sample container can be processed (e.g. pipetted, etc.), the first component and the second component are contained in the laboratory sample container in layers vertically separated from each other.

[0005] The laboratory sample container can be embodied as a conventional transparent laboratory sample container or as a laboratory sample tube used in an automated laboratory instrument. The method can be intended to detect the horizontal interface between different layers of a centrifuged blood sample, such as the interface between a serum or plasma layer and a separation medium layer or a coagulation (blood cell) layer.

[0006] The method comprises the following steps.

[0007] First data are generated, wherein the first data depends on the vertical position of the horizontally extending interface. The first data can be formed, for example, as a set of digital data characters, wherein, for a given vertical position, one of the set of digital data characters is specified, the digital data character depending on the nature of the component contained in the laboratory sample container at the given vertical position.

[0008] Generate second data in the form of digital picture data of a laboratory sample container containing the first component and the second component. The picture data can be embodied, for example, as raw picture data generated by means of a conventional image sensor, or can be embodied as picture data formed by compressing the raw data (e.g., in jpg format, etc.). The picture data may include color information of the laboratory sample container containing the first component and the second component.

[0009] A first probability distribution function is calculated in response to the first data, wherein the first probability distribution function assigns a probability of existence / occurrence of a horizontally extending interface to a vertical position.

[0010] A second probability distribution function is calculated in response to the second data, wherein the second probability distribution function assigns a probability of the presence of the horizontally extending interface to the vertical position. In order to determine or calculate the second probability distribution function, conventional image processing methods can be used. With the aid of image processing methods, for example, the properties of the first component, the properties of the second component, the properties of the laboratory sample container and / or the properties of the label / barcode attached to the laboratory sample container can be determined, wherein the second probability distribution function can be calculated or determined based on the following: the properties of the first component, the properties of the second component, the properties of the laboratory sample container, and / or the properties of the label / barcode attached to the laboratory sample container.

[0011] Finally, the vertical position of the horizontally extending interface is calculated depending on the first probability distribution function and depending on the second probability distribution function.

[0012] According to an embodiment, the first data are generated by sensing the transmittance through the laboratory sample container at different vertical positions. For this case, the first data may be formed, for example, as a set of digital data characters, wherein, for a given vertical position, one digital data character in the set of digital data characters corresponds to the transmittance at the given vertical position. The transmittance may be sensed for different wavelengths. The first probability distribution function may be based, for example, on the disclosure of EP 2 770 317 A1, in particular on claims 1 and 2, Figure 3 and Figure 4 and paragraph

[0049] to determine.

[0013] According to an embodiment, the vertical position of the horizontally extending interface is determined based on the first probability distribution function and the second probability distribution function by combining the first probability distribution function and the second probability distribution function based on Bayesian theorem, Dempster-Shafer theory and / or Kalman filtering. Reference is made to relevant technical documents.

[0014] According to an embodiment, the method comprises the following further steps: providing a test sample container, wherein the first test component and the second test component are contained in the test laboratory sample container in layers separated vertically from each other; determining the vertical position of the horizontally extending interface between the first test component and the second test component several times, for example 1000 times to 100000 times; determining a first frequency distribution of the determined vertical positions; and determining a first probability distribution function in response to the first data and in response to the first frequency distribution. The first data can be used, for example, to calculate the possible vertical positions of the horizontally extending interface between the first component and the second component based on the disclosure of EP 2 770 317 A1. Then, the first probability distribution function is determined by centering the first frequency distribution at the calculated (assumed) vertical position.

[0015] According to an embodiment, the method comprises the following further steps: providing a test sample container having a label attached to the test sample container; determining a geometric property of the label (e.g., the vertical position of a boundary of the label) several times, for example, 1000 to 100000 times; determining a second frequency distribution of the determined geometric property; and determining a second probability distribution function in response to the second data and in response to the second frequency distribution. The second data can, for example, be used to calculate the vertical position of the vertical boundary of the attached label, for example by means of image processing. The second frequency distribution can then be centered at the vertical position of the vertical boundary of the attached label. The resulting function can then be used to determine the second probability distribution function.

[0016] According to an embodiment, the method includes the following additional steps: providing a test sample container having a bar code label attached to the test sample container; determining a geometric property of the bar code label several times; determining a third frequency distribution of the determined geometric property; determining a second probability distribution function in response to the second data and in response to the third frequency distribution.

[0017] According to an embodiment, the method comprises the further step of generating second data comprising color information, wherein the second probability distribution function is determined in response to the color information.

[0018] According to an embodiment, the method comprises the further step of pipetting the first component and / or the second component in response to the determined vertical position of the horizontally extending interface.

[0019] The device is suitable for determining the vertical position of a horizontally extending interface between a first component and a second component, wherein the first component and the second component are contained in a laboratory sample container in the form of layers vertically separated from each other, and the device includes: a measuring unit, which is suitable for generating first data, wherein the first data depends on the vertical position of the horizontally extending interface; a digital camera, which is suitable for generating second data, wherein the second data is in the form of picture data of the laboratory sample container containing the first component and the second component; a processing unit, which is suitable for: determining a first probability distribution function in response to the first data, wherein the first probability distribution function assigns a probability of the existence of the horizontally extending interface to the vertical position; determining a second probability distribution function in response to the second data, wherein the second probability distribution function assigns a probability of the existence of the horizontally extending interface to the vertical position; and determining the vertical position of the horizontally extending interface depending on the first probability distribution function and the second probability distribution function.

[0020] According to an embodiment, the measuring unit is adapted to generate first data by sensing transmittance through the laboratory sample container at different vertical positions.The measuring unit may be embodied as disclosed in EP 2 770 317 A1 , in particular as disclosed in Fig. 1 of EP 2 770 317 A1 .

[0021] According to an embodiment, the device is adapted to perform the method as described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The present invention will now be described in detail with reference to the accompanying drawings, in which:

[0023] Figure 1a A test sample container is depicted, which contains a first test component and a second test component in layers vertically separated from each other, wherein a horizontally extending interface is formed between the first test component and the second test component,

[0024] Figure 1b depicts the frequency distribution of several determined vertical positions of the horizontally extending interface depicted in FIG1 , Figure 2 a depicts a laboratory sample container comprising a first component and a second component in layers vertically separated from each other, wherein a horizontally extending interface is formed between the first component and the second component, wherein a label is attached to the laboratory sample container, Figure 2 b depicts the first probability distribution function, where the first probability distribution function converts Figure 2 a, wherein the probability of the presence of the horizontally extending interface depicted in a is assigned to the vertical position, wherein the first probability distribution function is determined in response to first data generated by sensing the transmittance through the laboratory sample container at different vertical positions,

[0025] Figure 2c depicts the second probability distribution function, where the second probability distribution function converts Figure 2 The probability of the presence of the horizontally extending interface depicted in a is assigned to the vertical position, wherein the second probability distribution function is determined in response to second data in the form of picture data of a laboratory sample container.

[0026] Figure 3 a depicts a laboratory sample container comprising a first component and a second component in layers vertically separated from each other, wherein a horizontally extending interface is formed between the first component and the second component, wherein a label and a barcode label are attached to the laboratory sample container,

[0027] Figure 3 b depicts the first probability distribution function, where the first probability distribution function converts Figure 3 a, wherein the probability of the presence of the horizontally extending interface depicted in a is assigned to the vertical position, wherein the first probability distribution function is determined in response to first data generated by sensing the transmittance through the laboratory sample container at different vertical positions,

[0028] Figure 3 c depicts the second probability distribution function, where the second probability distribution function converts Figure 3 The probability of the presence of the horizontally extending interface depicted in a is assigned to the vertical position, wherein the second probability distribution function is determined in response to second data in the form of picture data of a laboratory sample container.

[0029] Figure 4 a depicts a laboratory sample container comprising a first component and a second component in layers vertically separated from each other, wherein a horizontally extending interface is formed between the first component and the second component, wherein a label is attached to the laboratory sample container,

[0030] Figure 4 b depicts the first probability distribution function, where the first probability distribution function converts Figure 4 a, wherein the probability of the presence of the horizontally extending interface depicted in a is assigned to the vertical position, wherein the first probability distribution function is determined in response to first data generated by sensing the transmittance through the laboratory sample container at different vertical positions,

[0031] Figure 4 c depicts the second probability distribution function, where the second probability distribution function converts Figure 4 a) assigns a probability of the presence of the horizontally extending interface depicted in a to the vertical position, wherein the second probability distribution function is determined in response to second data in the form of image data of a laboratory sample container, and

[0032] Figure 5 An apparatus for determining the vertical position of a horizontally extending interface between a first component and a second component is depicted highly schematically. DETAILED DESCRIPTION

[0033] Figure 1a A test sample container, i.e. a laboratory sample container for training purposes, is depicted, which comprises a first test component 2 (e.g., in the form of serum or an artificial equivalent) and a second test component 3 (in the form of a gel) in the form of layers vertically separated from each other, wherein a horizontally extending interface 1 is formed between the first test component 2 and the second test component 3.

[0034] For illustration purposes, the test sample container is depicted as being oriented horizontally. It goes without saying that the test sample container and the laboratory sample container 4 depicted in FIG. 1 and the following figures are actually oriented vertically, i.e., during the performance of the inventive method of the present invention, the longitudinal axis of the test container extends in the vertical direction denoted as z.

[0035] Based on the data generated by sensing the transmittance through the laboratory sample container 4 at different vertical positions z, the vertical position z_pv of the horizontally extending interface 1 is determined. Reference is made here to EP 2 770 317 A1. Without changing the measurement setup or the test setup, the vertical position z_pv of the horizontally extending interface 1 is determined repeatedly, for example 1000 times, wherein Figure 1b The resulting frequency distribution FD of the repeatedly determined vertical positions z_pv is depicted.

[0036] Figure 2 a depicts a laboratory sample container 4 containing a centrifuged blood sample, the blood sample comprising a first component 2 in the form of serum, a second component in the form of gel, and a third component 10 in the form of coagulated blood in layers vertically separated from each other. A horizontally extending interface 1 is formed between the first component 2 and the second component 3, wherein a label 5 is attached to the laboratory sample container 4.

[0037] Figure 2 b depicts the first probability distribution function f1, where the first probability distribution function f1 converts Figure 2 The probability P(z) of the presence of the horizontally extending interface 1 depicted in a is assigned to the vertical position z, wherein a first probability distribution function f1 is determined in response to first data generated by sensing the transmittance through the laboratory sample container 4 at different vertical positions z, as disclosed, for example, in EP 2 770 317 A1.

[0038] In order to determine or calculate the first probability distribution function f1, the first data are evaluated, wherein two vertical positions z_pv and z_pv2 are extracted as possible interface positions. The vertical position z_pv corresponds to the true vertical position of the interface 1. The vertical position z_pv2 corresponds to the vertical position of the vertical boundary of the tag 5, but with a reduced probability compared to the position z_pv. The vertical boundary of the tag 5 causes a change in the transmission, thereby potentially causing a false positive interface detection to occur.

[0039] After having calculated the two vertical positions z_pv and z_pv2 as possible interface positions, Figure 1b The frequency distribution FD depicted in is weighted with the probabilities at the corresponding positions z_pv and z_pv2, respectively, and is centered at the positions z_pv and z_pv2.

[0040] Figure 2 c depicts the second probability distribution function f2, where the second probability distribution function f2 converts Figure 2 The probability of the presence of the horizontally extending interface 1 depicted in a is assigned to the vertical position z, wherein a second probability distribution function f2 is determined in response to second data in the form of picture data of a laboratory sample container and its contents.

[0041] The second data are evaluated with the aid of image processing, which extracts the vertical position z_pv of the interface 1 and the vertical position z_pv2 of the vertical boundary of the label 5. The frequency distribution describing the changes when extracting the position of the interface 1 and the geometric properties of the label is weighted with the probabilities at the corresponding positions z_pv and z_pv2, respectively, and is centered at the positions z_pv and z_pv2. Finally, these values ​​are transformed into corresponding probabilities. Within the range of position z_pv, the probability of the existence of the horizontally extending interface 1 is positive, which has a maximum value at the position z_pv. Within the range of position z_pv2, the probability of the existence of the horizontally extending interface 1 is negative, which has a minimum value at the position z_pv2. It goes without saying that negative values ​​of the function f2 do not represent negative probabilities. On the contrary, negative values ​​are arithmetic values ​​representing unlikely interface positions.

[0042] In order to finally determine the vertical position z_pv of the horizontally extending interface 1 , the first probability distribution function f1 and the second probability distribution function f2 are combined, for example, based on the Bayesian theorem.

[0043] For example, using Gauβ distribution as the probability distribution function, according to the relationship between each measurement value (first data and second data) σ i The standard deviation of the combination σ comb Calculated as:

[0044]

[0045] Here n is the number of Gauβ distributions or sensors used. Thus, n>2 sensors or distributions from image processing devices can also be used to reduce the uncertainty of the measurement.

[0046] Since the vertical position z_pv of the horizontally extending interface 1 is determined based on two different probability distribution functions f1 and f2 (wherein the probability distribution functions f1 and f2 are based on different sensors), high-performance sensor fusion can be achieved, thereby providing more reliable interface detection compared to a solution using only one sensor (i.e., a transmission-based sensor or a camera-based sensor).

[0047] exist Figure 3 a / Figure 3 b / Figure 3 In the embodiment depicted in c, an additional barcode label 6 is attached to the laboratory sample container 4. As mentioned above, a first probability distribution function f1 is calculated. The barcode label 6 is detected by means of image processing, wherein a second probability distribution function f2 represents an unlikely probability within the range of the barcode label 6.

[0048] For example, a normalized Gauβ distribution with negative values ​​is placed so that the center is at the center of the barcode and the 3σ width has the width of the barcode. The negative Gauβ distribution can be normalized to a value between 0.1 and 0.8, in particular 0.3 or 0.5. This value can be derived from the generated model, i.e., depends on the frequency of the label edge coinciding with the serum separator / air interface. Alternatively, a constant negative probability with a value between 0 and 1, in particular 0.5, 0.75 or 0.9, can be specified along the barcode region.

[0049] exist Figure 4 a / Figure 4 b / Figure 4 In the embodiment depicted in c, the second probability distribution function f2 is generated additionally based on the color information in the second data. As depicted, if no color change occurs, the likelihood of the occurrence of the interface decreases.

[0050] In addition, different probabilities can be assigned to different color changes. For example, for the color change between air (transparent) and serum (e.g., pale yellow), a positive probability distribution according to the resolution of the measurement principle (e.g., camera resolution) can be assigned to the interface. Similarly, between serum and separation medium (almost white or blue or whatever color the separation medium has), a positive probability distribution can be assigned to be centered around the color change. The color of the separation medium can be identified by image processing of the entire tube, which is performed by the following steps: the type of tube is identified based on one or more or all of the length and width of the tube and the shape and color of the cap and the color of the liquid inside the tube. On the other hand, a negative distribution is assigned to the color change between the separation medium and the blood clot (dark red).

[0051] Figure 5 An apparatus for performing the method as described above is depicted.

[0052] The device 100 comprises a measuring unit 7a, 7b, 8a, 8b adapted to generate first data in the form of transmission data, wherein the first data depend on the vertical position z_pv of the horizontally extending interface 1 .

[0053] The device further comprises a camera 8 adapted to generate second data in the form of picture data of the laboratory sample container 4 containing the first component 2 and the second component 3 .

[0054] The device further comprises a processing unit 9 adapted to perform the inventive method of the present invention.

Claims

1. A method for determining the vertical position (z_pv) of a horizontally extending interface (1) between a first component (2) and a second component (3), wherein: The first component (2) and the second component (3) are contained in a laboratory sample container (4) in layers vertically separated from each other, and the method comprises the following steps: - generating first data, wherein the first data depends on the vertical position (z_pv) of the horizontally extending interface (1), - generating second data in the form of image data of the laboratory sample container (4) containing the first component (2) and the second component (3), - determining a first probability distribution function (f1) in response to the first data, wherein the first probability distribution function (f1) assigns a probability of the presence of the horizontally extending interface (1) to a vertical position, - determining a second probability distribution function (f2) in response to said second data, wherein said second probability distribution function (f2) assigns a probability of the presence of said horizontally extending interface (1) to a vertical position, and - determining a vertical position (z_pv) of the horizontally extending interface (1) depending on the first probability distribution function (f1) and the second probability distribution function (f2); The method further comprises the following steps: - providing a laboratory sample container (4), wherein the first test component (2) and the second test component (3) are contained in the laboratory sample container (4) in layers vertically separated from one another, - determining a vertical position (z_pv) of a horizontally extending interface (1) between the first test component (2) and the second test component (3) several times based on data generated by sensing the transmittance through the laboratory sample container (4) at different vertical positions, - determining a first frequency distribution (FD) of the determined vertical positions (z_pv), and - determining said first probability distribution function (f1) in response to said first data and in response to said first frequency distribution (FD).

2. The method according to claim 1, characterized in that - generating said first data by sensing the transmittance through said laboratory sample container (4) at different vertical positions.

3. The method according to any one of the preceding claims, characterized in that - performing a step of determining the vertical position (z_pv) of the horizontally extending interface (1) depending on the first probability distribution function (f1) and the second probability distribution function (f2) by combining the first probability distribution function (f1) and the second probability distribution function (f2) based on: - Bayes' theorem, and / or -Dempster-Shafer theory, and / or - Kalman filter.

4. The method according to claim 1 or 2, characterized in that: The method comprises the following steps: - providing a laboratory sample container (4) having a label (5) attached to the laboratory sample container (4), - determining the geometrical properties of said tag (5) several times, - determining a second frequency distribution of the determined geometric property, and - determining the second probability distribution function (f2) in response to the second data and in response to the second frequency distribution.

5. The method according to claim 1 or 2, characterized in that: The method comprises the following steps: - providing a laboratory sample container (4) having a barcode label (6) attached to the laboratory sample container (4), - determining the geometrical properties of said barcode label (6) several times, - determining a third frequency distribution of the determined geometric property, and - determining the second probability distribution function (f2) in response to the second data and in response to the third frequency distribution.

6. The method according to claim 1 or 2, characterized in that: The method comprises the following steps: - generating said second data including color information, - wherein said second probability distribution function (f2) is determined in response to said color information.

7. The method according to claim 1 or 2, characterized in that: The method comprises the following steps: - In response to the determined vertical position (z_pv) of the horizontally extending interface (1), the first component (2) and / or the second component (3) are pipetted.

8. A device (100) for determining the vertical position (z_pv) of a horizontally extending interface (1) between a first component (2) and a second component (3), wherein: The first component (2) and the second component (3) are contained in a laboratory sample container (4) in layers vertically separated from each other, and the device (100) comprises: - a measuring unit (7a, 7b, 8a, 8b) adapted to generate first data, wherein the first data depends on the vertical position (z_pv) of the horizontally extending interface (1), a camera (8) adapted to generate second data in the form of picture data of the laboratory sample container (4) containing the first component (2) and the second component (3), and - a processing unit (9) adapted to: - determining a first probability distribution function (f1) in response to the first data, wherein the first probability distribution function (f1) assigns a probability of the presence of the horizontally extending interface (1) to a vertical position, - determining a second probability distribution function (f2) in response to said second data, wherein said second probability distribution function (f2) assigns a probability of the presence of said horizontally extending interface (1) to a vertical position, and - determining a vertical position (z_pv) of the horizontally extending interface (1) depending on the first probability distribution function (f1) and the second probability distribution function (f2); The processing unit (9) is also suitable for - determining a vertical position (z_pv) of a horizontally extending interface (1) between the first component (2) and the second component (3) several times based on data generated by sensing the transmittance through the laboratory sample container (4) at different vertical positions, - determining a first frequency distribution (FD) of the determined vertical positions (z_pv), and - determining said first probability distribution function (f1) in response to said first data and in response to said first frequency distribution (FD).

9. The device (100) according to claim 8, characterized in that - the measuring unit (7a, 7b, 8a, 8b) is adapted to generate the first data by sensing the transmittance through the laboratory sample container (4) at different vertical positions z.

10. The device (100) according to claim 8 or 9, characterized in that - The device (100) is adapted to perform a method according to any one of claims 1 to 7.

Images