Biological sample measurement device and calibration method

By using a calibrator to determine distortion and liquid height conversion coefficients, the biological sample measuring device corrects for image distortion and misalignment, improving the accuracy of liquid volume measurements in biological samples.

WO2024236883A1PCT designated stage expired Publication Date: 2024-11-21HITACHI HIGH TECH CORP
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
PCT/JP2024/007835
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-17
Filing Date
2024-03-01
Publication Date
2024-11-21

AI Technical Summary

Technical Problem

Existing biological sample measuring devices face challenges in accurately measuring liquid volume due to misalignment and lens aberration, leading to decreased measurement accuracy.

Method used

The device acquires a distortion coefficient and liquid height conversion coefficient from a captured image of a calibrator with a pattern, correcting for image distortion and ensuring accurate liquid height measurement, thereby improving measurement accuracy.

Benefits of technology

This approach effectively avoids the decrease in liquid height measurement accuracy caused by device misalignment and lens aberration, enhancing the overall analysis accuracy of biological samples.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure JP2024007835_21112024_PF_FP_ABST
    Figure JP2024007835_21112024_PF_FP_ABST
Patent Text Reader

Abstract

The purpose of the present disclosure is to avoid deterioration in liquid level measurement accuracy due to deviation of a measurement device and lens aberration when measuring a biological sample having one or more component regions. A biological sample measurement device according to the present disclosure acquires a distortion coefficient and a liquid level conversion coefficient from a captured image of a calibrator having a pattern shape on a surface thereof, and corrects the liquid level of a sample on the basis of the distortion coefficient and the liquid level conversion coefficient (see FIG. 1).
Need to check novelty before this filing date? Find Prior Art

Description

Biological sample measuring device and calibration method

[0001] The present disclosure relates to an apparatus for measuring a biological sample separated into multiple component regions.

[0002] To improve the efficiency of clinical testing such as blood tests, there is a need for technology to automate the pre-analysis (pre-aliquoting) liquid volume confirmation process, which has traditionally been performed visually. The target specimens are separated into multiple layers by centrifugation or other methods, and technology is needed to measure only the liquid volume of the specimen to be analyzed, such as serum or plasma. Furthermore, when the specimen is a blood sample, the blood collection tubes that contain the specimen are affixed with barcode labels for patient identification, and technology is needed to enable measurement with the labels still attached. Furthermore, there is a demand for the option of installing a specimen measurement device on existing specimen transport equipment, which necessitates the miniaturization of the device.

[0003] As an example of such a biological sample measurement device, Patent Document 1 discloses a technology for detecting the height of a predetermined region of a sample separated into multiple layers by irradiating the sample separated into multiple layers with pulsed light of two wavelengths, which is switched in a time-division manner, and measuring the transmitted light while scanning the sample vertically.

[0004] Patent Document 2 discloses a technology for generating an image of a biological sample using a light source with two wavelength components emitted by the light source and an area camera, and then comparing the images generated using two or more exposure times or gains during imaging to select an image that most clearly identifies the portion of the biological sample to be measured.

[0005] US2012 / 0013889 JP2022-178784A

[0006] The technology disclosed in Patent Document 1 relates to a liquid volume measurement technology for measuring the liquid volume of a biological sample composed of multiple components by accurately identifying the boundary of a measurement target region from signals of transmitted light at two wavelengths that have different absorbances for the measurement target. However, Patent Document 1 requires measurement while vertically scanning the specimen (blood collection tube) in the longitudinal direction, making it difficult to miniaturize the device due to the need for a scanning mechanism.

[0007] The technology disclosed in Patent Document 2 relates to a liquid volume measurement technique that uses an area camera to capture near-infrared light transmission images, eliminating the need for sample rotation or vertical scanning mechanisms, even for labeled biological samples, and enabling the device to be more compact. This technique detects the area to be measured from the transmission image, determines the liquid height, and calculates the liquid volume. However, this requires a conversion coefficient to convert the number of pixels in the measurement area to length. Since the conversion coefficient is determined by the distance from the biological sample to the area camera, avoiding the effects of misalignment of the device due to installation or other factors is a challenge. Furthermore, images captured using an area camera are distorted due to lens aberration, and therefore, to accurately detect the liquid level, it is necessary to correct this distortion in the captured image.

[0008] The present disclosure has been made in consideration of the above-mentioned problems, and aims to avoid a decrease in the accuracy of liquid height measurement due to misalignment of the measuring device and lens aberration when measuring a biological sample having one or more component regions.

[0009] The biological sample measuring device according to the present disclosure obtains distortion coefficients and liquid height conversion coefficients from a captured image of a calibrator having a pattern shape on its surface, and corrects the liquid height of the sample based on these.

[0010] According to the biological sample measurement device of the present disclosure, when measuring a biological sample having one or more component regions, it is possible to avoid a decrease in liquid height measurement accuracy due to device misalignment and lens aberration, thereby improving analysis accuracy.

[0011] 6A is a diagram illustrating the configuration of a biological sample measuring device 101 according to embodiment 1. FIG. 6B is a diagram illustrating the principle of identifying a measurement target region. FIG. 6C is a diagram illustrating the principle of identifying a measurement target region. FIG. 6D is an example of a calibrator 301. FIG. 6E is an example of a calibrator 301. FIG. 6F is a diagram illustrating an example of a pattern 302. FIG. 6G is a diagram illustrating a method of imaging a calibrator 301. FIG. 6H is an explanatory diagram of a calibrator image 601 obtained by imaging a calibrator 301 with a constant light and dark stripe spacing using the configuration of FIG. 5. FIG. 6H is a diagram illustrating the distribution of the number of stripe spacing pixels [pixel] in the analysis range 602 of FIG. 6A when there is lens distortion aberration. FIG. 6B is a diagram illustrating the liquid height conversion coefficient [mm / pixel] obtained by dividing the actual stripe spacing length [mm] of FIG. 6B by the stripe period [pixel]. FIG. 6H is a flowchart illustrating the process by which a distortion coefficient calculation unit 109 acquires a distortion coefficient. FIG. 6I is a flowchart illustrating a method of determining a liquid height conversion coefficient after performing distortion correction on a calibrator image 601. FIG. 11 is a flowchart explaining a method for acquiring a liquid height conversion coefficient as a distribution (function) with vertical coordinates as variables, based on the distortion coefficient and distortion function acquired in FIG. 8. FIG. 12 is a flowchart explaining a liquid volume measurement method using the constant liquid height conversion coefficient acquired in FIG. 9. FIG. 13 is a flowchart explaining a liquid volume measurement method using the distribution of liquid height conversion coefficients acquired by the processing in FIG. 10. FIG. 14 is a diagram showing the configuration of a biological sample measuring device 101 pertaining to embodiment 2. FIG. 15 is a flowchart explaining a method for adjusting the imaging position of a biological sample 102 from the value of the liquid height conversion coefficient.

[0012] First Embodiment The biological sample to be measured in the embodiments of the present disclosure is assumed to be a labeled specimen before opening (before analysis) and separated into multiple component layers (typically 1 to 3 layers) by centrifugation or not. The measurement target is a portion of the specimen separated into multiple layers, such as plasma or serum, that is to be analyzed (dispensed) by a biochemical analyzer or the like.

[0013] FIG. 1 is a configuration diagram of a biological sample measurement device 101 according to a first embodiment of the present disclosure. The biological sample measurement device 101 is a device that measures a biological sample 102. The biological sample 102 is a sample composed of multiple layers as described above. The biological sample measurement device 101 identifies a measurement target region 103 in the biological sample 102 and measures its liquid volume. The biological sample measurement device 101 includes a surface illumination light source 104, an area camera 105, a time-sharing control driver 106, a calibration data acquisition unit 107, and a liquid volume measurement unit 108.

[0014] The surface illumination light source 104 is configured to be able to switch the wavelength of light emitted between two wavelengths and illuminate the biological sample 102 with that light. The light emitted by the surface illumination light source 104 can simultaneously illuminate two or more component layers constituting the biological sample 102 (i.e., across two or more component layers). Furthermore, it can illuminate the upper surface of the top layer (the boundary between the sample and the air layer) regardless of the component layer. Switching the wavelength of light does not necessarily require emitting light having only a single wavelength; it is sufficient to switch the wavelength component with the strongest intensity (e.g., wavelength λ1 = 970 ± 100 nm, wavelength λ2 = 1550 ± 100 nm). For biological samples 102 with one or three component layers, if the number of component layers is known prior to measurement, it is sufficient to use one of the two wavelengths, and there is no need to switch wavelengths. The reason for this and the method for selecting the wavelength will be explained in the measurement principle section below.

[0015] The area camera 105 captures an image of the surface illumination light transmitted through the biological sample 102, thereby generating a two-dimensional image of the biological sample 102. The area camera 105 has sensitivity characteristics that enable it to detect light in the wavelength band emitted by the surface illumination light source 104. The area camera 105 can be configured, for example, by an InGaAs camera. The area camera 105 also captures a calibrator image. Details of the calibrator will be described later.

[0016] The time-division control driver 106 switches the wavelength of light emitted by the surface illumination light source 104 in a time-division manner. In synchronization with the wavelength switching, the time-division control driver 106 adjusts the exposure time (or gain) of the area camera 105 to a time appropriate for that wavelength. The timing of image capture by the area camera 105 is controlled by the time-division control driver 106 in synchronization with the wavelength emitted by the surface illumination light source 104. The time-division control driver 106 may receive a processing result from the liquid volume measurement unit 108 and control re-imaging in accordance with the result.

[0017] The calibration data acquisition unit 107 analyzes the calibrator image acquired by the area camera 105. The calibration data acquisition unit 107 includes a distortion coefficient calculation unit 109, a liquid height conversion coefficient calculation unit 110, and a calibration data storage unit 111. The distortion coefficient calculation unit 109 calculates a distortion coefficient that represents image distortion caused by the influence of lens aberration. The liquid height conversion coefficient calculation unit 110 calculates a conversion coefficient that converts the number of pixels in the liquid height region of the image of the biological sample 102 acquired by the area camera 105 into a liquid height [mm]. The calculation methods for the distortion coefficient and liquid height conversion coefficient will be described later. The calibration data storage unit 111 stores the distortion coefficient acquired by the distortion coefficient calculation unit 109 and the liquid height conversion coefficient acquired by the liquid height conversion coefficient calculation unit 110.

[0018] The liquid volume measurement unit 108 extracts the measurement target region 103 from the captured image acquired by the area camera 105 and acquires the number of liquid height pixels [pixels] in the vertical direction corresponding to the liquid height in the captured image. The liquid volume measurement unit 108 accesses the calibration data storage unit 111 to acquire the stored distortion coefficient and liquid height conversion coefficient, and uses these to convert the number of liquid height pixels [pixels] acquired from the captured image to a liquid height [mm], and further calculates the liquid volume using information on the container diameter of the biological sample 102. Details of the processing performed by the liquid volume measurement unit 108 will be described later.

[0019] 2A and 2B are diagrams illustrating the principle of identifying a measurement target area. In the configuration of Fig. 1, an area camera 105 captures a two-dimensional transmission image of a biological sample 102, as shown in Fig. 2A and Fig. 2B.

[0020] 2A shows an example in which the biological sample 102 is separated into three layers. FIG. 2B shows an example in which the biological sample 102 is separated into two layers. The blood clot 202 is formed in the lower layer when the biological sample 102 is centrifuged. The separation material 201 is mixed in to separate the blood clot 202 from the measurement target region 103.

[0021] Comparing the first wavelength (wavelength 1) and the second wavelength (wavelength 2) emitted by the surface illumination light source 104, the transmittance when wavelength 1 passes through the blood clot 202 is almost the same as the transmittance when wavelength 2 passes through the blood clot 202. Therefore, there is only a slight difference between the image of the blood clot 202 acquired using wavelength 1 and the image of the blood clot 202 acquired using wavelength 2. Similarly, with respect to the separation material 201, the transmittance of wavelength 1 and the transmittance of wavelength 2 are almost the same, so there is only a slight difference between the two images.

[0022] In contrast, the transmittance when wavelength 1 passes through measurement target area 103 is significantly different from the transmittance when wavelength 2 passes through measurement target area 103. Therefore, there is a significant difference between the captured image of measurement target area 103 acquired using wavelength 1 and the captured image of measurement target area 103 acquired using wavelength 2. By identifying this difference, measurement target area 103 can be extracted from the captured image.

[0023] The wavelength bands used for wavelength 1 and wavelength 2 must be selected in advance so that a significant difference occurs between the wavelengths in the measurement target region 103, but almost no difference occurs in other areas, as illustrated in Figures 2A and 2B. As long as this condition is met, the specific wavelength values ​​may be arbitrary. That is, it is sufficient that at least the difference between the transmittance when wavelength 1 passes through the measurement target region 103 and the transmittance when wavelength 2 passes through the measurement target region 103 is greater than the difference between the transmittance when wavelength 1 passes through areas other than the measurement target region 103 and the transmittance when wavelength 2 passes through areas other than the measurement target region 103.

[0024] By using the above measurement principle, it is possible to identify the measurement target region 103 when the biological sample 102 is separated into multiple component layers. The liquid volume measurement unit 108 identifies the measurement target region 103 according to this principle. The number of component layers is not important, but typical biological samples such as plasma and serum are separated into one to three layers. In either case, the measurement target region 103 can be identified with high accuracy.

[0025] For biological samples 102 with one component layer and biological samples 102 with three component layers, if the number of component layers is known in advance of measurement, measurement can be performed using a transmission image of only one wavelength. In the case of three layers as shown in Figure 2A, a wavelength can be selected that exhibits a difference in absorptance between the measurement target region 103 and the separation material 201 or air layers located above and below the measurement target region 103. In the case of a single layer, a wavelength can be selected that exhibits a difference in absorptance between the measurement target region 103 and the air layer above the measurement target region 103.

[0026] The liquid volume measuring unit 108 detects the boundary of the measurement target area 103 obtained as described above, obtains the number of liquid height pixels corresponding to the liquid height, and calculates the liquid height [mm] using the obtained number of liquid height pixels and the liquid height conversion coefficient [mm / Pixel] stored in the calibration data storage unit 111 of Fig. 1. The method for calculating the liquid height conversion coefficient will be described later. The liquid volume measuring unit 108 calculates the liquid volume based on the liquid height [mm] obtained in this manner and diameter information of the blood collection tube.

[0027] <Embodiment 1: Purpose of Calibration> As mentioned above, in order for the biological sample measuring device 101 to determine the liquid volume, a liquid height conversion coefficient [mm / Pixel] is required to convert the number of liquid height pixels obtained from the captured image into liquid height [mm]. The liquid height conversion coefficient can be determined from the field of view [mm] captured by the area camera 105 and the number of sensor pixels of the area camera 105, but in order to determine the liquid height [mm] with high precision, the liquid height conversion coefficient must be determined accurately, which poses the following two challenges:

[0028] The first point concerns the deviation in the distance (working distance) between the imaging position of the biological sample 102 and the area camera 105. The field of view of the area camera 105 expands and contracts depending on the working distance, and the liquid height conversion coefficient is also determined by the working distance. If the imaging position of the biological sample 102 or the position of the area camera 105 differs from device to device due to installation or other reasons, the working distance will change and the value of the liquid height conversion coefficient will also differ, directly leading to a decrease in the accuracy of liquid height [mm] detection. Therefore, in order to more accurately determine the liquid height conversion coefficient, calibration is required to determine the liquid height conversion coefficient for each individual biological sample measuring device 101.

[0029] The second point concerns image distortion caused by lens aberration. When capturing an image using a general lens that is not a telecentric optical system, lens distortion aberration causes image distortion from the center to the edge of the image. As will be explained in more detail below, when such image distortion is present, the liquid height conversion coefficient will change depending on the pixel position. Therefore, to obtain an accurate liquid height conversion coefficient, it is necessary to detect image distortion and correct the liquid height conversion coefficient in accordance with the image distortion. In particular, because the biological sample measuring device 101 measures the liquid height in the measurement target area 103, correction for distortion in the liquid height direction is necessary.

[0030] To solve the above two problems, the calibration described below is performed for each individual biological sample measuring device 101. During calibration, with the aim of determining a highly accurate liquid height conversion coefficient, a distortion coefficient that represents image distortion and a liquid height conversion coefficient that calculates and corrects the liquid height based on the distortion coefficient are obtained for each device, and these are stored in the calibration data storage unit 111. The liquid volume measurement unit 108 obtains the distortion coefficient and liquid height conversion coefficient stored in the calibration data storage unit 111, and by using these in calculating the liquid height, it is possible to calculate the liquid height [mm] with high accuracy.

[0031] 3A and 3B show an example of a calibrator 301. The calibration data acquisition unit 107 acquires an image of the calibrator 301 captured by the area camera 105, and analyzes the shape pattern formed on the surface of the calibrator 301 to acquire a distortion coefficient and a liquid height conversion coefficient.

[0032] The calibrator 301 has a pattern 302 formed on a flat housing as shown in FIG. 3A or a cylindrical housing as shown in FIG. 3B. The pattern 302 in FIG. 3 is an example of a vertical black and white striped pattern. The shape of the calibrator 301 may be a flat plate, a cylinder, or a rectangular parallelepiped, but regardless of the shape, the pattern 302 is formed on the surface of the housing. If the shape of the calibrator 301 is the same as or similar to the shape of the container for the biological sample 102, the computational load for calculating the distortion coefficient and the liquid height conversion coefficient can be reduced.

[0033] The housing of calibrator 301 is made of a material such as resin or plastic that transmits at least one wavelength of the surface illumination light source 104 of the biological sample measurement device 101. Pattern 302 may be printed directly on the housing, or pattern 302 may be printed on paper and attached to the surface of the housing. It may also be formed on the housing by any other appropriate method. Since calibration data acquisition unit 107 analyzes the contrast of pattern 302 acquired by area camera 105, pattern 302 is composed of areas that transmit and areas that do not transmit the wavelength of the surface illumination light source 104. By calculating each coefficient using light that has transmitted through calibrator 301, the device settings (position of the object to be imaged, optical configuration, etc.) used when imaging calibrator 301 can be used as is when imaging the sample.

[0034] 4A and 4B are diagrams showing examples of a pattern 302. The duty ratios of light and dark are different between FIGS. 3A and 4A. The pattern 302 is a stripe pattern in FIG. 4A or a grid pattern in FIG. 4B, which have a vertical light and dark pattern. As long as the stripe spacing [mm] of the pattern 302 is known, it does not have to be a stripe or grid pattern, and the black and white duty ratio can be any. When a stripe or grid pattern with repeated light and dark at equal intervals is used, the amount of calculation required for the distortion coefficient acquisition process performed by the calibration data acquisition unit 107 can be reduced. A random pattern may also be used as long as it is a known pattern.

[0035] The printing range of the pattern 302 must be a range that can cover the vertical area of ​​the biological sample 102 to be imaged, and the pattern 302 must be present in an area that is at least wider than the height range of the measurement target area 103. Specifically, when the area camera 105 images a 90 mm range of a 100 mm long blood collection tube, the pattern 302 must be present across the entire vertical field of view of the area camera 105. If the range of the image analyzed by the liquid volume measurement unit 108 is narrower than the field of view of the area camera 105, the amount of calculation required for the distortion coefficient acquisition process by the calibration data acquisition unit 107 can be reduced by adjusting the printing range to the analysis range of the liquid volume measurement unit 108.

[0036] 5 is a diagram illustrating a method for capturing an image of the calibrator 301. The calibrator 301 is positioned parallel to the lens surface of the area camera 105, and the area camera 105 captures an image of the transmitted light when the calibrator 301 is irradiated with light from the surface illumination light source 104. Only one wavelength of light is required to be irradiated onto the calibrator 301. When using a surface illumination light source 104 that irradiates light of two wavelengths, such as 970 nm and 1550 nm, the contrast of the pattern 302 can be improved and the calibrator 301 can be analyzed with high accuracy by capturing an image of the calibrator 301 using light of the shorter wavelength (970 nm), which has less blurring due to chromatic aberration.

[0037] The calibrator 301 may be placed at any position as long as the distance from the area camera 105 to the calibrator 301 is known. If the placement position of the calibrator 301 coincides with the imaging position of the biological sample 102, the amount of calculation required to obtain the distortion coefficient and the liquid height conversion coefficient can be reduced.

[0038] <Embodiment 1: Method for correcting image distortion due to lens aberration> Figures 6A and 6B are diagrams illustrating distortion in a captured image. Figure 6A is an explanatory diagram of a calibrator image 601 captured using the configuration of Figure 5 of a calibrator 301 with a constant spacing between light and dark fringes. Figure 6B is a diagram showing the distribution of the number of pixels [pixels] between fringes in the analysis range 602 of Figure 6A when lens distortion aberration is present. The horizontal axis of Figure 6B represents the vertical coordinate of the calibrator image 601, and the vertical axis represents the number of pixels [pixels] between light and dark fringes.

[0039] Even when imaging a calibrator 301 with a constant stripe spacing in the light-dark pattern, if there is lens distortion, the stripe spacing will not be constant, and the number of pixels between stripes will decrease as you approach the edge of the image, as shown in Figure 6B. This distribution of stripe spacing in the calibrator 301 represents image distortion.

[0040] FIG. 7 illustrates the liquid height conversion coefficient [mm / pixel] calculated by dividing the actual stripe spacing length [mm] in FIG. 6B by the stripe period [pixel]. Because the number of stripe spacing pixels changes due to image distortion, the distribution of the liquid height conversion coefficient also depends on the vertical coordinate. In other words, the liquid height conversion coefficient varies depending on the liquid surface position of the biological specimen 102. If the liquid height conversion coefficient calculated from the stripe period at the image center is used directly to convert the number of liquid height pixels at the image edge, the liquid height accuracy will decrease. To accurately calculate the liquid height [mm], it is necessary to convert the number of liquid height pixels taking into account the distributions in FIGS. 6B and 7.

[0041] The distortion coefficient calculation unit 109 obtains a distortion coefficient that represents the distortion described above. The liquid height conversion coefficient calculation unit 110 obtains a liquid height conversion coefficient based on the distortion coefficient. A specific method for doing this will be described below.

[0042] FIG. 8 is a flowchart showing the process by which the distortion coefficient calculation unit 109 acquires distortion coefficients. The calibrator image 601 captured by the area camera 105 is acquired (S801). A vertical fringe period distribution, as shown in FIG. 6, is acquired from the luminance distribution of an analysis range 602 set in the central region of the image (S802). The width of the analysis range 602 is set to a maximum of approximately the diameter of a blood collection tube, and the fringe period distribution is acquired by averaging the horizontal pixel values ​​of the analysis range 602, thereby reducing the influence of noise. A distortion coefficient is acquired by fitting the acquired fringe period distribution to a distortion function (S803). Equation 1 is an example of a distortion function. yd is the distortion function, y is the vertical coordinate, and k1 and k2 are distortion coefficients. The distortion coefficient is set so as to minimize the error between Equation 1 and the fringe period distribution acquired in S802, and the acquired distortion coefficient is stored in the calibration data storage unit 111 (S804).

[0043] Since only distortion in the vertical direction (liquid height direction) affects the liquid height accuracy, distortion in the horizontal direction does not need to be taken into consideration, and the distortion function can be a single direction with the vertical coordinate as a variable. When detecting the blood collection tube diameter, as described below, a horizontal distortion function can be calculated in the same way, or the vertical distortion function can also be used as the horizontal distortion function. Calculating only a single-direction distortion function has the advantage of reducing the amount of calculation required.

[0044]

[0045] The liquid height conversion coefficient calculation unit 110 acquires the liquid height conversion coefficient based on the distortion coefficient stored in the calibration data storage unit 111. Figures 9 and 10 are flowcharts illustrating a method for determining the liquid height conversion coefficient. There are two methods for determining the liquid height conversion coefficient taking image distortion into consideration. The first method is to determine a constant liquid height conversion coefficient after correcting the fringe period distribution in Figure 6B by correcting the distortion of the calibrator image 601 (Figure 9). The second method is to determine the distribution of the liquid height conversion coefficient from the distortion function (Figure 10).

[0046] FIG. 9 is a flowchart illustrating a method for calculating a liquid height conversion coefficient after correcting distortion of the calibrator image 601. First, the calibrator image 601 and the distortion coefficients stored in the calibration data storage unit 111 are acquired (S901). Next, the pixel spacing is scaled according to a distortion function, such as Equation 1, defined using the distortion coefficients, thereby correcting distortion in the vertical direction of the calibrator image (S902). Next, the light-dark cycle distribution of the calibrator image 601 after distortion correction is acquired (S903). Through the process of S902, the light-dark cycle distribution of FIG. 6B acquired in S903 is flattened to match the distribution of the actual calibrator cycle. Finally, the actual calibrator cycle [mm] is divided by the light-dark cycle [pixel] to obtain a constant liquid height conversion coefficient [mm / pixel] (S904), which is then stored in the calibration data storage unit 111 (S905). In step S902, distortion correction is performed on the calibrator image 601. However, instead of the calibrator image 601, the vertical fringe period distribution may be corrected using a distortion function.

[0047] Fig. 10 is a flowchart illustrating a method for acquiring a liquid height conversion coefficient as a distribution (function) with vertical coordinates as variables, based on the distortion coefficients and distortion function acquired in Fig. 8. First, the calibrator image 601 and distortion coefficients are acquired (S1001), and a reference value for the liquid height conversion coefficient is acquired using the fringe period of the central region of the calibrator image 601, for example (S1002). Next, the distortion function of Fig. 8 is set based on the distortion coefficient, and the inverse of the distortion function is used to acquire a distribution of the liquid height conversion coefficient, correcting the reference value for the liquid height conversion coefficient depending on the vertical coordinate (S1003). The acquired distribution of the liquid height conversion coefficient is similarly stored in the calibration data storage unit 111 (S1004).

[0048] In either method shown in Figures 9 and 10, a liquid height conversion coefficient that takes image distortion into account can be obtained, and it is sufficient to implement and execute either method. However, depending on whether Figure 9 or Figure 10 is executed, the contents of the liquid height conversion coefficient to be stored will differ, and the flow of the liquid volume measurement process described below will also differ.

[0049] The above process is performed in advance as calibration for each device. When measuring the liquid volume of the biological sample 102, the distortion coefficient and liquid height conversion coefficient stored in the calibration data storage unit 111 are acquired and the liquid volume is measured. This makes it possible to avoid deviations in the working distance due to installation, etc., and a decrease in the accuracy of liquid height measurement due to image distortion.

[0050] <Embodiment 1: Liquid volume measurement method> Figures 11 and 12 are flowcharts illustrating the process of measuring liquid volume by the liquid volume measurement unit 108. Figure 11 shows a method that uses the liquid height conversion coefficient obtained by the method in Figure 9, and Figure 12 shows a method that uses the distribution of liquid height conversion coefficients obtained by the method in Figure 10. It is possible to switch between using Figure 11 and Figure 12 depending on the process content (Figure 9 or Figure 10) executed by the liquid height conversion coefficient calculation unit 110.

[0051] FIG. 11 is a flowchart illustrating a liquid volume measurement method using the constant liquid height conversion coefficient acquired in FIG. 9 . The liquid volume measurement unit 108 acquires an image of the biological sample 102 captured by the area camera 105 (S1101), accesses the calibration data storage unit 111, and acquires a distortion coefficient and a liquid height conversion coefficient (constant) (S1102). Next, the distortion coefficient is used to correct image distortion in the captured image of the biological sample 102 using the same method as S902 in FIG. 9 (S1103). Image distortion correction is performed at least in the vertical direction. When performing the detection of blood collection tube diameter described below, horizontal distortion correction is also performed similarly to the correction of the calibrator image 601. The measurement target region 103 is extracted from the biological sample image after image distortion correction by detecting the brightness gradient of the boundary based on the principle described in FIG. 2 (S1104). The number of pixels in the vertical direction of the extracted measurement region is set as the number of liquid height pixels, and the liquid height [mm] is obtained by multiplying the number of liquid height pixels by the liquid height conversion coefficient obtained in S1102 (S1105). The liquid volume is calculated using the liquid height [mm] and information on the blood collection tube diameter (S1106). The blood collection tube information may be obtained before measurement by information on the blood collection tube to be used, or may be obtained from an image of the biological sample 102 by a method described below.

[0052] 11 can also acquire a one-dimensional luminance signal by processing such as extracting a central region from the captured image of the biological specimen 102, and detect the boundary of the measurement target region 103. In this case, one-dimensional luminance signal acquisition processing may be performed on the distortion-corrected image output by the biological specimen image distortion correction processing (S1103), or one-dimensional luminance signal acquisition processing may be performed before S1103, and distortion correction in S1103 may be performed on the acquired one-dimensional luminance signal. By performing distortion correction and measurement target region 103 extraction processing on the one-dimensional luminance signal, the amount of calculation can be reduced.

[0053] FIG. 12 is a flowchart illustrating a liquid volume measurement method using the distribution of liquid height conversion coefficients obtained in the process of FIG. 10 . Unlike the process of FIG. 11 , the process of FIG. 12 does not perform distortion correction on the captured image of the biological sample 102. Instead, a measurement target area is extracted from the captured image obtained in S1201 using a process similar to S1104 of FIG. 11 (S1203). In S1204, a liquid height conversion coefficient corresponding to the coordinates (in pixels) of the measurement target area detected in S1203 is obtained from the liquid height conversion coefficient distribution obtained in S1202, and the liquid height [mm] is obtained by integrating the liquid height conversion coefficients within the range of the measurement target area in the vertical direction. As with the process of FIG. 11 , S1203 may also perform boundary detection on one-dimensional luminance signals. The liquid volume calculation process of S1205 is similar to S1106 of FIG. 11 .

[0054] By performing the above-described processing shown in FIGS. 11 and 12, image distortion can be corrected, and therefore the liquid height can be obtained and the liquid amount can be measured with high accuracy.

[0055] <Embodiment 1: Use of Blood Collection Tube Diameter> As a factor affecting the accuracy of the liquid height, deviations in the working distance, which affect the value of the liquid height conversion coefficient, have been described. However, in addition to deviations due to device installation, the working distance also varies depending on the blood collection tube diameter, which affects the accuracy of the liquid height measurement. For example, for a blood collection tube with a diameter of 13 mm and a blood collection tube with a diameter of 16 mm, the distance from the camera to the surface of the blood collection tube will differ by 1.5 mm. One way to avoid the effects of this deviation is to prepare a liquid height conversion coefficient corresponding to the blood collection tube diameter and use it depending on the blood collection tube diameter information obtained during measurement.

[0056] Possible methods for obtaining a liquid height conversion coefficient corresponding to the blood collection tube diameter include the following: (a) preparing cylindrical calibrators 301 as shown in FIG. 3B with the same diameter as the blood collection tubes, for each type of blood collection tube diameter, and performing calibration using each of them to obtain and store the distortion coefficient and liquid height conversion coefficient corresponding to the blood collection tube diameter; (b) taking into account the difference between the diameter of the calibrator 301 and the diameter of the actual blood collection tube, obtaining a scaled liquid height conversion coefficient according to the ratio of each working distance, and storing this in advance in the calibration data storage unit 111, or obtaining the scaled liquid height conversion coefficient each time the liquid volume is measured.

[0057] When a planar calibrator 301 such as that shown in Figure 3A is used, the liquid height conversion coefficient obtained from the calibrator 301 is scaled according to the ratio between the working distance when imaging the calibrator 301 and the working distance when imaging the biological sample 102, which is determined by the diameter of the blood collection tube, and the resulting liquid height conversion coefficient on the surface of the blood collection tube is obtained.

[0058] The liquid volume measurement unit 108 acquires blood collection tube information to calculate the liquid volume from the liquid height. The blood collection tube information may be stored in the liquid volume measurement unit 108 in advance, or the blood collection tube diameter may be detected for each measurement from a captured image of the biological sample 102. In either case, by acquiring the blood collection tube diameter at the time of measurement, a liquid height conversion coefficient corresponding to the blood collection tube diameter can be used. Methods for acquiring the blood collection tube diameter from a captured image of the biological sample 102 include detecting the boundary from the brightness difference between the blood collection tube and the background edge and then calculating the diameter.

[0059] 13 is a configuration diagram of a biological sample measuring device 101 pertaining to embodiment 2 of the present disclosure. In addition to the configuration described in embodiment 1, this embodiment is equipped with an imaging position adjustment section 1301, a gripping mechanism 1302, and a gripping mechanism control driver 1303. The other configuration is the same as embodiment 1.

[0060] The gripping mechanism 1302 is a mechanism that grips the biological sample 102 and the calibrator 301. The gripping mechanism control driver 1303 is a driver that controls the position of the gripping mechanism 1302 based on the output result of the imaging position adjustment unit 1301. The imaging position adjustment unit 1301 is composed of an imaging position storage unit 1304 and a biological sample imaging position acquisition unit 1305. The imaging position storage unit 1304 stores the imaging position of the calibrator 301 and the imaging position of the biological sample 102 acquired by the biological sample imaging position acquisition unit 1305. The biological sample imaging position acquisition unit 1305 acquires information on the imaging position of the calibrator 301 stored in the imaging position storage unit 1304 and information on the liquid height conversion coefficient stored in the calibration data storage unit, and acquires the imaging position of the biological sample 102 using these.

[0061] In Figure 13, the imaging position of the calibrator 301 and the biological sample 102 is adjusted using a gripping mechanism 1302, but the imaging position can also be adjusted by a method other than gripping, such as providing a jig for fixing the bottom of the biological sample 102 and the calibrator 301 and a rail for adjusting the position of the jig, and controlling the position of the jig with a control driver.

[0062] <Embodiment 2: Adjustment of the calibrator and biological sample imaging position> Since the liquid height conversion coefficient is determined by the working distance, in order to accurately determine the liquid height, it is necessary to match the imaging position of the pattern 302 surface of the calibrator 301 with the position of the measurement surface of the biological sample 102.

[0063] When using a cylindrical calibrator 301 having the same diameter as the blood collection tube to be measured, the imaging positions of the calibrator 301 and the biological sample 102 are aligned. The imaging position of the calibrator 301 stored in the imaging position storage unit 1304 is acquired by the biological sample imaging position acquisition unit 1305, and the imaging position of the biological sample 102 is sent to the gripping mechanism control driver 1303.

[0064] If the diameter of the calibrator 301 differs from the diameter of the blood collection tube, or if the position of the printed surface of the pattern 302 on the calibrator 301 is misaligned with the position of the measurement surface of the biological sample 102 due to differences in shape, such as a plate-shaped calibrator, the biological sample imaging position acquisition unit 1305 acquires an imaging position shifted by the amount of misalignment from the imaging position of the calibrator 301, and images the biological sample 102 at the acquired imaging position. This makes it possible to determine the liquid height using the distortion coefficient and liquid height conversion coefficient acquired by calibration as is. In this case, the liquid height conversion coefficient can also be corrected by scaling it up or down as described in embodiment 1.

[0065] Third Embodiment In the second embodiment, a method for adjusting the imaging position of the biological sample 102 based on the imaging position of the calibrator 301 was described. This method allows for accurate measurement of the liquid height using an accurate liquid height conversion coefficient. However, if the position of the area camera 105 or the gripping mechanism 1302 is shifted due to installation or other reasons, the working distance of the biological sample 102 will differ between devices. The difference in working distance will expand or contract the field of view of the area camera 105, resulting in different resolutions between devices.

[0066] In the third embodiment of the present disclosure, to address this issue, the imaging position of the biological sample 102 is adjusted based on the value of the liquid height conversion coefficient obtained by calibration in the configuration of Fig. 13. This allows a common working distance to be set between devices.

[0067] FIG. 14 is a flowchart illustrating a method for adjusting the imaging position of the biological sample 102 based on the value of the liquid height conversion coefficient. First, image 1 of the calibrator 301 is acquired at a first imaging position stored in the imaging position storage unit (S1401). Next, the gripping mechanism control driver 1303 moves the calibrator 301 to a second imaging position stored in the imaging position storage unit (S1402), and acquires image 2 (S1403). The second imaging position is shifted from the first imaging position in the depth direction relative to the area camera 105 (either in the positive or negative direction). For each of image 1 and image 2, a liquid height conversion coefficient is acquired using the method of embodiment 1. The acquisition method may be either the method shown in FIG. 9 or FIG. 10. In the case of FIG. 9, a constant liquid height conversion coefficient is acquired, and in the case of FIG. 10, a reference value for determining the distribution of the liquid height conversion coefficient is acquired (S1404). The biological specimen imaging position acquisition unit 1305 sets a linear function with the horizontal axis representing the coordinate of the imaging position and the vertical axis representing the liquid height conversion coefficient, based on the two acquired liquid height conversion coefficients (or reference values ​​of the liquid height conversion coefficients) and the coordinates of each imaging position (S1405). From the acquired linear function, the coordinate of the imaging position at which a liquid height conversion coefficient common to all devices can be obtained is acquired (S1406). The acquired imaging position is stored in the imaging position storage unit 1304 and used when measuring the biological specimen 102. By performing the above procedure during calibration for each device, the biological specimen 102 can be imaged at a working distance (and a common liquid height conversion coefficient) common to all devices.

[0068] The effect of capturing images at a common working distance is that images can be captured at the same resolution between devices, and lens focus deviations due to deviations in working distance can also be reduced.

[0069] <Regarding Modifications of the Present Disclosure> The present disclosure is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those including all of the described configurations. Furthermore, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, or to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0070] The calibration in the first embodiment may be performed for each apparatus, such as when the apparatus is installed, or may be performed using the calibrator 301 each time a measurement is performed.

[0071] In the first embodiment, a specific example of wavelengths used to obtain a transmission image was described assuming a specific configuration of the biological sample 102 (a serum sample). However, the present disclosure is not limited to this and can be applied to other types of samples. In other words, by appropriately selecting wavelengths for obtaining transmission images such as those illustrated in FIGS. 2A and 2B based on the absorbance characteristics of the sample, the present disclosure can be applied to other samples.

[0072] The calibration data acquisition unit 107, the liquid volume measurement unit 108, and the imaging position adjustment unit 1301 can be configured by hardware such as a circuit device that implements their functions, or by software that implements their functions being executed by a computing device such as a CPU (Central Processing Unit).

[0073] In the above embodiment, the calibrator 301 can be a component of the biological sample measuring device 101 , or can be configured as a separate member from the biological sample measuring device 101 .

[0074] In the above embodiments, the procedure for calculating the distortion coefficient and liquid height conversion coefficient does not necessarily have to be performed when measuring the biological sample 102; for example, it may be performed in advance before measuring the biological sample 102, and the results may be stored in the biological sample measuring device 101.

[0075] 101: Biological sample measuring device 104: Surface illumination light source 105: Area camera 106: Time-division control driver 107: Calibration data acquisition unit 108: Liquid volume measurement unit 301: Calibrator 1301: Imaging position adjustment unit

Claims

1. A biological sample measuring device for measuring a biological sample having one or more component regions, comprising: an imaging unit configured to generate a first captured image of the biological sample using light that has passed through the biological sample, and to generate a second captured image of the calibrator using transmitted light that has passed through a calibrator having a pattern shape formed on its surface; a distortion coefficient calculation unit that uses the second captured image to calculate a distortion coefficient for correcting distortion in the image captured by the imaging unit; a conversion coefficient calculation unit that uses the distortion coefficient to calculate a conversion coefficient for converting the number of pixels in the image captured by the imaging unit into length; and a liquid volume measurement unit that measures a liquid height of the biological sample by applying the conversion coefficient to the first captured image.

2. The biological sample measuring device according to claim 1, characterized in that: the liquid volume measuring unit is configured to identify a target portion of the biological sample to be measured from an image of the biological sample; the imaging unit generates a third image of the biological sample using a first wavelength component of the light that has passed through the biological sample; the imaging unit generates a fourth image of the biological sample using a second wavelength component of the light that has passed through the biological sample; the liquid volume measuring unit identifies an amount resulting from a difference between a portion of the third image generated using the first wavelength component and a portion of the fourth image generated by the second wavelength component; and the liquid volume measuring unit identifies a portion where the identified amount is equal to or greater than a threshold value, thereby identifying the range of the target portion.

3. The biological sample measuring device according to claim 1, characterized in that the pattern shape is made up of a plurality of shapes arranged along the liquid height direction of the biological sample.

4. The biological sample measuring device according to claim 1, further comprising a light source that irradiates the calibrator and the biological sample with a first light having a first wavelength component or a second light having a second wavelength component, the light source irradiates the calibrator with the first light or the second light, whichever has the shorter wavelength, and the imaging unit generates the second captured image using the light that has passed through the calibrator.

5. The biological sample measuring device according to claim 1, characterized in that the conversion coefficient calculation unit calculates the conversion coefficient for each position in the liquid height direction of the biological sample, and the liquid volume measurement unit measures the liquid height of the biological sample by applying the conversion coefficient corresponding to the position in the liquid height direction of the first captured image.

6. The biological sample measuring device according to claim 1, characterized in that: the pattern shape is composed of a plurality of shapes periodically arranged along the liquid height direction of the biological sample; the distortion coefficient calculation unit obtains the period of the pattern shape in the second captured image for each position in the liquid height direction; the distortion coefficient calculation unit obtains a distortion function representing the distortion by performing function fitting on the period of the pattern shape obtained from the second captured image; and the distortion coefficient calculation unit uses parameters constituting the distortion function as the distortion coefficients.

7. The biological sample measuring device according to claim 1, characterized in that the conversion coefficient calculation unit corrects the distortion of the second captured image using the distortion coefficient, and the conversion coefficient calculation unit calculates the conversion coefficient by dividing an actual size of the pattern shape of the calibrator by the number of pixels of the pattern shape of the corrected second captured image.

8. The biological sample measuring device of claim 1, characterized in that: the distortion coefficient calculation unit calculates the distortion coefficient for distortion that depends on the position of the biological sample in the liquid height direction; the conversion coefficient calculation unit obtains a reference value of the conversion coefficient by dividing the actual size of the pattern shape of the calibrator at a reference position of the second captured image by the number of pixels of the pattern shape of the second captured image; and the conversion coefficient calculation unit calculates the conversion coefficient that changes depending on the position in the liquid height direction by applying a function constructed using the distortion coefficient to the reference value.

9. The biological sample measuring device according to claim 1, characterized in that the liquid volume measuring unit corrects distortion of the first captured image using the distortion coefficient, and the liquid volume measuring unit measures the liquid height of the biological sample by applying the conversion coefficient to the corrected first captured image.

10. The biological sample measuring device according to claim 8, characterized in that the liquid volume measuring unit measures the liquid height of the biological sample by applying a value of the conversion coefficient calculated for each position in the liquid height direction, the value corresponding to the position in the liquid height direction of the first captured image.

11. The biological sample measuring device of claim 1, characterized in that the conversion coefficient calculation unit calculates the conversion coefficient for each size of the container using the second captured image of each of a plurality of calibrators, each of which has a different size for each size of the container that holds the biological sample, or the conversion coefficient calculation unit enlarges or reduces the conversion coefficient determined from the second captured image in accordance with a ratio between the working distance when the imaging unit captures the second captured image and the working distance when the imaging unit captures the first captured image.

12. The biological sample measuring device according to claim 1, further comprising a position adjustment unit that adjusts the position of the object imaged by the imaging unit, and the position adjustment unit adjusts the position of the biological sample so that the position of the pattern shape when the second image used to calculate the distortion coefficient and the conversion coefficient is captured is the same as the position of the measured surface of the biological sample when the first image to which the distortion coefficient and the conversion coefficient are applied is captured.

13. The biological sample measuring device according to claim 1, further comprising a position adjustment unit which adjusts the position of the object imaged by the imaging unit, wherein the imaging unit captures an image of the calibrator when the calibrator is in a first position, and the imaging unit captures an image of the calibrator when the calibrator is in a second position in the depth direction different from the first position, and the conversion coefficient calculation unit calculates, as the conversion coefficient, a first coefficient at the first position and a second coefficient at the second position, and the conversion coefficient calculation unit calculates a function expressing the relationship between the position of the calibrator and the conversion coefficient using the first coefficient and the second coefficient, and the conversion coefficient calculation unit identifies, from the relationships described by the function, a pair of the conversion coefficient and the position of the biological sample that is used in common for each individual biological sample measuring device, and the position adjustment unit places the biological sample at the position identified by the pair, and the liquid volume measurement unit uses the conversion coefficient identified by the pair.

14. The biological sample measuring device according to claim 1, wherein the container for holding the biological sample and the calibrator have the same shape or similar shapes.

15. A method for calibrating a biological sample measuring device that measures a biological sample having one or more component regions, comprising the steps of: generating an image of the calibrator using transmitted light that has passed through a calibrator having a pattern shape formed on its surface; using the captured image to calculate a distortion coefficient for correcting distortion in the image captured by the biological sample measuring device; and using the distortion coefficient to calculate a conversion coefficient for converting the number of pixels in the image captured by the biological sample measuring device into a length.

Citation Information

Patent Citations

  • Biological sample measuring device

    JP2022178784A

  • Method and Device for Determining the Position of a Boundary Between Two Phased in a Sample Tube

    US20120013889A1

  • Automatic water level identification method and device

    CN110287953A

  • Methods and apparatus adapted to identify 3D center location of a specimen container using a single image capture device

    WO2022174239A1