Automated blood drawing device and automated blood drawing method

Abstract

The present invention provides an automated blood drawing device and an automated blood drawing method whereby a more accurate puncture suitability value that accounts for the depth as well as the visibility of blood vessels may be calculated when a blood drawing site such as a finger or an arm is presented.　Provided is an automated blood drawing device for drawing blood by means of needle puncture of a biological site including a finger or an arm, said device comprising: a light reception unit that receives light in a wavelength range that can penetrate a living body to the maximum depth of blood vessels included in a target puncture range and light in a wavelength range that can penetrate only to the vicinity of the surface layer; an image acquisition unit that acquires measurement images, each of which is a measurement image of backscattered light in a respective wavelength range; an emphasized observation unit that selectively emphasizes and observes blood vessels that exist intermediately between the vicinity of the surface layer and the maximum depth on the basis of the difference between the measurement images for the respective wavelength ranges; a puncture suitability value calculation unit that calculates a puncture suitability value in the form of a value indicative of suitability or unsuitability for puncture on the basis of obtained characteristics of blood vessel density and / or thickness; and a puncture suitability determination unit that makes a puncture suitability determination on the basis of the puncture suitability value.

Description

Automatic blood sampling device and automatic blood sampling method

[0001] The present invention relates to a method and apparatus for automating blood sampling from a living body using a machine, and more particularly to an automatic blood sampling apparatus and method that automatically determine a needle insertion position suitable for obtaining a sufficient amount of blood.

[0002] Blood tests have become essential in medical care as they provide basic information for diagnosing various diseases. Demand for blood tests is expected to continue to increase in the future, as they are expected to be used for early diagnosis of disease signs in order to curb ever-increasing medical costs. With this growing demand, there are concerns about the increased burden on medical professionals who collect blood samples. Therefore, in recent years, attention has been focused on automated blood collection devices that mechanize blood collection with minimal manual intervention. In particular, there is a demand for small, space-saving blood collection devices that can be installed without taking up space in limited medical settings.

[0003] A conventional example of this type of automatic blood collection device is the device described in Patent Document 1. In this device, when a fingertip is presented at a specific position on the device, a needle automatically moves to puncture the fingertip, creating a small wound, and the blood that leaks out is collected in a collection container. By using the fingertip as the blood collection site, the device is made compact. However, while some fingertip areas are suitable for blood collection and others are not, the device described above does not assess suitability and only discloses mechanically puncturing the needle at a fixed position. As a result, individual differences in the finger and changes in the posture and position of the presented finger can potentially lead to insufficient blood collection. When manually collecting blood from the arm in medical settings, doctors and nurses typically search for a large vein that can be visually identified and then puncture that vein. While it is difficult to see the veins on the fingertip with the naked eye as on the arm, visualizing the veins in some way and targeting similarly large veins is thought to stabilize the amount of blood collected. Therefore, the device described in Patent Document 2 discloses a method of taking an image of the fingertip using a near-infrared camera that can see through blood vessels, identifying the blood vessel from the captured image through image analysis, and determining the clearest point on the blood vessel where blood is expected to be concentrated locally. This makes it possible to select a location where a large amount of blood can be expected to be collected, and perform blood collection, just as if a doctor or nurse were doing it visually.

[0004] Approaches aimed at improving blood collection volume have focused not only on the clarity of blood vessels but also on the depth at which the blood vessels are located. Because blood vessels run three-dimensionally within subcutaneous biological tissue, their depth varies depending on the location. When collecting blood, if the inserted needle does not accurately reach the blood vessel, it is natural that an adequate amount of blood cannot be collected. To address this issue, for example, Patent Document 3 discloses a method for measuring the depth of the blood vessel using stereoscopic vision with two cameras and controlling the puncture so that the needle reaches the blood vessel in accordance with the measurement data. This enables reliable blood collection by targeting the blood vessel.

[0005] Patent No. 6994910 JP 2023-049080 JP 8-168477

[0006] The above-mentioned prior art assumes that the target site and conditions for blood collection are accepted as presented and that the best blood collection is performed under those conditions. However, if the condition is unsuitable for blood collection, it may be preferable to take measures such as not collecting blood or changing the site from which blood is collected. In particular, when the fingertip is used as the target site for blood collection, since there are multiple fingers, even if one finger is unsuitable, another finger may be more suitable.

[0007] The automatic blood collection device described in Patent Document 2 also calculates a suitability index (hereinafter referred to as puncture suitability) based on the clarity of blood vessels, specifically the curvature described below, through image analysis to quantitatively determine whether a vessel is suitable for puncture. However, simply focusing on image features such as vascular clarity may result in a lack of correlation with blood collection volume due to factors other than those appearing as image features. It is necessary to calculate puncture suitability in a manner that removes factors that affect blood collection. One of the influencing factors is that even if the clarity of the vessel is sufficient, the vessel may be located deep. Even if the vessel is located deep, if its clarity is greater than its depth, the puncture suitability for that vessel will show a high value. However, in reality, blood collection itself is difficult due to the difficulty of reaching the needle, and the amount of blood collected is not as expected from the clarity of the vessel in the image. Another factor is that the signal-to-noise ratio of the captured image may be reduced due to the influence of the surrounding environment during image capture, particularly external light, which may prevent the actual clarity of the vessel from being accurately reflected in the image. For example, if ambient light is reflected onto the imaging surface, such as a fingertip, the signal-to-noise ratio of blood vessels will decrease relatively. If the puncture suitability is calculated while leaving the degradation of image quality due to such environmental factors as it is, a reversal will occur, such as a low puncture suitability but a large amount of blood being collected. Thus, in order to obtain a high correlation between puncture suitability and blood collection volume, it is necessary to make a comprehensive judgment that takes into account information other than that which has been focused on so far.

[0008] Regarding the depth of blood vessels, as mentioned above, Patent Document 3 discloses a method for an automatic blood collection device that reliably reaches a certain depth of a blood vessel by controlling the needle's depth. However, depending on the blood collection method, the needle does not necessarily reach a certain depth of the blood vessel. In particular, in the case of the lancet blood collection method disclosed in Patent Document 1, in which a needle is used to make a minute incision in the fingertip and collect blood that bleeds from the wound, needles that can only penetrate to a shallow depth of about 2 mm are often used. Furthermore, using a needle that can penetrate deep and monitoring and controlling the needle's penetration depth in real time during puncture requires precise needle control and highly accurate measurement, which complicates the device mechanism and creates problems in terms of cost and maintainability of the blood collection device. Therefore, when selecting a blood vessel to puncture, it is desirable to be able to determine that medium-sized and thick blood vessels in shallow layers that are easily reached by the needle are more suitable for puncturing than thicker and thicker blood vessels in deeper layers that are difficult to reach by the needle. To achieve this, a measurement and calculation method that prioritizes superficial blood vessels is required when calculating puncture suitability.

[0009] Regarding image quality, by evaluating the quality of the captured image itself before calculating the puncture suitability, it is possible to prevent erroneous calculations of the puncture suitability due to low image quality and to correct the puncture suitability to match the quality degradation. Therefore, a method is needed to detect degradation of image quality that may affect the calculation of the puncture suitability and to quantitatively evaluate the degree of degradation.

[0010] Therefore, the present invention provides an automatic blood sampling device and an automatic blood sampling method that can calculate a more accurate puncture suitability by taking into account not only the clarity of the blood vessels but also the depth, by simply presenting the blood sampling site such as a finger or arm.

[0011] In order to solve the above problems, the automatic blood sampling device of the present invention is an automatic blood sampling device that samples blood by needle puncturing a biological part including a finger or an arm, and is characterized by comprising: a light receiving unit that receives light in a wavelength range that can penetrate into the living body to the maximum depth of blood vessels included in the puncturing target area, and light in a wavelength range that can penetrate only to near the surface; an image acquiring unit that acquires measurement images of backscattered light in each wavelength range; an enhancement observation unit that selectively emphasizes and observes blood vessels that exist between the near the surface and the maximum depth based on the difference between the measurement images in each wavelength range; a puncturing suitability calculation unit that calculates an index value indicating whether or not puncturing is suitable from the obtained characteristics of the blood vessel density and / or thickness as puncturing suitability; and a puncturing suitability determination unit that determines whether puncturing is suitable based on the puncturing suitability.

[0012] Furthermore, the automatic blood sampling method according to the present invention is an automatic blood sampling method for sampling blood by needle puncturing a biological part including a finger or an arm, and is characterized by having: a light receiving step in which a light receiving unit receives light in a wavelength range that can penetrate into the living body to the maximum depth of blood vessels included in the puncturing target area, and light in a wavelength range that can penetrate only to near the surface; an image acquiring step in which an image acquiring unit acquires measurement images of backscattered light in each wavelength range; an enhancement observation step in which an enhancement observation unit selectively enhances and observes blood vessels that are located between the near the surface and the maximum depth based on the difference between the measurement images in each wavelength range; a suitability calculation step in which a puncturing suitability calculation unit calculates an index value indicating whether or not puncturing is suitable from the obtained characteristics of the blood vessel thickness and / or thickness as puncturing suitability; and a puncturing suitability determination step in which a puncturing suitability determination unit determines whether or not puncturing is suitable based on the puncturing suitability.

[0013] According to the present invention, it is possible to provide an automatic blood sampling device and an automatic blood sampling method that can calculate a more accurate puncture suitability by taking into account not only the clarity of the blood vessel but also the depth by presenting the blood sampling site such as a finger or an arm. Problems, configurations, and effects other than those described above will become clear from the description of the following embodiments.

[0014] 12 is a configuration diagram of an automatic blood sampling device according to a first embodiment of the present invention. It is a functional block diagram of the calculation unit shown in FIG. 1. It is a top view of the drive unit shown in FIG. 1. It is a diagram outlining a method for selectively measuring blood vessels in a shallow layer using the automatic blood sampling device according to the first embodiment of the present invention. It is a diagram showing an example of sensitivity characteristics in each RGB wavelength range. It is a diagram showing an example of a difference image when a color camera is placed directly under a finger. It is a flowchart showing the processing flow of the automatic blood sampling device according to the first embodiment of the present invention. It is a schematic explanatory diagram of a method for automatically adjusting an α value. It is a functional block diagram of a calculation unit constituting an automatic blood sampling device according to a second embodiment of the present invention. It is a flowchart showing the processing flow for image quality evaluation. It is a diagram showing an example of puncture point selection. It is a functional block diagram of a calculation unit constituting an automatic blood sampling device according to a third embodiment of the present invention. It is a schematic explanatory diagram of the processing of the continuity evaluation unit 19 and reliability calibration unit 20 shown in FIG.

[0015] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0016] FIG. 1 is a block diagram of an automatic blood sampling device according to a first embodiment of the present invention, and is a schematic diagram of the hardware configuration. As shown in FIG. 1, a calculation unit 101 built into the automatic blood sampling device 1 further comprises a CPU 102, memory 103, auxiliary storage 104, and interface (I / F) 105, which are interconnected by an internal bus. These may also be interconnected by a dedicated bus for purposes such as increasing speed. The CPU 102 is a calculation device that executes programs. This CPU 102 also executes the calculation process from an image of the puncturing suitability and reliability evaluation, which will be described later. The memory 103 stores the processing program itself and data required for processing. Data that needs to be permanently stored is written to and read from a recording medium such as a flash memory or hard disk, shown as auxiliary storage 104.

[0017] The interface (I / F) 105 connects the calculation unit 101 with various other functional blocks to exchange data. For example, when connected to the communication unit 109, the calculation unit 101 can perform various communications with the outside. The communication unit 109 allows some or all of the processing associated with blood collection to be shared with an external device, i.e., a host calculation device, or allows linkage with various services on the cloud.

[0018] A functional block for automatically collecting blood samples is connected to the calculation unit 101 via an interface (I / F) 105. The image input unit 106 is connected to a color camera 112, captures an image of a biological part 115 (a finger in this example) from which blood is to be collected, and sends the image data to the calculation unit 101 as digital image data. The image input unit 106 converts the output signal from the color camera 112 into a format that is easy for the calculation unit 101 to handle. However, the color camera 112 may have a built-in equivalent function, in which case the image input unit 106 simply functions as a data relay. The captured image is stored in a memory 103 and processed by a CPU 102. At this time, two light sources, a long wavelength light source 110 and a short wavelength light source 111, connected to a light source control unit 108, irradiate the finger (i.e., biological part 115) with light from the same side as the color camera 112. A program executed by the CPU 102 feedback-controls the output and balance of the long wavelength light source 110 and the short wavelength light source 111 to maximize the clarity of the vascular pattern in the captured image. The output of the long wavelength light source 110 and the short wavelength light source 111 is adjusted by varying the power supply level using PWM (Pulse Width Modulation) control or the like based on a setting value from the CPU 102. The long wavelength light source 110 and the short wavelength light source 111 are light sources in the long wavelength and short wavelength ranges, respectively, and LEDs (Light Emitting Diodes), which offer a high degree of freedom in wavelength selection, are suitable components for use. However, LEDs generally have strong directionality, and in the case of a reflected light imaging method in which light is irradiated from the same side as the color camera 112 as shown in FIG. 1, only a portion of the illuminated object tends to be illuminated with excessively strong light. Therefore, attaching a diffuser 113 to the long wavelength light source 110 and the short wavelength light source 111 to illuminate the illuminated object evenly over a wide area can result in images with less brightness unevenness. Furthermore, some of the light emitted from the long wavelength light source 110 and the short wavelength light source 111 does not penetrate into the inside of the finger, which is the biological part 115, but is directly reflected (specularly reflected) on the skin surface and reaches the color camera 112, which may cause noise during photography and degrade image quality. Therefore, a polarizing plate 114 is further added to the long wavelength light source 110 and the short wavelength light source 111, and a similar polarizing plate 114 is installed on the color camera 112 side so as to be out of phase with the light source side.The light emitted from the long-wavelength light source 110 and the short-wavelength light source 111 reaches the finger, which is the biological part 115, with its phase limited by the polarizing plate 114. However, while the light specularly reflected from the surface of the finger remains unchanged in phase, the light that penetrates into the finger and returns by backscattering is distorted in phase. Therefore, the polarizing plate 114 attached to the color camera 112 attenuates the specularly reflected light while allowing the backscattered light reflecting the blood vessels to pass through intact. This enables vascular imaging with reduced specular reflection. The captured image thus obtained is analyzed by the CPU 102, undergoes vascular pattern enhancement processing, and then calculates the puncture suitability and evaluates reliability. The most suitable point in the image is then selected, and once it is determined that puncture should be performed there, the coordinate values ​​of the puncture point are sent to the mechanism control unit 107, which controls the drive unit 116 to move the puncture needle 117 to the corresponding position on the finger, which is the actual biological part 115, and perform the puncture. Prior to this, calibration is performed in advance so that the coordinates of any point on the finger captured on the captured image match the actual point on the finger to be punctured by the puncture needle 117, and the correspondence is stored. After puncturing, the puncture needle 117 moves away from the finger, and the blood collection tube 118 receives the blood that has been bled, and blood collection is completed after waiting until a specified volume has been collected. Note that the puncture needle 117 may be a thin tubular needle similar to an injection needle, and may be connected to the blood collection tube 118 with a tube or the like, and the needle may be left punctured until the specified volume has accumulated in the blood collection tube 118.

[0019] 2 is a functional block diagram of the calculation unit shown in FIG. 2. As shown in FIG. 2, the calculation unit 101 includes an enhancement observation unit 11, a puncture suitability calculation unit 12, a puncture suitability determination unit 13, a puncture position selection unit 14, a blood vessel evaluation unit 15, a memory 103, an auxiliary storage 104, and an interface (I / F) 105, which are connected to one another via an internal bus. Here, the enhancement observation unit 11, the puncture suitability calculation unit 12, the puncture suitability determination unit 13, the puncture position selection unit 14, and the blood vessel evaluation unit 15 are realized by the CPU 102 shown in FIG. 1 and the memory 103 storing various programs. The enhancement observation unit 11 selectively enhances and observes blood vessels located between the surface layer and the maximum depth based on the difference between measured images of backscattered light in a wavelength range that can penetrate into the living body to the maximum depth of blood vessels included in the puncture target range and a wavelength range that can penetrate only to the surface layer, via the image input unit 106, which is an image acquisition unit. The puncturing suitability calculation unit 12 calculates an index value indicating whether puncturing is suitable based on the obtained blood vessel thickness and / or thickness characteristics as puncturing suitability. The puncturing suitability determination unit 13 determines whether puncturing is suitable based on the calculated puncturing suitability. The puncturing suitability determination unit 13 may be configured to obtain a correlation between puncturing suitability and blood collection volume by prior data analysis, and to perform blood collection only when the puncturing suitability indicates that a blood collection volume equal to or greater than a predetermined standard is expected. The puncturing position selection unit 14 selects the most suitable position for puncturing within the target biological region according to the level of the puncturing suitability. Furthermore, as shown in FIG. 2 , the blood vessel evaluation unit 15 is designed to indicate a higher or lower puncturing suitability value for each point on the image when the point is located on a blood vessel, the thickness and darkness of the blood vessel are more pronounced, and the blood vessel is located shallower.

[0020] FIG. 3 is a top view of the drive unit 116 shown in FIG. The drive unit 116 here is similar to the blood collection mechanism of the compact blood collection device disclosed in Patent Document 1. Here, the drive unit 116 is composed of a turntable rotated by a motor or the like and a mechanism for pushing the puncture needle 117 from the turntable surface toward the finger, which is the biological site 115, and has two degrees of freedom: rotation and pushing. The puncture needle 117 and the blood collection tube 118 are arranged side by side on the same circular orbit on the turntable. When the drive unit 116 rotates and the puncture needle 117 comes to rest directly below the finger, the puncture needle 117 is pushed out and punctures the finger. After puncturing, the turntable rotates, and the blood collection tube 118 comes to rest directly below the finger, where the bleeding blood is collected. With this drive system, blood collection can be achieved by simply providing two degrees of freedom: rotation of the turntable and pushing out the puncture needle 117. Another advantage is that consumables that are used only once per blood collection, such as the puncture needle 117 and the blood collection tube 118, can be easily managed together on the turntable surface. The simple structure is also advantageous for realizing a compact device. However, the range in which the puncture needle 117 can move is limited to a circular orbit.

[0021] Fig. 4 is a diagram showing an outline of a method for selectively measuring blood vessels in a shallow layer using the automatic blood sampling device 1 according to this embodiment. In Fig. 4, a visible light captured image 400 shows an example of capture when the color camera 112 is placed directly under the finger. The visible light captured image 400 is a color image, and the fingertip, which is the biological part 115, is captured in the center while being simultaneously illuminated by the long wavelength light source 110 and the short wavelength light source 111. Within the fingertip area, information about the fingertip surface, such as a fingerprint, and information about the blood vessel pattern, which is difficult to distinguish with the naked eye, are superimposed and observed.

[0022] In this embodiment, a visible light captured image 400, which is a color image, is first divided into a long-wavelength component image 402 and a short-wavelength component image 404. The long-wavelength component image 402 is primarily composed of light irradiated from the long-wavelength light source 110, penetrated into the finger, and returned as backscattered light, while the short-wavelength component image 404 is primarily composed of light irradiated from the short-wavelength light source 111 and similarly backscattered light. Generally, in a color camera 112, filters of the three primary colors of RGB are alternately arranged corresponding to each pixel of the image sensor according to a rule known as a Bayer array, so that the R component of RGB selectively transmits long wavelengths, the B component transmits short wavelengths, and the G component transmits intermediate wavelengths. This allows the intensity of each wavelength range to be converted into digital data as pixel brightness values ​​and reconstructed as an image. By utilizing this principle as it is, component images for each captured wavelength range are extracted.

[0023] Fig. 5 is a diagram showing an example of the sensitivity characteristics of each RGB wavelength range. As shown in Fig. 5, although there is overlap among the three wavelength ranges, if we focus on the two wavelength ranges of R and B, they can be considered to be almost independent. In other words, the biological information obtained by irradiation with the long wavelength range light source 110 (Fig. 1) is captured in the R component image, and the biological information obtained by irradiation with the short wavelength range light source 111 (Fig. 1) is captured in the B component image, each with high independence.

[0024] Of the component images obtained in this manner, the long-wavelength component image 402 ( FIG. 4 ) is in the red visible light wavelength range with a wavelength of approximately 620 nm, and has the property of relatively emphasizing the vascular pattern compared to the short-wavelength component image 404 ( FIG. 4 ). This is because the depth to which light incident on the skin can penetrate biological tissue varies depending on the wavelength, with longer wavelengths reaching deeper. Relatively large blood vessels, such as those targeted for puncture, are located in deeper layers and are difficult to observe without long-wavelength light. However, if the wavelength is too long, it will penetrate too deeply, clearly depicting blood vessels that the puncture needle 117 cannot easily reach. Specifically, light in the near-infrared region is said to penetrate to a depth of approximately 5 mm, depicting blood vessels located much deeper than the 2 mm puncture depth of the needle used in lancet blood collection methods. On the other hand, the penetration depth of visible light is approximately 2 mm to 3 mm, and blood vessels that are beyond the reach of the needle will not be depicted or will appear faint or unclear, even if they are actually thick and large. Since the density of blood vessels is an important parameter in calculating the puncture suitability, thin blood vessels have a low puncture suitability and are less likely to be selected as puncture points. Since the penetration depth varies depending on the wavelength even within visible light, the maximum depth of the blood vessel to be punctured can be freely determined by appropriately selecting the wavelength of the long-wavelength light source 110 (FIG. 1).

[0025] The other short-wavelength component image 404 is in the blue visible light wavelength range of approximately 470 nm, and has the property of relatively emphasizing skin surface patterns such as fingerprints and wrinkles compared to the long-wavelength component image 402 ( FIG. 4 ). This is because light in the short-wavelength range does not penetrate deep into the skin, but instead returns as backscattered light from the epidermis to the upper dermis, making it less susceptible to absorption by blood vessels inside the body. Here, the skin surface patterns are also reflected in the long-wavelength component image 402 ( FIG. 4 ), and when focusing on the observation of blood vessels, they are buried in noise and have low contrast. Therefore, by subtracting the short-wavelength component image 404 ( FIG. 4 ) from the long-wavelength component image 402 ( FIG. 4 ), a difference image 406 ( FIG. 4 ) is obtained, in which only the blood vessels are emphasized. This difference image (blood vessel image) 406 is adjusted so that only the superficial blood vessels appear darker, making it suitable as a base image for calculating the puncture suitability. The subsequent calculation process of the puncture suitability can be carried out basically in accordance with the procedure described in Patent Document 2.

[0026] FIG. 6 shows an example of a difference image 406 obtained when the color camera 112 is placed directly under the finger. As shown in FIG. 6 , the difference image (blood vessel image) 406 captures the fingertip, which is the biological part 115, at the center, and a blood vessel pattern 301 appears within the finger area. Points on this blood vessel pattern indicate areas with relatively high blood volume. Furthermore, blood vessels located deep within the puncture needle 117, which are defined by the wavelength set for the long-wavelength light source 110, are not clearly visible. The arcuate trajectory 300 in FIG. 6 indicates the range of movement of the puncture needle 117 in the structure shown in FIG. 3 . In this case, the optimal puncture point must be extracted from each point on this arcuate trajectory 300, and the intersection point with a blood vessel estimated to have a high blood volume is selected. Generally, because the blood vessel pattern 301 is formed in a mesh-like pattern, the arcuate trajectory 300 and the blood vessel pattern 301 often intersect at multiple points. Therefore, among these intersection points, the point with the highest puncture suitability is extracted to determine the actual insertion point for the puncture needle 117. In the above example, the movable range of the puncture needle 117 is limited to the arc trajectory 300, but the method for realizing the movement of the puncture needle 117 is not limited to the mechanism shown in Figure 3, and the movable range will change depending on the realization method. If the degree of freedom of movement of the puncture needle 117 is increased, the movable range will also be wider, and it is possible to make it a band-like range with a certain width rather than a line as described above. Of course, it does not have to be arc-shaped, and it can also be a straight line, a rectangle, or a fan shape.

[0027] FIG. 7 is a flowchart showing the processing flow of the automatic blood sampling device 1 according to this embodiment. As shown in FIG. 7 , the process begins in step S501 and includes various processes to be executed, as well as hardware initialization. In step S502, the color camera 112 is activated and begins capturing images. The captured images are stored in a memory area accessible by the CPU 102. In the following step S503, a process for detecting a body part 115, such as a finger, is executed. Specifically, image recognition is performed on the input image, particularly to detect significant image changes resulting from the presentation of a finger or arm to the device. In step S504, if the image change detected in step S503 is significant, it is determined that a body part 115 has been presented, and the process proceeds to step S505. In step S504, preparation for blood sampling begins, and a user is notified as necessary. A dedicated sensor can also be used to detect the body part. The method of notifying the user that a body part has been presented can be selected from among text or an icon displayed on a display or the like, an audio announcement, an alarm sound, or other methods optimal for the application environment. On the other hand, if the image change is not large, the process returns to step S502 and the subsequent processes are repeated.

[0028] In step S505, the long-wavelength light source 110 and short-wavelength light source 111 ( FIG. 1 ) for image capture are turned on, and the light intensity of the light source is adjusted until the finger, which is the biological part 115, is captured at an appropriate brightness. Whether the brightness is appropriate is determined by analyzing the image input each time and determining whether the average brightness of the finger area is within a predetermined range and whether there are any extremely bright or dark areas (step S506). In the following step S507, the balance of the light intensity of the long-wavelength light source 110 and short-wavelength light source 111 ( FIG. 1 ) is confirmed. While the outline of FIG. 4 outlines the process of calculating the difference between the long-wavelength and short-wavelength components, if the balance is poor, the difference value may be extremely small, resulting in an image with low gradation and insufficient information extracted from the difference. This makes it difficult to accurately calculate the puncture suitability. Therefore, the light intensity of each of the long-wavelength light source 110 and short-wavelength light source 111 ( FIG. 1 ) is controlled to increase the information content of the difference image (vascular image) 406. The amount of information can be easily determined from the average brightness of the difference image (blood vessel image) 406. However, if balance is given too much priority, one side may become too bright or too dark. Therefore, light intensity adjustment is performed by returning to step S502 and confirming that both light intensity and balance are satisfied. At this time, if excessive light intensity causes saturation of a portion of the target area in the image, no information on that portion may be obtained. Therefore, priority is given to controlling the amount of light to prevent excessive light intensity. In addition, some color cameras 112 allow color balance to be set using items such as color balance and color temperature, which can be used for adjustment as needed. Once light adjustment is completed in this manner, the process proceeds to step S509 to extract the optimal puncture point.

[0029] The first step S509 of the optimal puncture point extraction process is the separation of long and short wavelength components of the blood vessel. As described above, data for each RGB component captured in a Bayer array is arranged so that R and B each form a single image, and component images are formed as long and short wavelength components, respectively. That is, a long wavelength component image 402 (FIG. 4) and a short wavelength component image 404 (FIG. 4) are formed. Some color cameras 112 output color information packed in a format such as RGB, YUV, or YCbCr for each pixel or for each set of adjacent pixels. In such cases, the packed color information is converted to an RGB format with a clear correspondence to wavelength, and then the R component image and the B component image are separated and extracted. In the subsequent process 510, the difference between the R component and B component images is calculated. In this process, the luminance values ​​of the component images are not simply subtracted from each other, but weighted using a weight α to satisfy the following equation (1): diff = R - α × B (1). Since the brightness of the image portions relating to the skin surface contained in the long wavelength component image 402 and the short wavelength component image 404 differs between the component images, an appropriate α is set and the level is adjusted so that the skin surface image is offset by the difference. The value of α varies depending on the imaging conditions and the color of the presented finger, so it is desirable to automatically adjust it appropriately according to the image being captured.

[0030] Figure 8 is a schematic diagram illustrating the automatic α value adjustment method. The top diagram in Figure 8 illustrates the above equation (1). The bottom diagram shows three vertically arranged example difference images obtained when the value of α in equation (1) is changed. The top image 410 is the image obtained when α = 0, and as is clear from equation (1), it is completely identical to the long-wavelength component image 402. The bottom image 414 is the image obtained when α = 1, resulting in R-B, i.e., a simple difference between the long-wavelength component image 402 and the short-wavelength component image 404. Image 412 shows the state in which an appropriate α value m has been found, removing the image components of the skin surface and emphasizing only the image components of the blood vessels. Note that the image components of the skin surface are not smooth and uniform; the presence of a fingerprint or other object causes subtle variations in brightness. Obtaining a cross-sectional brightness profile for a portion of the finger region in the difference image (α = 0) 410 yields a profile curve 416, for example, which shows high-low brightness variations. On the other hand, if we calculate a profile of the same portion of the difference image (α = 1) 414, we obtain a profile that resembles the upside-down version of profile curve 416, as shown in profile curve 420, which exhibits high-low brightness variations. This inversion occurs because an excessively large value of α causes concave portions to become overfilled and convex. Conversely, if α is set to an appropriate value, the unevenness of the skin surface disappears, resulting in a smooth profile. Therefore, when the curvature is calculated for the concave or convex portion of the profile, the optimal value m is the value at which this value becomes zero. Here, the curvature represents the rate of change in brightness at each point of the cross-sectional profile, taking a maximum positive value at the bottom of the concave portion where brightness is lowest and a minimum negative value at the peak of the convex portion where brightness is highest. The higher the curvature, the sharper the image change, which can be said to result in a fingerprint or blood vessel with higher definition. Here, the curvature C(x) at point P(x) on the x-coordinate on the profile can be expressed as follows, where δ is the width of the irregularity to be calculated: C(x) = {P(x + δ, y) - P(x, y)} + {P(x - δ, y) - P(x, y)} ... (2) In this case, the change in the curvature from α = 0 to α = 1 can be roughly approximated by a linear line, and by finding the curvature when α = 0 and the curvature when α = 1, it is possible to easily find α = m when the curvature becomes 0.However, since it is uncertain to use the change in curvature at only one point as the α for the entire image, α is calculated from the change in curvature at each point within the finger region, and the average is used as the α for the entire image. Since a statistically stable value is required, sampling can be appropriately thinned out as long as a sufficient number of samples can be obtained. When calculating the curvature, setting δ in the above formula (2) to a value that easily captures brightness changes caused by fingerprints makes it easier to obtain results that meet the objective of vascular enhancement. However, some brightness changes in the image are caused by vascular patterns, and it is possible that the above formula (2) may pick up such brightness changes and reflect them in the curvature. Therefore, by excluding brightness changes in the area near the blood vessels, the ability to remove image components on the skin surface can be improved, thereby enhancing the effectiveness of vascular enhancement. However, since it is difficult to extract the vascular pattern in the subtraction image (α=0) 410 or the subtraction image (α=1) 414, α is first calculated from the inflection rates of all points in the finger region, and then a provisional vascular-enhanced image 412 is synthesized using this α, and vascular extraction is performed from this vascular image. After identifying the region near the blood vessels using the image in which the blood vessels are relatively clearly visible, the subtraction image (α=0) 410 and the subtraction image (α=1) 414 are re-calculated for the inflection rates of each point in the finger region excluding the region near the blood vessels, and α is derived from the average of these curvature rates. This allows the enhancement observation unit 11 ( FIG. 2 ) to more appropriately remove information about the skin surface and obtain a less-noise subtraction image (α=m) 412 (vascular-enhanced image), which is expected to improve the accuracy of calculating the puncture suitability.

[0031] Returning now to the description of FIG. 7 , in step S511, the puncture suitability calculation unit 12 ( FIG. 2 ) calculates the puncture suitability of each point within the finger region for the difference image (α=m) 412 (vessel-enhanced image) obtained by the enhancement observation unit 11 ( FIG. 2 ). To calculate the puncture suitability, in this embodiment, the puncture suitability calculation unit 12 ( FIG. 2 ) extracts a vascular pattern from the difference image (α=m) 412 (vessel-enhanced image) and calculates the puncture suitability for each point on the center line based on the density and / or thickness and length of the blood vessels. This can be indexed by analyzing image features such as the brightness level at each point, the width between vascular boundaries, and the continuity of the blood vessels as lines. To simply determine the optimal puncture point on the trajectory of the puncture needle 117 shown in FIG. 6 , the puncture suitability is calculated for only points that intersect with blood vessels on the trajectory. In step S512, the puncture suitability of each point thus obtained is compared, and the puncture suitability calculation unit 12 (FIG. 2) extracts the point that shows the maximum suitability within the movable range of the puncture needle 117.

[0032] In this embodiment, the objective is not simply to find the optimal puncture point in the captured image, but also to estimate the expected blood collection volume and determine whether puncture is feasible. To obtain the data on which the blood collection volume is based, a sufficient number of subjects are recruited in advance to ensure statistical significance. For the actual puncture points, the puncture suitability is calculated from images taken immediately before puncture, and a correlation is obtained between the puncture suitability and the blood collection volume obtained by puncturing. This correlation data is used to determine a regression equation between the puncture suitability and the blood collection volume, and this equation is used to estimate the blood collection volume from the puncture suitability. If the correlation coefficient is not sufficiently high, the estimated value will be uncertain. Therefore, the internal parameters used in calculating the puncture suitability are adjusted to increase the correlation coefficient. For example, by changing the weighting of features such as blood vessel density, thickness, length, and depth, the characteristics of the puncture suitability can be flexibly changed, and a combination of parameters with a high correlation with the blood collection volume is ultimately selected. If the puncturing suitability determination unit 13 determines that the amount of blood to be collected estimated from the obtained puncturing suitability does not reach a preset amount, it quantitatively determines whether to postpone blood collection or change the target area. If the puncturing suitability determination unit 13 determines that blood collection is appropriate, in step S513 the puncturing position selection unit 14 controls the drive unit 116 ( FIG. 1 ), which moves the needle to and punctures the actual body area corresponding to the optimal puncturing point selected above. This initiates blood collection, and ends when the required amount of blood is collected. Then, in step S514, the automatic blood collection device 1 is returned to its initial state, preparing for blood collection from the next user.

[0033] As described above, according to this embodiment, by calculating the puncture suitability based on an index that reflects the depth in addition to the thickness and density of the blood vessel, it becomes possible to determine the suitability of puncture with a higher correlation with the amount of blood to be collected.

[0034] In this embodiment, the long wavelength light source 110 and the short wavelength light source 111 are simultaneously irradiated to separate components from a color image. However, the present invention is not limited to simultaneous irradiation. Alternatively, the long wavelength light source 110 and the short wavelength light source 111 can be alternately irradiated to obtain two types of captured images, which can then be used as the long wavelength component image 402 and the short wavelength component image 404. In this case, overlap between the long and short wavelength regions does not occur in principle, allowing for more independent image analysis. However, with alternate irradiation, the target biological part 115 may move or fluctuate between the two captures. This fluctuation must be corrected before using the two long wavelength component image 402 and the short wavelength component image 404. Furthermore, instead of using two light sources, the long wavelength light source 110 and the short wavelength light source 111, a white light source capable of irradiating light with a wider wavelength range can also be used. Only one light source is required, and conditions such as the spread of the light beam can be matched for both the long and short wavelength components. However, if the wavelength range of the white light source is too broad in the long wavelength range, the penetration depth into the skin will be too great, which contradicts the purpose of the present invention, which is to selectively measure only shallow blood vessels. Therefore, when using a white light source, the characteristics of its long wavelength range must be carefully considered. To ensure uniform luminous flux conditions, it is possible to use a light source component that incorporates multiple long and short wavelength light-emitting elements in the same package. Furthermore, a multi-wavelength light source with additional long wavelength light-emitting elements can also be used. By adjusting the emission wavelength in the long wavelength range according to the blood vessel depth to be measured, it is possible to select the optimal wavelength according to the depth that the puncture needle can reach.

[0035] The puncture suitability is calculated for each point on the image by the vascular evaluation unit 15, which is designed to indicate a higher or lower value the more prominent the thickness and darkness of the blood vessel and the shallower the blood vessel is located on the image. Furthermore, a more stable and accurate calculation of the puncture suitability is achieved by using the total output value of the vascular evaluation unit 15 for each point in the entire neighborhood centered on the point to be calculated. Furthermore, when calculating the total value for the neighborhood, the vascular evaluation unit output value is weighted by the distance from the center of the neighborhood, so that points with high vascular evaluation unit output values ​​concentrated near the center indicate a higher suitability.

[0036] As described above, this embodiment makes it possible to provide an automatic blood sampling device and an automatic blood sampling method that can calculate a more accurate puncture suitability by taking into account not only the clarity of the blood vessels but also the depth of the blood vessels when the blood sampling site on the finger, arm, etc. is presented. Furthermore, this embodiment makes it possible to determine whether puncture is appropriate with a higher correlation to the amount of blood to be sampled.

[0037] FIG. 9 is a functional block diagram of a calculation unit 101a constituting an automatic blood sampling device according to this embodiment. As shown in FIG. 9, the calculation unit 101a according to this embodiment differs from the first embodiment in that it further includes a reliability evaluation unit 16, a usage restriction unit 17, and a puncturing suitability correction unit 18. Hereinafter, components similar to those in the first embodiment are designated by the same reference numerals, and redundant description will be omitted. In this embodiment, we focus on the need to minimize the impact of image quality degradation in order to achieve a puncturing suitability that is highly correlated with the amount of blood collected. Therefore, by pre-evaluating the quality of the captured image itself before calculating the puncturing suitability, we can prevent erroneous calculation of the puncturing suitability due to low image quality and correct the puncturing suitability to a level that is commensurate with the quality degradation. Such image quality evaluation is effective not only for calculating the puncturing suitability based on superficial vascular measurement using visible light as described above, but also for vascular extraction using near-infrared light as described in Patent Document 2.

[0038] As shown in FIG. 9 , the calculation unit 101a constituting the automatic blood sampling device according to this embodiment further includes a reliability evaluation unit 16, a usage restriction unit 17, and a puncturing suitability correction unit 18. The reliability evaluation unit 16, usage restriction unit 17, and puncturing suitability correction unit 18 are realized by the CPU 102 shown in FIG. 1 and memory 103 storing various programs. The reliability evaluation unit 16 quantitatively determines the image quality of the blood vessel at the candidate puncturing position in the captured blood vessel image using, for example, the curvature of brightness, as the reliability. If the determined reliability is low, the usage restriction unit 17 also determines the puncturing suitability as low reliability and restricts its use. The puncturing suitability correction unit 18 corrects the puncturing suitability in accordance with the reliability.

[0039] FIG. 10 is a flowchart showing the processing flow for image quality evaluation. Image quality evaluation is performed on the original image as captured to prevent fundamental factors from being obscured by other image processing. In step S601, initialization processing for the image quality evaluation process is performed. Next, in step S602, the finger region is extracted, and in the following step S603, the vascular pattern is extracted. The focus on the vascular pattern is because the effects of image quality degradation are most readily apparent in the vascular region, and blood vessels in regions with degraded image quality tend to lose clarity (sharpness). While the vascular pattern is also extracted as described above when calculating the puncture suitability, in this case, the vascular pattern can be extracted more reliably by performing a smoothing process or normalizing the extraction process to reduce brightness unevenness. In particular, measurement of the vascular pattern using visible light tends to be noisy because it deals with relatively small amounts of difference. Therefore, performing a process to suppress noise and emphasize the vascular region using smoothing or other processes is highly effective. However, the vascular pattern thus obtained is subject to general noise and other image quality degradation factors specific to the target captured image, making accurate sharpness evaluation difficult. Therefore, it is necessary to perform vascular pattern extraction from the original image separately from the vascular pattern extraction performed in the puncture suitability calculation. The reliability evaluation unit 16 then calculates a sharpness index value (contrast index value on the blood vessels) for each point on the extracted blood vessels in the original image (step S604). Here, the curvature of each point is used as the simplest index. The sharper the blood vessels, the more abrupt the change in luminance in the blood vessel portion, resulting in a larger curvature value. In this case, the original image, as is, is somewhat noisy and rough, and the curvature values ​​may vary significantly between adjacent points. Therefore, in image quality evaluation, the sum of the curvatures included in a certain range of areas near the target point is calculated, and this is used as the image quality evaluation value for the target point (step S605). In this embodiment, the sum of the curvatures included in a certain range of area near the target point is used, but the maximum value, minimum value, median, etc. may also be used depending on the tendency of the image characteristics obtained.In step S606, the obtained image quality evaluation value is sent to a higher-level execution program as the reliability of the target point, and finally a finalization process is performed. The higher-level program may be the program shown in the processing flow of Fig. 7 described in the above-mentioned first embodiment, and can evaluate image quality when a puncture candidate point is selected (after step S512 in Fig. 7), and if the quality is low, it can determine that the calculation result of the puncture suitability is also low in reliability and not to be adopted.

[0040] 11 is a diagram showing an example of puncture point selection. This example shows a case in which the point most suitable for puncturing is selected on the trajectory of the puncture needle 117. Puncture suitability is calculated for multiple points on the trajectory of the puncture needle 117 that intersect with blood vessels. If the reliability of the point with the highest puncture suitability value is lower than a predetermined threshold T, the point with the highest puncture suitability among the second and lower ranking points, whose reliability is equal to or greater than the predetermined threshold T, is selected as the puncture candidate point. This makes it possible to select a puncture candidate based on a highly reliable determination result.

[0041] Although the image quality evaluation or reliability calculation described above focuses on the clarity of blood vessels, a similar evaluation may be performed by focusing on differences in the clarity of patterns on the skin surface, such as fingerprints. In the case of the skin surface, for example, when light reflection occurs, changes in image features may occur, such as excessive emphasis on the unevenness of the surface. Therefore, it is possible to determine that the image quality is low when there are many points with extremely high curvatures.

[0042] As described above, according to this embodiment, in addition to the effects of Example 1, it is possible to eliminate the disturbance of the puncturing suitability due to unreliable information and to increase the correlation with the amount of blood to be collected. Furthermore, it is possible to exclude points where the estimation of the amount of blood to be collected is unstable from the list of puncturing candidate points and to preferentially select only points where the estimation is stable as puncturing candidate points.

[0043] FIG. 12 is a functional block diagram of a calculation unit 101b constituting the automatic blood collection device according to the present embodiment. As shown in FIG. 12, the calculation unit 101b according to the present embodiment differs from the second embodiment in that the reliability evaluation unit 16 includes a continuity evaluation unit 19 and a reliability calibration unit 20. Hereinafter, the same components as those in the second embodiment are denoted by the same reference numerals, and redundant explanations will be omitted. The continuity evaluation unit 19 and the reliability calibration unit 20 are realized by the CPU 102 shown in FIG. 1 and the memory 103 storing various programs. The continuity evaluation unit 19 evaluates the continuity of blood vessels. When a sudden discontinuity due to noise, epidermal damage, or the like is confirmed, the reliability calibration unit 20 interpolates the discontinuity based on the continuity of blood vessels to calibrate the reliability.

[0044] FIG. 13 is a schematic diagram illustrating the processing of the continuity evaluation unit 19 and the reliability calibration unit 20 shown in FIG. 12 . When calculating the puncture suitability, sudden discontinuities due to epidermal damage or the like may occur in blood vessels in the captured image. This distorts the image features around the relevant location, regardless of information such as the actual blood vessel density, thickness, and depth. Therefore, when calculating the puncture suitability, the continuity evaluation unit 19 checks the continuity of blood vessels near the target candidate point. If a discontinuity is found midway through a blood vessel determined to have continuity, the reliability calibration unit 20 calculates an interpolated path for the discontinuity based on the continuity of the blood vessels. The puncture suitability value is also estimated by interpolating the puncture suitability of the blood vessels before and after the interpolated path. The continuity determination method involves performing line tracing for a certain section from the location where the blood vessel discontinuity occurs, determining a direction vector in that section, and evaluating whether the direction vector, when inverted and extrapolated from the discontinuity, smoothly connects to another discontinuity. In this case, it is possible that the cause of the discontinuity is the three-dimensional movement of blood vessels under the skin, and that the discontinuity occurs simply because the vessel has burrowed deeper. In cases such as these, interpolation is not necessarily appropriate, and therefore interpolation of the puncture suitability should not be performed either. On the other hand, if the discontinuity is based on such natural depth changes, the brightness change caused by the discontinuity will not be sudden, but will tend to fade gradually and become invisible. Therefore, by focusing on the steepness of the brightness change in the discontinuous part, interpolation is performed only in cases of abrupt changes.

[0045] As described above, according to this embodiment, even if a blood vessel is discontinuous, the value of the puncture suitability that should be present can be compensated for from its continuity, and all puncture points that are expected to produce a sufficient amount of blood can be selected without omission.

[0046] The present invention is not limited to the above-described embodiments, but includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is 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, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment.

[0047] DESCRIPTION OF SYMBOLS 1...Automatic blood sampling device 11...Emphasis observation unit 12...Puncturing suitability calculation unit 13...Puncturing suitability determination unit 14...Puncturing position selection unit 15...Blood vessel evaluation unit 16...Reliability evaluation unit 17...Usage restriction unit 18...Puncturing suitability correction unit 19...Continuity evaluation unit 20...Reliability calibration unit 101, 101a, 101b...Calculation unit 102...CPU 103...Memory 104...Auxiliary storage 105...Interface (I / F) 106...Image input unit 107...Mechanism control unit 108...Light source control unit 109...Communication unit 110...Long wavelength range light source 111...Short wavelength range light source 112...Color camera 113...Diffuser (diffusion plate) 114...Polarizing filter 115...Body part 116...Drive unit 117...Puncturing needle 118...Blood sampling tube 300: Needle movable range 301: Blood vessel pattern 400: Visible light photographed image 402: Long wavelength component image 404: Short wavelength component image 406: Difference image 410: Difference image (α=0) 412: Difference image (α=m) 414: Difference image (α=1) 416: Brightness cross-sectional profile of difference image 410 418: Brightness cross-sectional profile of difference image 412 420: Brightness cross-sectional profile of difference image 414

Claims

1. An automatic blood collection device that performs blood collection by puncturing a living body part including a finger or an arm, the device comprising: a light receiving unit that receives light in a wavelength range capable of penetrating into the living body up to the maximum depth of a blood vessel included in the puncture target range and light in a wavelength range that can penetrate only near the surface layer; an image acquisition unit that acquires measurement images of the backscattered light in each wavelength range; an enhanced observation unit that selectively emphasizes and observes a blood vessel existing between near the surface layer and the maximum depth based on the difference between the measurement images in each wavelength range; a puncture suitability calculation unit that calculates, as a puncture suitability index value, an index value indicating whether or not it is suitable for puncture from the characteristics of the obtained blood vessel density and / or thickness; and a puncture suitability determination unit that determines whether or not puncture is suitable based on the puncture suitability.

2. The automatic blood collection device according to claim 1, wherein the light receiving unit receives light in a wavelength range capable of penetrating into the living body up to the maximum depth of a blood vessel included in the puncture target range and light in a wavelength range that can penetrate only near the surface layer, simultaneously or alternately.

3. The automatic blood collection device according to claim 2, wherein the puncture suitability determination unit obtains a correlation between the puncture suitability and the blood collection amount by pre-analyzing data, and performs blood collection only when the puncture suitability is such that a blood collection amount equal to or greater than a determined standard can be expected.

4. The automatic blood collection device according to claim 3, further comprising: a puncture position selection unit that selects, within the target living body part, the position most suitable for puncture according to the level of the puncture suitability; and a drive unit that moves the puncture to the position selected by the puncture position selection unit to perform puncture.

5. The automatic blood collection device according to claim 2, wherein the puncture suitability is calculated by a blood vessel evaluation unit designed such that, for each point on the image, the higher or lower the value, the more likely the point is on a blood vessel, the more prominent the thickness and density of the blood vessel, and the shallower the blood vessel.

6. The automatic blood collection device according to claim 2, further comprising: a reliability evaluation unit that quantitatively determines, as reliability, the image quality of the blood vessel at the puncture candidate position in the captured blood vessel image based on the change curvature of the luminance; a use restriction unit that restricts the use of the puncture suitability as low reliability when the reliability is low; and a puncture suitability correction unit that corrects the puncture suitability according to the reliability.

7. The automatic blood collection device according to claim 6, wherein the reliability evaluation unit includes a continuity evaluation unit that evaluates the continuity of blood vessels, and a reliability calibration unit that, when a sudden interruption due to noise or a wound on the epidermis is confirmed, interpolates the interrupted portion based on the continuity of the blood vessels to calibrate the reliability. The automatic blood collection device is characterized by this.

8. An automatic blood collection method for collecting blood by puncturing a living body part including a finger or an arm, comprising: a light receiving step in which a light receiving unit receives light in a wavelength range capable of penetrating into the living body to the maximum depth of a blood vessel included in a puncture target range and light in a wavelength range capable of penetrating only near the surface layer; an image acquisition step in which an image acquisition unit acquires measurement images of backscattered light in each wavelength range; an enhanced observation step in which an enhanced observation unit selectively enhances and observes a blood vessel existing between the near-surface layer and the maximum depth based on the difference between the measurement images in each wavelength range; a fitness calculation step in which a puncture fitness calculation unit calculates an index value indicating whether or not it is suitable for puncture, as the puncture fitness, from the characteristics of the obtained blood vessel density and / or thickness; and a puncture suitability determination step in which a puncture suitability determination unit determines puncture suitability based on the puncture fitness. The automatic blood collection method is characterized by this.

9. The automatic blood collection method according to claim 8, wherein in the light receiving step, the light receiving unit receives light in a wavelength range capable of penetrating into the living body to the maximum depth of a blood vessel included in a puncture target range and light in a wavelength range capable of penetrating only near the surface layer, simultaneously or alternately. The automatic blood collection method is characterized by this.

10. The automatic blood collection method according to claim 9, wherein the puncture suitability determination unit obtains the correlation between the puncture fitness and the blood collection amount by pre-analyzing data, and performs blood collection only when the puncture fitness is such that a blood collection amount equal to or greater than a determined criterion can be expected. The automatic blood collection method is characterized by this.

11. The automatic blood collection method according to claim 9, comprising: a step in which a reliability evaluation unit quantitatively obtains, as the reliability, the curvature of change in luminance of the image quality of a blood vessel at a puncture candidate position in a photographed blood vessel image; a step in which a utilization restriction unit restricts utilization as low reliability of the puncture fitness when the reliability is low; and a step in which a puncture fitness correction unit corrects the puncture fitness according to the reliability. The automatic blood collection method is characterized by this.

12. The automatic blood sampling method according to claim 9, comprising: a continuity evaluation unit evaluating the continuity of a blood vessel; and a reliability calibration unit, when a sudden interruption due to noise or an epidermal wound is confirmed, interpolating an interrupted portion based on the continuity of the blood vessel to calibrate the reliability.

Images