Analysis system and program

Abstract

To appropriately manage an original image after arithmetic processing.SOLUTION: In an analysis system, an arithmetic part builds new created image pixel information by performing arithmetic operation including Hadamard product of at least one or more basic image pixel information and at least one or more operation image pixel information. Moreover, when the created image pixel information includes data that cannot be within a range of binary data of a predetermined prescribed bit number, (A) the arithmetic part performs scaling for including the data within the range of the prescribed bit number with respect to the entire created image pixel information and outputs modified created image pixel information after scaling to a storage part, or (B) the arithmetic part performs arithmetic operation to reduce the bit number of the image pixel information so that it is included within the range of the predetermined prescribed bit number and outputs to the storage part the modified created image pixel information within the range of the predetermined prescribed bit number.SELECTED DRAWING: Figure 7

Description

[Technical field]

[0001] The present invention relates to any analytical imaging device that takes photographs, videos, or short-term photographs, including magnetic resonance imaging (MRI), positron emission tomography (PET), computer tomography (CT), optical microscope imaging devices, confocal microscope imaging devices, electron microscopes, and mass spectrometry imaging devices. [Background technology]

[0002] In image analysis devices, it is desirable for pixel signal strength and / or signal-to-noise ratio to be high. However, signal strength and / or signal-to-noise ratio depend on the amount, concentration, and observation time of the observed object, and the low resolution and low brightness images of current image analysis devices have limited utility.

[0003] For example, proton density (PD) images, T1 weighted images, and T2 weighted images are commonly used in MRI, and the shorter the relaxation time, the brighter the T1 weighted image (positive contrast image) and the darker the T2 weighted image (negative contrast image). The presence of paramagnetic substances, which have lower image brightness than PD images, shortens the relaxation time due to the relaxation enhancement effect (proton relaxation enhancement; PRE). In brain functional MRI that utilizes this property, T2* weighted images (T2 star weighted images) are used, but there is a problem of insufficient image brightness. In order to obtain high brightness, the number of integrations is increased. In addition, in order to obtain high resolution images, the number of encodings and projections is increased, which causes a problem of long imaging times. Medical analysis images require clear images, and patients are restrained for long periods of time without moving, so there is a strong demand for short imaging times. In parallel with improvements in analytical images through photographic techniques, various improvements have been made through computational processing of analytical images. For example, smoothing correction involves calculations such as moving average filter calculations, low-pass filter calculations, two-dimensional convolution filter calculations, and least squares method calculations, which smooth out changes in the shading (intensity) of an image. Moreover, noise removal by calculation is commonly performed. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] US 4891593 [Patent Document 2] WO92 / 04640 [Patent Document 3] WO94 / 03824 [Non-patent literature]

[0005] [Non-Patent Document 1] W. Youngdee et al, Optimization of field-cycled PEDRI for in vivo imaging of free radicals, Phys, Med. Biol. 46 (2001) 2531-2544 [Non-Patent Document 2] D. Lurei et al. Field-cycled PEDRI imaging of free radicals with detection at 450 mT, Magn. Reson. Imaging 23,175-181(2005) [Non-Patent Document 3] George PH Styan, Hadamard Products and Multivariate Statistical Analysis, Linear Algebra and Its Application 6: 217-240 (1973) Summary of the Invention [Problem to be solved by the invention]

[0006] Among such image processing operations, there are a few that significantly increase the bit depth of the image.

[0007] Medical image monitors used in clinical settings for reading images generally have a display specification of 256 levels (8 bits), and high-quality (16 bits, etc.) original images after processing are converted to 8 bits before being displayed on the monitor. However, a user (analyst) performing an analysis may sometimes feel the need to perform an analysis using the original image. In such a case, the user displays the image on a monitor that can display a high-quality image after processing. However, some arithmetic processes create image pixel information in binary data with an unnecessarily high bit count. This creates a problem in that applications are unable to handle the higher-than-expected bit count and are unable to display the original image. Furthermore, this type of arithmetic process causes the problem that there are no monitors or applications that can display the original image at that quality. This type of arithmetic process that increases the bit count involves a traceability problem in that the original image cannot be accessed after the process, and so has not been used.

[0008] The present invention aims at proper management of original images after computational processing. This specification discloses various aspects, many of which aim to improve operability. [Means for solving the problem]

[0009] One aspect of the present invention has the following configuration in order to solve the above problems. The image construction unit includes a memory unit and a calculation unit, the memory unit stores at least basic image pixel information and operational image pixel information, the basic image pixel information and the operational image pixel information are of the same sample and include the same analysis target area, the basic image pixel information includes a predetermined number of pixels captured by a shooting method different from the shooting method used to capture the operational image pixel information, and does not include analysis information or, even if it does, does not affect the analysis, the operational image pixel information is image pixel information that includes analysis information and has a brightness and / or resolution lower than that of the basic image pixel information, and the calculation unit calculates a Hadamard product for at least one of the basic image pixel information and at least one of the operational image pixel information. and constructing new original image pixel information by performing an operation including: when the original image pixel information includes data that does not fit within a predetermined range of binary data of a predetermined number of bits, the operation unit performs one of the following processes: (A) scaling the entire original image pixel information to fit within a range of binary data of a predetermined number of bits, and outputting the original image pixel information corrected by the scaling to the memory unit; or (B) performing an operation to reduce the number of bits of the image pixel information so that it fits within a predetermined range of binary data of a predetermined number of bits, and outputting the original image pixel information corrected to fit within the range of binary data of a predetermined number of bits to the memory unit. Effect of the Invention

[0010] It is now possible to properly manage original images. [Brief description of the drawings]

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

[0012] The inventors present an operation involving the Hadamard product as an example of an operation process that increases the number of bits of binary data of an image.

[0013] Before describing the specific examples and numerous experiments included in the example embodiments, the inventors will explain the concept of the Hadamard product.

[0014] (Hadamard product) The Hadamard product as referred to in the present invention is as defined in mathematics. In mathematics, the Hadamard product is a matrix multiplication determined by multiplying each element of a matrix of the same size. In this specification, the operator symbol for the Hadamard product is defined as "o". The Hadamard product can be explained using a 3x3 matrix as shown in the following equation (1).

number

[0015] (Image and image pixel information) An image is made up of pixels, and when you want to emphasize the pixels contained in the image, you say "image pixel information." Also, when you want to emphasize the shape of an object captured in an image, you say "image." For example, basic image is used in expressions such as "basic image displayed on a display." Basic image pixel information is used when you want to emphasize that calculations are being performed on pixels, for example, in expressions such as "calculating basic image pixel information." The use of image pixel information for an image is similarly applied to an operational image and an operational image pixel information.

[0016] (Number of pixels) The Hadamard product is an operation that requires two matrices of the same size. For fundamental image pixel information and differential image pixel information that have the same number of pixels, performing a Hadamard product operation on the fundamental image pixel information means that the fundamental image pixel information and the differential image pixel information can be expressed as matrices of the same size.

[0017] (Images containing the same area of ​​the same sample) The same sample means that the subject of analysis is the same. For example, if the sample is a patient, it means that the sample is the same patient. Also, the same region means that the image includes the same region. For example, if a whole-body image is captured, the analysis region may be appropriately cut out according to the purpose of analysis. The basic image and the operational image may be the whole-body image itself.

[0018] The embodiment is an invention of an analysis system (analysis device 1). The region in the embodiment is synonymous with the region that a user of the analysis system (analysis device 1) (the user may be an analyst or a photographer; hereinafter, simply referred to as "user") is attempting to analyze. For example, if a user is attempting to discover metastatic cancer cells by analyzing the entire body of a patient, the region may be a 3D image of the entire body of the patient. Also, if the entire liver of a patient is the analysis target, the entire liver becomes the region. The user decides what the region is. The image may include any area to be analyzed and may be a 2D or 3D image of the patient's entire body, or, if the liver is the subject of analysis, a 2D or 3D image of the entire liver.

[0019] (Dynamic image pixel information) The operational image pixel information includes analytical information, for example, when cancer cells are the analysis target, the operational image pixel information must include analytical information that can distinguish between cancer cells and normal cells. Being distinguishable means being distinguishable by an image, and means that the pixels of the part where normal cells are photographed have almost no brightness, while the pixels of the part where cancer cells are photographed have brightness. As described later, the brightness of the pixels of the analysis object part included in the operational image pixel information may be very low sensitivity, i.e., extremely low brightness. Rather, the embodiment described later makes analysis easier when the pixels of the area where the analysis object is photographed have extremely low brightness and / or the number of pixels is small. In other words, the analysis device 1 of the embodiment makes it possible to analyze even a low-resolution operational image in which the area of ​​normal cells is dark and has no brightness, while the area where cancer cells are present has barely any brightness.

[0020] (Basic image pixel information) The basic image pixel information may or may not include analytical information, but if it does include analytical information, it will not affect the analysis of the operational image pixel information. The total number of pixels contained in the basic image pixel information is preferably large, and is greater than the total number of pixels in the operational image pixel information. When the created image pixel information is created, the outline of the analysis target (e.g., the liver) is determined by the basic image pixel information, so the greater the total number of pixels, the clearer the created image pixel information will be. The total number of pixels in the basic image pixel information and the differential image pixel information is the total number of pixels in the image to be analyzed, which is obtained by photographing the same area and then cropping the image to include the analysis area. The total number of pixels is important for operations including the Hadamard product, which will be described later. The Hadamard product is an operation between matrices that have the same size columns and the same size rows. Therefore, when performing an operation including a Hadamard product on the basic image pixel information and the differential image pixel information, the basic image pixel information and the differential image pixel information become pixel matrices with the same column size and row size (and the total number of pixels is necessarily the same).

[0021] (Generated image pixel information) The concept of generated images was first introduced by the inventors. Generated images enable new analyses that have never been performed before.

[0022] (Explanation of symbols) The symbols used in this specification are defined as follows:

[0023] (Matrix notation) When a "matrix" is written in mathematical symbols, it is written in bold and / or italic letters. However, patent specifications do not allow for bold and italic notation. Therefore, the mathematical notation of a "matrix" in this specification is written in non-bold alphabets with underlines, as shown in [Math 2] below. In addition, in [Mathematical formula 2], matrices are written in bold italic letters.

[0024]

number

[0025] An image has pixels (x, y) in two dimensions and voxels (x, y, z) in three dimensions. The inventor will use pixels as an example for explanation, but will not exclude a three-dimensional image capture device. The brightness CI (Contrast Intensity) (x, y) (t) of a pixel (x, y) of an image at a certain time t is shown by the formula [Equation 3], and is the sum of the signal intensity SI (x, y) (t) of each pixel and noise Noise (x, y) (t).

number

[0026] (Brightness of basic image pixel information and differential image pixel information) The pixel information of a binarized image contains only information of 0 or 1, and does not contain any intermediate gradation information. On the other hand, the basic image pixel information is information that includes luminance information expressed by [Equation 3]. In other words, the basic image pixel information includes multi-level luminance information that includes intermediate gradations. The differential image pixel information is similar and includes multi-level luminance information including intermediate gradations.

[0027] (Image pixel information) The image pixel information of an image consisting of m × n pixels will be a square matrix if m and n are the same number. The shape of the pixels themselves does not matter.

[0028] (Basic and operational images) The following description describes the base images and the working images as n×n square matrix images. The operational image and the basic image must be captured or acquired by different imaging methods of the same analysis target area of ​​the same sample. The operational image and the basic image may be cropped from a larger image to obtain the analysis target area. However, both the operational image pixel information and the basic image pixel information must include the same analysis target area.

[0029] The matrix representation of the operational image pixel information is given by adding the subscript "op" to the matrix: CI op It is written as follows. The matrix representation of the basic image pixel information is given by adding the subscript "core" to the matrix: CI core It is written as follows. The matrix representation of the created image pixel information is given by adding the word "created" as a subscript to the matrix: CI created It is written as follows. Furthermore, when it is desired to include, in the matrix notation of the created image pixel information, the origin information of the operational image pixel information used for creating the created image pixel information that has already been created, the notation is as follows. For example, if the working image pixel information used is a DNP image CI created Add DNP after CI created-DNP It is written as follows. The basis and operational images do not have to be raw images taken with the same modality. A DPN image is an image created by subtracting two images taken with different imaging techniques, as described below. The base and working images may originate from computed images or from raw captured images.

[0030] (Total number of pixels and pixel interpolation) The total number of pixels is expressed as 32 × 32, etc. The Hadamard product is not limited to square matrices, but can be applied to matrices of n × m. For simplicity, the matrices described in this specification are square matrices. The inventor will explain the notation when pixel interpolation is performed. For example, when pixel information of a total pixel number (matrix number) of 32×32 is interpolated to image pixel information of a total pixel number (matrix number) of 128×128, the notation is represented by using arrows, such as "32×32→128×128". In addition, if image pixel information with a total pixel count of 128 x 128 is reduced to 32 x 32 by pixel count conversion and then pixel-interpolated to 256 x 256, it will be written as "128 x 128 → 32 x 32 → 256 x 256", etc.

[0031] (Notation of Hadamard product in the present invention) The above basic image pixel information (pixel matrix CI core ) and the dynamic image pixel information (pixel matrix CI op ) and perform the Hadamard product (see Non-Patent Document 3) shown in Equation [4] to obtain new high-luminance generated image pixel information (pixel matrix CI created ) is obtained.

number

[0032] (Best base image) Since hydrogen is contained everywhere in the living body, proton density (PD) imaging is an imaging method that can image organs and tissues without omission and can obtain images with high resolution and brightness. Therefore, many of the experiments disclosed in this specification use proton density (PD) images as basic images. Note that proton density (PD) images were suitable basic image pixel information in the case of many of the experiments described in this specification, but they may be inappropriate depending on the analysis purpose or analysis target. The basic image pixel information plays a role of providing shape information of the analysis area by performing calculations including the Hadamard product on the differential image pixel information, which often lacks shape information due to its low total pixel number, resolution, and brightness. Therefore, an image in which only the area to be analyzed has high resolution and brightness and has a large total pixel number is suitable as basic image pixel information. Furthermore, it is also important that the basic image pixel information has a large total number of pixels, because a large total number of pixels in the basic image pixel information increases the resolution and provides clear shape information. Since the basic image pixel information is pixel information used to assist in the analysis of the analytical information contained in the operational image pixel information, it is required that the basic image pixel information does not contain the desired analytical information, or, even if it does contain the desired analytical information, it does not affect the analysis of the operational image pixel information. If the basic image pixel information contains too much analytical information, it will overwhelm the analytical information contained in the operational image pixel information, making it difficult to analyze the operational image pixel information.

[0033] [Experiment 1] The magnetic resonance imaging device is FC-DNP-MRI (Patent Documents 2 and 3, Non-Patent Documents 1 and 2). The hydrogen nuclear spin of water molecules that are hyperfinely bound to a paramagnetic substance undergoes dynamic nuclear polarization (DNP) by exciting the polarization of the paramagnetic substance under a static magnetic field with electron spin resonance (ESR). The inventor attempted imaging using the DNP-MRI method (Patent Document 1), which visualizes the hydrogen nuclear spin.

[0034] Two test tubes (outer diameter 13 mm, height 30 mm) containing water and 2 mM paramagnetic substance Cmp (3-carbamoyl-PROXYL) free radical solution were prepared. The phantom used in Experiment 1 was created by placing these two test tubes under a kitchen rubber glove (containing polyvinyl alcohol) that resembled a human arm. A sample tube containing 2 mM paramagnetic substance Cmp (3-carbamoyl-PROXYL) free radical solution was attached to the ESR excitation coil. The DNP-MRI method is an analysis method in which an analysis image (hereafter referred to as a "DNP image") is obtained by subtracting image pixel information with electron spin resonance (ESR) excited from image pixel information without ESR excitation. Experiment 1 was controlled by this method. The total pixel count of all images was standardized to 124 × 124 pixels.

[0035] (How to create a DNP image) Figure 1 is an explanatory diagram of experiment 1. Both images in Figure 1(b) and Figure 1(c) are the original images used to create the DNP operating image in Figure 1(d). The DNP operating image is created based on images taken with two different pulse sequences. In detail, the image in Figure 1(b) is an ESR non-excitation image (ESR=0W) obtained at td=200 ms, a so-called T1 weighted image. Figure 1(c) is an image taken at td=200 ms after the ESR excitation (ESR=6.5W for 1300 ms). The DNP operating image in Fig. 1(d) is a difference image obtained by subtracting the pixel information of the images in Fig. 1(b) to Fig. 1(c). The image in Fig. 1(d) is generally called a DNP image. Fig. 1(e) is a DNP created image. Figure 2 shows the pulse sequence used in the experiment. d is the time when the magnetic field is converted to a high magnetic field (300 mT) and is set to 0 seconds. TR is the echo time when imaging using echo.

[0036] (Image from experiment 1) The DNP image in Figure 1(d) is very blurry and the entire image is covered with coarse dot noise. Note that the background is white because the noise makes the pixel brightness negative when subtraction is performed.

[0037] The DNP image in Figure 1(d) was used as the operating image. The object to be analyzed was a test tube containing 2 mM Cmp solution, which appears as a white disk. The difference can be seen when comparing it with the test tube containing water on the right side of the image. In this experiment 1, the test tube containing 2 mM Cmp solution is brighter because a test tube containing a high concentration (2 mM) Cmp solution was used. Actual DNP images of living organisms are less bright and unclear. In addition, actual DNP images of living organisms are more difficult to analyze because paramagnetic substances are localized and organs and their tissues are hardly visible. The DNP operating image pixel information in Figure 1(d) is “128 x 128 CI op-DNP " is written.

[0038] (Base image of experiment 1) The image in Fig. 1(a) was used as the base image. The acquisition conditions for the image in Fig. 1(a) were TR = 500 ms, flip angle = 60 degrees, bandwidth = 20 kHz, phase encoding number = 65, frequency encoding number = 96, and slice thickness = 5 mm. The basic image, Fig. 1(a), clearly shows the glove, the test tube containing the high concentration (2 mM) Cmp solution, and the test tube containing water. This image was taken using a different method from Fig. 1(d), which was selected as the operating image. The basic image in FIG. 1(a) meets the requirements of a basic image, that is, high resolution or high brightness, and a shooting method different from that of the operational image. CI core " is written. The base image pixel information is an image that does not contain analytical information or that may contain analytical information but does not affect the analysis, and is an image with high brightness and / or resolution. The operational image pixel information is image pixel information that contains analytical information and has lower brightness and / or resolution than the base image pixel information. Since the base image is determined based on the analysis target and analysis information, the same image can be used as both a base image and an operational image. Suppose there are images of an analysis target (for example, paramagnetic substances contained in the kidneys) and a non-analysis target (for example, the large intestine). If the analysis target is the paramagnetic substances contained in the kidneys, the image contains a large amount of information about the target to be analyzed, so even if it becomes an operational image, it cannot become a base image. However, if the analysis target is the large intestine, the image can become a base image if the large intestine is photographed with high resolution and high brightness.

[0039] (Hadamard product of experiment 1) 128×128 CI op-DNP ○ 128×128 CI core The following calculation is performed. As a result of the calculation, the DNP generated image ( CI created-DNP The image with 128 × 128 pixels, called the DNP operating image ( CI op-DNP ), the operating image has a background covered with noise (dots), the outline of the glove is unclear, and the contrast between the Cmp solution-filled test tube and the water-filled test tube is weak. In response to this, DNP Creative Image ( CI created-DNP ) the background noise has been eliminated, the outline of the glove is clearly visible, and the brightness of the Cmp solution-filled test tube has been greatly improved.

[0040] As explained above, the Hadamard product is a matrix-matrix multiplication operation performed by multiplying matrices of the same size component by component. Operations involving the Hadamard product of differential image pixel information and foundation image pixel information can be likened to multiplication, where the luminance of the differential image pixel information is used as the base, and the luminance of the foundation image pixel information is used as the "amplification factor." If the brightness of a pixel in the analysis area is 3 in the working image pixel information and 100 in the foundation image pixel information, the brightness will be amplified 100 times to 300. On the other hand, overlay (the superposition of images) is the addition of pixels, and the total only equals 3 + 100 = 103. On the other hand, pixel-by-pixel noise generation in the spatial noise domain is irregular and is reduced by the Hadamard product. It will be understood that the Hadamard product of the present invention and the conventional overlay are completely different concepts.

[0041] The basic image pixel information is not included in the operational image pixel information, and even if it is included, it does not affect the analysis of the operational image pixel information. The reason for this restriction is that only the analysis information in the operational image pixel information is to be analyzed. In contrast, basic image pixel information is used to clarify the appearance of tissues, organs, and the like. The analytical information contained in the operational image pixel information is amplified using the shading of the appearance of the tissue or organ as an amplification factor, allowing the user to analyze the presence of the analytical information in light of the appearance of the tissue or organ. If the basic image pixel information has too much analytical information, a synergistic effect occurs with the analytical information contained in the operational image pixel information, causing the pixel brightness to saturate, making it impossible to see the appearance of tissues and organs contained in the analysis area. For example, a blood vessel may be visible in the basic image pixel information, but the blood vessel may be completely invisible in the generated image pixel information.

[0042] The present invention makes it possible to analyze the analytical information contained in the operational image pixel information while taking advantage of the appearance of tissues and organs, which is why it is preferable that the basic image pixel information does not contain analytical information. A user of the analysis system of the embodiment is aware of the analysis area at the time when he or she wishes to analyze the operational image pixel information, and by comparing the basic image pixel information with the created image pixel information, can know the distribution of the analysis target in relation to the external appearance of the tissue or organ.

[0043] (Results of Experiment 1) Figure 3 shows the results of Experiments 1 and 2, and is a table of numerical analysis of SI (signal intensity), SNR (signal-to-noise ratio), and Increase (rate of increase) for the 2 mM CPM-filled test tube, water-filled test tube, and glove regions. Numerical analysis was performed to verify whether the visual evaluation was correct. For each image, the MRI region of interest (ROI) was set in the sample tube region and spatial noise region. Then, the brightness of each image was calculated using ImageJ, an analysis software provided by the National Institutes of Health (NIH), and numerical analysis was performed according to the AAPM Report. The SI (signal intensity) in the region of the 2 mM CPM-encapsulated test tube, which contains a high concentration of paramagnetic substance, increased from 764 to 7,589,150, an increase of 9,930 times. The most notable result was the increase in SNR (signal-to-noise ratio), which increased by 39 times. The water-filled test tube and glove regions also showed roughly the same trend, and the increase rate was also large in regions with low concentrations of the paramagnetic substance being analyzed, supporting the visual results mentioned above.

[0044] [Experiment 2] FIG. 4 is an explanatory diagram for explaining Experiment 2, which examines the effect of the total number of pixels of the operational image pixel information on the analysis. FIG. 4(a) is an image with a total number of pixels of 128×128 used as the basic image pixel information. FIG. 4(b) is a T1 weighted image, with a total number of pixels of 16×16 used to create a DNP image. FIG. 4(c) is an image excited by electron spin resonance (ESR), with a total number of pixels of 16×16 used to create a DNP image. FIG. 4(d) is an image that is a difference obtained by performing a subtraction operation on the pixel information of the image in FIG. 4(b) to that in FIG. 4(c). The image in FIG. 4(d) is generally called a DNP image. FIG. 4(e) is an image obtained by pixel-interpolating the image pixel information of FIG. 4(d), with a total number of pixels of 16×16, to 128×128, the same as the total number of pixels of the basic image pixel information. FIG. 4(f) is a generated image obtained by calculating the Hadamard product for the differential image pixel information of FIG. 4(e) and the basic image pixel information of FIG. 4(a).

[0045] In fact, the DNP images used as a comparison control were taken using a test tube containing 2 mM CPM, which is expected to have a strong SI (signal intensity). When using a living organism as a subject, paramagnetic substances are not contained in high concentrations. Generally, when the SI (signal intensity) is extremely weak, the signal intensity per pixel is increased by increasing the area of ​​each pixel (resulting in a reduction in the total number of pixels). DNP images taken of an actual living organism have a small total number of pixels and are very low resolution. In experiment 2, basic image pixel information with a total pixel count of 124 × 124 and differential image pixel information with a total pixel count of 16 × 16 were captured and the Hadamard product was applied. Experiment 2 was conducted as an experiment closer to the situation of biometric imaging.

[0046] (Pixel Interpolation) Since the Hadamard product is an operation that requires two matrices of the same size, the pixels contained in the fundamental image pixel information and the differential image pixel information must be the same. Therefore, the 16x16 captured dynamic image pixel information was converted to the same total pixel number of 124x124 as the base image pixel information by an appropriate pixel interpolation method.

[0047] The pixel interpolation of the pixel matrix may be a one-dimensional or multidimensional linear interpolation method, and the selection of the pixel interpolation method is arbitrary. In addition, since the basic image pixel information and the operational image pixel information are required to be captured using different imaging methods, they may not be captured simultaneously. Therefore, the position of the analysis target area may be shifted due to the basic image pixel information and the operational image pixel information being captured separately. In such a case, alignment is required. Such alignment may be performed by utilizing affine transformation or the like to match the position and size of the analysis target area of ​​both images. Note that various methods other than affine transformation have been developed as a means for performing alignment, and it is preferable to adopt an alignment means appropriate to the image to be used. The present invention also encompasses filtering in k-space as a method for reducing noise in dynamic images. In the present invention, "operation including the Hadamard product" does not preclude the inclusion of appropriate operations other than the Hadamard product, such as filtering, affine transformation, and pixel interpolation.

[0048] 16x16 in Fig.4(e) → 128x128 CI op-DNP The image is 128 × 128 in Fig. 1(d). CI op-DNP The image in Fig. 4(e) was blurred overall. The image in Fig. 4(e) was affected by pixel interpolation, and the edges were particularly unclear. However, to my surprise, the image in Fig. 4(f) obtained by the Hadamard product was 16x16 → 128x128. CI created-DNP The generated image is Fig. 1(e) 128 × 128 CI created-DNP The created image was much clearer than the original image. The inventors performed a numerical analysis to confirm that the image was clear. The data of Experiment 2 in Figure 3 is the result of the numerical analysis. The SNR (signal-to-noise ratio) increase in the test tube containing 2 mM CPM was 60 times, which was a better result than that in experiment 1.

[0049] (Results from Hadamard product) From the above experiments, it was clear that the Hadamard product sharpens the region to be analyzed. Furthermore, it was surprisingly found that a clearer image can be obtained by using an image with a smaller total number of pixels and lower resolution as the operating image. The inventors performed experiments not described in Patent Document 4 and Patent Document 4 to confirm this result. As a result, it was found that there is an optimal value for the total number of pixels. As shown in Figure 3, the SNR (signal-to-noise ratio) increase (increase rate) differs between the glove region and the 2 mM CPM-encapsulated test tube region, and the increase (increase rate) can differ depending on the tissue. Furthermore, it was found that the resolution itself was improved.

[0050] Depending on the type of analysis, it is preferable that the analysis area is clearly defined, but it may also be desirable for the shape of the surrounding tissue to be clearly defined. Furthermore, the optimum total number of pixels may vary depending on the type of organ, etc. Since the optimal value for the total number of pixels varies depending on the purpose of analysis, the type of organ, etc., this means that the optimal value cannot be determined without creating creative image pixel information from operational image pixel information captured with various total pixel numbers.

[0051] [Experiment 3] Experiment 3 was an experiment to verify whether the same results as in Experiment 2 could be obtained even when the differential image pixel information was used with the total number of pixels reduced by calculation. The operational image pixel information was captured with a total pixel count of 128 x 128. The basic image pixel information was captured with a total pixel count of 256 x 256. The operational image pixel information with a total pixel count of 128 x 128 was integrated through calculations so that the total pixel count became 32 x 32. After that, the operational image pixel information with the total pixel count reduced to 32 x 32 was pixel-interpolated using a linear interpolation method. As a result, operational image pixel information with a total pixel count of 256 x 256, the same as the basic image pixel information, was obtained. The obtained operational image pixel information can be expressed as 128x128 → 32x32 → 256x256. CI op It is. In experiment 3, the following calculations were performed using these image pixel information. 256×256 CI core ○ 128×128 → 32×32 → 256×256 CI op =128×128→32×32→256×256C I created And, 256×256 CI core ○ 128×128→256 CI op =128×128→256×256 CI created was compared to. As a result, 128x128 → 32x32 → 256x256 CI created The resulting image was clearer. From the above results, it has been discovered that there is an optimum value for the total number of pixels of the differential image pixel information of the present invention.

[0052] (Inventor's discovery) As described above, it has been discovered that, regardless of whether the image is captured by photography or by artificial means, the image will be clearer if the operational image pixel information having a smaller total number of pixels is subjected to pixel interpolation to match the total number of pixels of the basic image pixel information and then the Hadamard product is calculated. As shown in the above experiments, the generated images have dramatically higher brightness and resolution, making it possible to perform analyses and diagnoses that were previously impossible.

[0053] Therefore, the inventor has invented the following method of using the creative image. The user artificially or photographs working image pixel information with different total pixel numbers, and creates respective creative image pixel information from each working image pixel information. Then, the user stores them in the storage unit 3. The display unit 6 displays the obtained multiple types of generated image pixel information in a form that allows comparison. The user can select generated image pixel information generated from the operational image pixel information with the optimum total number of pixels in light of the analysis purpose. The display unit 6 can also display the operational image pixel information that is the basis of the creative image pixel information. This embodiment is preferable for performing a more precise analysis.

[0054] (Method of reducing the number of image matrices) It was surprising to find that even when the number of pixels in the captured dynamic image pixel information was reduced by software, the signal to noise ratio (SNR) of the resulting image pixel information created from the dynamic image pixel information improved. Furthermore, the inventors tried simple methods such as averaging as a means of reducing the number of pixels, but this also improved the signal-to-noise ratio (SNR) of the generated image pixel information. If there is operational image pixel information captured with a high quality matrix number (number of pixels), the number of pixels can be reduced as desired by software. Even if a large number of operational image pixel information is captured to obtain an optimal matrix number (number of pixels), the user can obtain operational image pixel information with the optimal matrix number (number of pixels) by software. The results of the simulation experiment showed that the present invention is universally applicable to images from any imaging method (X-ray, ultrasound, PET, etc.) Moreover, it was found that the present invention is applicable even if the basic image and the operational image are images of different modalities, such as X-ray and MRI.

[0055] (Multiple operational image pixel information) [Experiments 4-6] FIG. 5 is an explanatory diagram of Experiments 4 and 5 using biological samples. [Image A'], [Image B'], and [Image C'] in Figure 5 are three images taken in advance for use in Experiments 4 to 6, and the shooting conditions were the same as in Experiment 1. [Image A'] is an image taken under ESR excitation for obtaining a DNP image. [Image B'] is an image taken under the same conditions as [Image A'] except that ESR excitation was not used, and is an image called a T1 weighted image. [Image C'] is a PD image. [Image D'] shows the regions of interest (ROIs) set on the 2 mM Cmp sealed sample tube (Cmp2Mm), the area near the ESR coil surface, the wing area, and the void noise area. The brightness of each ROI was measured, and a numerical analysis was performed in the same manner as in Experiment 1. Figure 6 shows tables of the numerical analysis of signal intensity (SI) and signal-to-noise ratio (SNR). (a) is a table of the results of Experiment 4, and (b) is a table of the results of Experiments 5 and 6. The left columns of the tables in Figure 6(a) and (b) show the data for the ROIs of [Image D'], which were set in the order of wing, Cmp2mM, and surface from the top.

[0056] [Image B'] is a non-excited ESR image and was used not only for creating DNP images but also as a T1 operating image. The image pixel information of [Image A'] and [Image B'] was subtracted to obtain the conventional DNP image [Image E'] shown in Figure 5(a) and used as a control and as a DNP operating image.

[0057] In experiment 4, the T1 weighted image ([Image B']) was used as the working image, and the PD base image ([Image C']) was used. CI op-T1 〇 CI core-PD The Hadamard product was calculated using the Hadamard product. The resulting T1 image is shown in Figure 5(e). CI created-T1 It is.

[0058] Experiment 5 was conducted as shown in Fig. 5(b). CI op-DNP 〇 CI core-PDIn the experiment in which the Hadamard product was performed, the DNP generated image pixel information obtained was as follows: CI created-DNP It is. In addition, in experiment 6, Figure 5(b) was used as the basic image for creation, ( CI op-DNP 〇 CI core-PD ) core 〇 CI op-T1 The Hadamard product is calculated and the DNP·T1 generated image pixel information is CI created-DNP·T1 The term "created base image" means that the created image created in the previous operation including the Hadamard product is used as the base image for the next operation.

[0059] Here, the inventor will explain the difference between basic image pixel information and created basic image pixel information. The term "generated basic image pixel information" is used to describe the process of repeatedly performing operations including the Hadamard product on a plurality of working image pixel information in succession for one basic image pixel information.

[0060] The creation base image pixel information may be image pixel information that includes analysis information since the working image pixel information is captured during creation. In contrast, the basic image pixel information is an image that does not contain analytical information, or that may contain analytical information but does not affect the analysis, and is an image with high brightness and / or high resolution. The creation basis image pixel information may or may not include analysis information. The creation basis image pixel information is fundamentally different from the basis image pixel information in that it may include analysis information.

[0061] There is an image composition section, The image construction unit includes a storage unit and a calculation unit, The storage unit stores at least basic image pixel information and differential image pixel information, the foundation image pixel information and the operational image pixel information are of the same sample and include the same region of interest; The basic image pixel information includes a predetermined number of pixels captured by a different imaging method from that of the operational image pixel information, and does not include analysis information or, even if it does include analysis information, does not affect the analysis; the operational image pixel information is image pixel information including analytical information and having a lower brightness and / or resolution than the basic image pixel information; The calculation unit performs a calculation including a Hadamard product on at least one of the basic image pixel information and at least one of the differential image pixel information to construct new creative image pixel information, An analysis system characterized by using the created image pixel information as created basic image pixel information, and constructing another new created image pixel information by performing operations including a Hadamard product on the created basic image pixel information and another operational image pixel information.

[0062] Also disclosed is a program for operating the analysis system.

[0063] (Effect of repeating operations including Hadamard products) As described above, the inventor has explained the manner in which a plurality of pieces of differential image pixel information are successively subjected to the Hadamard product operation for one piece of basic image pixel information. For example, when making a diagnosis, the target disease may be the same, but there may be multiple features that indicate the disease. In order to make an accurate diagnosis, it is necessary to confirm that multiple features are present at the same time. For the sake of explanation, the features are referred to as feature A (reflected in the image captured by imaging method A) and feature B (reflected in the image captured by imaging method B). In order to narrow down the disease, first, a calculation including the Hadamard product is performed on the basic image and the operational image captured by imaging method A to generate a generated image. Next, a calculation including the Hadamard product is performed on another operational image captured by imaging method B and the obtained generated image (generated basic image), and a generated image reflecting both feature A and feature B is created. In this way, a creative image obtained by calculating the Hadamard product of a plurality of operational images for one basic image is a combination of the features of the individual operational images, enabling new diagnoses. This aspect allows multiple features to be utilized to extract or narrow down a disease.

[0064] Working image pixel information is image pixel information that includes analytical information and has a lower brightness and / or resolution than the basis image pixel information. Under typical functional imaging conditions, repeated imaging may result in a clear image with high brightness and / or resolution. When the imaging time is shortened to avoid putting a burden on the patient, the image may have low brightness and / or resolution and may still meet the requirements for operational image pixel information.

[0065] Thus, it is important that the basic image and the operational image are captured using different imaging methods, but the resolution and brightness of the operational image can be artificially modified.

[0066] Figure 6(a) shows the results of the T1 generated image in experiment 4, which shows an increase of more than 2,000 times in SI (signal intensity) and more than 20 times in SNR (signal-to-noise ratio) compared to the conventional T1 weighted image used as a comparison example. From the results of this numerical analysis, it was found that the T1 generated image of the present invention is an image that is extremely easy to analyze in terms of SI (signal intensity) and SNR (signal-to-noise ratio) compared to the conventional technology.

[0067] The results of the numerical analysis of the DNP generated image (Experiment 5) and the DNP·T1 generated image (Experiment 6) in the table of Figure 6(b) show that the SI of the DNP generated image of the present invention is about 2000 times higher and the SNR (signal to noise ratio) is 20 times higher than that of the conventional method. Furthermore, the DNP·T1 generated image of the present invention has an SI (signal intensity) that is one million times higher and an SNR (signal to noise ratio) that is more than 100 times higher than that of the conventional method, resulting in an image with astonishingly high resolution and image brightness.

[0068] The generated image of the present invention can improve the accuracy of diagnosis by the doctor who uses it, if the purpose is to find patients who are likely to have a disease. Even if it is mistakenly determined that there is a disease, a definitive diagnosis can be made by conducting a separate detailed examination. Since clear images are obtained, the doctor who reads the image is less likely to overlook a disease that is a false negative. The generated image of the present invention has the potential to revolutionize diagnosis. Furthermore, the present invention makes it possible to analyze analytical images that have been difficult to analyze due to problems such as resolution.

[0069] The dramatic improvement in SNR (signal-to-noise ratio) opens new possibilities. Early cancer is difficult to detect through image analysis. This means that users can obtain generated image pixel information with improved SNR (signal-to-noise ratio) by using multiple operating image pixel information with different modalities and shooting conditions. The present invention can contribute to the discovery of difficult-to-analyte targets, such as early-stage cancer.

[0070] (Example of an analysis device) In the above description, no particular reference numerals were added to the generated image pixel information and the operational image pixel information. In the following description of the embodiment, the operation by calculation is explained in order to make it easier to understand. CI created 35 (creative image pixel information) and CI op This is done using codes and equations such as 31 (operating image pixel information). 7 is an explanatory diagram of a magnetic resonance imaging apparatus (analysis device 1) equipped with an imaging section 2. The analysis device 1 is equipped with a storage section 3, a calculation section 4, the imaging section 2, and a display section 6.

[0071] [Storage] The storage unit 3 does not have to be a single device. Even if image data of the same patient is distributed among the storage units 3, such as servers, in multiple medical institutions, it can be easily collected as long as it is formatted in DICOM, the international standard for medical data communication.

[0072] [Memory section (operating image pixel information)] In the embodiment, the data storage unit 3 stores the captured image. CI op 31 (operating image pixel information) and CI core The two images of 32 (basic image pixel information) are stored. These images are both of the same sample, contain the same analysis target region, and are images captured by a plurality of different imaging methods. moreover, CI op 31 (operating image pixel information) is a set of a plurality of types of image data that are all of the same sample, include the same analysis target area, and are captured by a plurality of different imaging methods. CI op 31 (differential image pixel information). These CI op 31 (differential image pixel information) are both stored in the memory unit 3.

[0073] [Memory section (basic image pixel information)] The data storage unit 3 includes: CI core At least one of 32 (basic image pixel information) is stored. CI core 32 (basic image pixel information) CI op 31 (operating image pixel information) is taken using a different imaging method and is the same image including the same area of ​​the same patient (sample). CI core 32 (basic image pixel information) and CI op The image pixel information 31 (operational image pixel information) need only include an analysis area that is of interest to the user, and does not have to be a whole body image.

[0074] [Shooting date and time of basic image pixel information and operational image pixel information] CI core 32 (basic image pixel information) and CI op It is desirable that 31 (operating image pixel information) be photographed at close intervals. CIcore 32 (basic image pixel information) and CI op If the image pixel information 31 (operating image pixel information) is taken at significantly different times, the body shape of the same patient may change. Users should be aware that some organs, such as the intestines and stomach, may constantly change position. Prior to the Hadamard product calculation, CI core 32 (basic image pixel information) and CI op If there is no significant change in position due to the difference in shooting time, the position can be adjusted by an operation such as affine transformation. Furthermore, the analysis can be performed even if other organs move as long as the organ to be analyzed does not move. Also, CI op 31 (operational image pixel information) may be images taken of the same patient at different dates and times. By comparing different dates and times, it is possible to compare changes in the size of tumors, etc. If there are images taken every year for health checkups, etc., CI op 31 (operational image pixel information), it is also possible to look back and identify the time when a tumor developed.

[0075] [Proper management of images] The memory unit 3 stores the 16-bit binary data. CI op 31 (operating image pixel information) or stored as 16-bit binary data CI core It stores 32 (basic image pixel information). In general, 3D medical images are generally stored as 16-bit binary data. Medical images stored as 16-bit binary data according to the DICOM standard can be transmitted within or outside the hospital via communication.

[0076] The calculation unit 4 calculates the data stored in the memory unit 3. CI op 31 (operating image pixel information) andCI core 32 (basic image pixel information) is read, the Hadamard product is calculated, CI created Create 35 (created image pixel information). The result of the calculation process is CI created 35 (created image pixel information) becomes 32-bit binary data. When Hadamard is superimposed regardless of the reduction in the number of pixels, the SI becomes larger and larger. The number of bits of the image increases unnecessarily. The calculation unit 4 monitors the number of bits of the image, and is designed to be able to determine the number of bits to be monitored in advance to a predetermined number of bits. For example, the user can set the predetermined number of bits to 16 bits, which is the number for a general medical image.

[0077] The calculation unit 4 calculates a predetermined number of bits, 16 bits, by Hadamard product calculation. CI created The program is operated to detect when the pixel information of the generated image 35 (generated image pixel information) is exceeded. CI created Scaling is performed on the entirety of 35 (created image pixel information) to fit it within the range of bit binary data with a specified number of bits. This is a new image created by arithmetic processing. CI created If even a part of the pixels included in 35 (created image pixel information) exceeds the predetermined bit number of 16 bits, the calculation unit 4 performs the following process. CI created Scaling processing is performed on the entirety of 35 (created image pixel information) to fit it within the range of bit binary data with a predetermined number of bits. Scaled to fit into 16 bits CI created 35 (created image pixel information) is sent from the calculation unit 4 to the storage unit 3.

[0078] (Relationship between bit reduction and SNR) As mentioned above, the inventor has proposed a Hadamard overlapping mode in which the creation image pixel information is the creation basic image pixel information, and the creation basic image pixel information and another operational image pixel information are further subjected to a calculation including the Hadamard product to construct another new creation image pixel information. When the Hadamard product is overlapped in this way, the SI (signal intensity) becomes larger and larger. Therefore, the inventors have invented an embodiment in which the number of bits is suppressed to a predetermined number of bits. In this invention, although the SI (signal strength) decreases compared to before the bit number suppression, the improved state of the SNR (signal-to-noise ratio) is maintained. Furthermore, even if the number of bits is reduced, quantitativeness can be ensured by maintaining pixel units.

[0079] The embodiment is CI core 32 (basic image pixel information), multiple CI op This is particularly effective when processing the Hadamard product of 31 (differential image pixel information). To prevent the number of bits from becoming unnecessarily large, the user can resize the original 16-bit binary data of the image as necessary. CI created 35 (creative image pixel information) is guaranteed to be accessed and analyzed. The arithmetic process for increasing the number of bits of binary data of image pixel information is not limited to the Hadamard product, and the present invention is in principle applicable to any analysis system that performs arithmetic process for increasing the number of bits of binary data of image pixel information.

[0080] (Variations in the number of bits in binary data) The calculation unit 4 of the embodiment calculates a predetermined number of bits, 16 bits, by Hadamard product calculation. CI created It was programmed to detect when 35 (creative image pixel information) was exceeded. It is a new image created through computational processing. CI created In some cases, the number of bits in 35 (created image pixel information) can be easily predicted. When it is expected that the number of bits of the binary data of the image pixel information will increase, the calculation unit 4 performs the following process. The calculation unit 4 performs a calculation to reduce the number of bits of the image pixel information so that the number of bits falls within the range of the binary data of the predetermined number of bits. Next, the calculation unit 4 calculates the new image pixel information that falls within the range of the binary data of the predetermined number of bits. CI created 35 (created image pixel information) is output to the memory unit 3.

[0081] Moreover, if the number of bits of the binary data of the image pixel information increases without any need for prediction, the calculation unit 4 does not need to make the prediction. As a result, regardless of the image pixel information, the calculation unit 4 performs a calculation to reduce the number of bits of the image pixel information so that the image pixel information falls within the range of the binary data of the predetermined number of bits. Then, the calculation unit 4 calculates the new image pixel information that falls within the range of the binary data of the predetermined number of bits. CI created 35 (created image pixel information) is output to the memory unit 3.

[0082] [Analyzer] The concept of multimodality is also explained. When X-ray CT and magnetic resonance imaging are selected as modalities, the X-ray CT and magnetic resonance imaging can be arranged in tandem in the analysis device 1 to facilitate imaging. The specimen (patient) is carried on a bed between the gantry of the X-ray CT and the gantry of the magnetic resonance imaging. The specimen (patient) is then immediately CI op 31 (operating image pixel information) and CI core 32 (basic image pixel information) is captured. In this way, CI op 31 (operating image pixel information) was taken using a different shooting method CI core The imaging method of 32 (basic image pixel information) may be of a different modality. Of course, the X-ray CT and the magnetic resonance imaging device may be placed in separate locations and connected via a network to form the imaging unit 2 of a single analysis device 1.

[0083] For the sake of explanation, the calculation unit 4 is described as a single device that also includes shooting control, but it goes without saying that each function may be separated. The analysis device 1 of the embodiment includes an imaging unit 2, but the imaging unit 2 may be present or absent. An analysis device 1 that does not include an imaging unit 2 is connected to a network and is configured to be able to obtain image pixel information of the same patient. As communication standards for medical data such as DICOM are widespread, image pixel information can be transmitted from the network to CI op 31 (operating image pixel information) and CI core Obtaining 32 (basic image pixel information) is easy.

[0084] (SNR display) It is not generally known that an image created by, for example, overlaying two original images has a better SNR (signal-to-noise ratio) than the original images. As described above, the Hadamard product brings about a remarkable improvement in SNR (signal-to-noise ratio) when used as in the embodiment.

[0085] FIG. 8 is an explanatory diagram of a screen display displayed on the display unit 6. The display unit 6 can display a screen having a first image display area 61 (basic image display area), a second image display area 62 (operational image display area), and a third image display area 63 (creative image display area). Also, on the screen, a first image SNR display area 611, a second image SNR display area 621, and a third image SNR display area 631 are provided below each of the display areas (61 to 63). This allows the user to grasp the degree of improvement in SNR (signal-to-noise ratio). The user can also judge whether the adopted basic image pixel information is good or bad. The user can obtain and compare multiple creative image pixel information while replacing various basic image pixel information with the same operational image pixel information, and thus select the optimal basic image pixel information. Similarly, the SNR display areas (611·621·631) also serve to select the operational image pixel information. As described above, one piece of creative image pixel information can be obtained from a plurality of pieces of operational image pixel information and one piece of basic image pixel information. Furthermore, when performing analysis using the concept of the above-mentioned creation basic image pixel information, displaying the SNR (signal to noise ratio) improves convenience. The user can view the third image display area 63 (creation image display area) and the third image SNR display area 631 during the process of successively calculating the operational image pixel information using the Hadamard product. The user can judge the quality of the adopted operational image pixel information from the SNR (signal to noise ratio).

[0086] (Specify the analysis target) When calculating the SNR (signal-to-noise ratio), the user can specify the analysis area and background area. Specifying the area clearly distinguishes between noise (background area) and signal (analysis area), and allows the SNR of only the analysis area to be calculated. The background region may be automatically selected by the calculation unit 4. If the user has not specified an analysis region, the user can set the entire photographed object as the analysis region, in which case the SNR (signal-to-noise ratio) is calculated with the background and the entire photographed object as the analysis targets. The user can specify the analysis region by, for example, encircling the region using an interface such as a mouse. When the entire photographed object is to be the analysis target, the calculation unit 4 can automatically specify the analysis target. In this manner, the calculation unit 4 accepts the designation of the analysis region either by the user or automatically.

[0087] In view of the above, this specification also discloses an invention that displays the following SNR:

[0088] (Invention of an analysis system that displays SNR) An image forming unit and a display unit are provided, The image construction unit has a storage unit and a calculation unit, The storage unit stores a captured image as image pixel information, The calculation unit is A calculation process is performed on the image pixel information sent from the storage unit to create new image pixel information, receiving a designation of an analysis area included in the new image pixel information, and calculating a signal-to-noise ratio between the designated analysis area and a background area; The display unit has a display area for displaying the new image pixel information and a display area for displaying the signal-to-noise ratio of the image displayed thereon.

[0089] The "arithmetic processing" in this invention includes any arithmetic processing other than the Hadamard product, and therefore the "new image pixel information" in this invention is not limited to the created image pixel information obtained as a result of the Hadamard arithmetic processing.

[0090] (Invention of a program that operates an analysis system that displays SNR) A program for operating an analytical system, comprising: The image construction unit has a storage unit and a calculation unit, The storage unit stores a captured image as image pixel information, The calculation unit, Operate the image pixel information sent from the memory unit to perform arithmetic processing to generate new image pixel information; The calculation unit, receiving a designation of an analysis area included in the new image pixel information, and calculating a signal-to-noise ratio of the designated analysis area to a background area; A program for operating an analysis system, characterized in that the display unit has a new image pixel information display area, a signal-to-noise display area for displaying the signal-to-noise ratio of the image displayed therein, and causes the new image pixel information and the signal-to-noise ratio to be displayed, respectively.

[0091] (Invention of an analysis system that displays the SNR of generated image pixel information) There is an image composition section, The image construction unit includes a storage unit and a calculation unit, The storage unit stores at least basic image pixel information and differential image pixel information, the foundation image pixel information and the operational image pixel information are of the same sample and include the same region of interest; The basic image pixel information includes a predetermined number of pixels captured by a different imaging method from that of the operational image pixel information, and does not include analysis information or, even if it does include analysis information, does not affect the analysis; the operational image pixel information is image pixel information including analytical information and having a lower brightness and / or resolution than the basic image pixel information; The calculation unit performs a calculation including a Hadamard product on at least one of the basic image pixel information and at least one of the differential image pixel information to construct new creative image pixel information, A signal-to-noise ratio is calculated based on a designation of an analysis area included in the original image pixel information, and a signal-to-noise ratio is calculated based on the designation of an analysis area and a background area. The display unit has a generated image pixel information display area and a signal-to-noise display area for displaying the signal-to-noise ratio of the image displayed therein.

[0092] (Invention of a program that operates an analysis system that displays the SNR of generated image pixel information) A program for operating an analytical system, comprising: The image construction unit has a storage unit and a calculation unit, The storage unit stores at least basic image pixel information and differential image pixel information, the foundation image pixel information and the operational image pixel information are of the same sample and include the same region of interest; The basic image pixel information includes a predetermined number of pixels captured by a different imaging method from that of the operational image pixel information, and does not include analysis information or, even if it does include analysis information, does not affect the analysis; the operational image pixel information is image pixel information including analytical information and having a lower brightness and / or resolution than the basic image pixel information; The calculation unit is operated to perform a calculation including a Hadamard product on at least one of the basic image pixel information and at least one of the operational image pixel information, thereby constructing new creative image pixel information; Furthermore, the calculation unit is operated to receive a designation of an analysis area included in the generated image pixel information, and to calculate a signal-to-noise ratio of the designated analysis area and a background area; A program for operating an analysis system, characterized in that the display unit has a display area for displaying original image pixel information and a display area for displaying a signal-to-noise ratio of the image displayed therein, and causes the original image pixel information and the signal-to-noise ratio to be displayed, respectively.

[0093] (Overlay image and Hadamard product image) Recently, PET (positron emission tomography) has been widely used to detect cancer cells. PET images contain information about the location of cancer cells but not about the shape of organs, so images taken with other modalities that contain information about the shape of organs are overlaid and used for analysis. Images including the shape of the organs are appropriately selected depending on the purpose of analysis, such as X-ray CT images or MRI images. In general, an analysis system capable of combining PET and X-ray CT images by overlay is called PET-CT, and an analysis system capable of combining PET and MRI images by overlay is called PET-MRI.

[0094] The generated image 7 (Hadamard product image) created by the Hadamard product improves the SNR (signal-to-noise ratio) and has the potential to contribute to analysis, but since this is an image first proposed by the inventor, there is almost no accumulated clinical data, etc. On the other hand, there is a sufficient amount of accumulated clinical data on analysis systems that provide overlay images8 such as PET-CT and PET-MRI, and these images are familiar to users.

[0095] Therefore, the inventor recommends a display mode in which the overlay image 8 and the original image 7 (Hadamard product image) are displayed simultaneously on the display unit 6. The user can perform accurate analysis by comparing the overlay image 8 with the generated image 7 (Hadamard product image). FIG. 9 is an explanatory diagram in which an overlay image 8 and an original image 7 (Hadamard product image) are simultaneously displayed on a display unit 6. As shown in FIG. As the amount of clinical data analyzed using the generated image 7 (Hadamard product image) increases, clinical cases in which the generated image 7 (Hadamard product image) is appropriate and clinical cases in which the overlay image 8 is appropriate may be discovered. In such cases, the display format shown in FIG. 9 allows for accurate analysis.

[0096] This specification discloses the following invention for displaying overlay images and origin image 7 (Hadamard product image). (Invention of an analysis system that displays overlay images and Hadamard images) There is an image composition section and a display section. The image construction unit includes a storage unit and a calculation unit, The storage unit stores at least basic image pixel information and differential image pixel information, the foundation image pixel information and the operational image pixel information are of the same sample and include the same region of interest; The basic image pixel information includes a predetermined number of pixels captured by a different imaging method from that of the operational image pixel information, and does not include analysis information or, even if it does include analysis information, does not affect the analysis; the operational image pixel information is image pixel information including analytical information and having a lower brightness and / or resolution than the basic image pixel information; The calculation unit performs a calculation including a Hadamard product on at least one of the basic image pixel information and at least one of the differential image pixel information to construct new creative image pixel information, The display unit is characterized in that it simultaneously displays a created image, which is a Hadamard product image obtained by imaging the created image pixel information, and an overlay image in which the basic image pixel information and the operational image pixel information are overlaid.

[0097] This specification also discloses a program invention for operating the analysis system invention for displaying the overlay image and the Hadamard image. (An invention of a program that operates an invention of an analysis system that displays overlay images and Hadamard images) A program for operating an analytical system, comprising: There is an image composition section and a display section. The image construction unit includes a storage unit and a calculation unit, The storage unit stores at least basic image pixel information and differential image pixel information, the foundation image pixel information and the operational image pixel information are of the same sample and include the same region of interest; The basic image pixel information includes a predetermined number of pixels captured by a different imaging method from that of the operational image pixel information, and does not include analysis information or, even if it does include analysis information, does not affect the analysis; the operational image pixel information is image pixel information including analytical information and having a lower brightness and / or resolution than the basic image pixel information; Operate the calculation unit to construct new creative image pixel information by performing a calculation including a Hadamard product on at least one of the basic image pixel information and at least one of the operational image pixel information; A program for operating an analysis system, characterized by operating the display unit to simultaneously display a creative image, which is a Hadamard product image obtained by imaging the creative image pixel information, and an overlay image obtained by overlaying the basic image pixel information and the operational image pixel information.

[0098] (Unit display) Although not limited to this embodiment, the basic image pixel information, the operational image pixel information, and / or the generated image pixel information displayed on the display unit 6 may be configured to display the unit when the pixel has a unit. For example, the unit of a pixel in X-ray-CT is called HU value (previously called CT value, etc.). Also, the unit of a pixel in PET-CT may be SUV (standardized uptake value), which is a quantitative value obtained by correcting the radioactivity concentration measured from the image by the administered dose and body weight. Also, MRI quantitative images may be described in units of quantitative values, for example, milliseconds for MRI relaxation time images, and diffusion coefficients for MRI diffusion images. Therefore, the display unit 6 displays the pixel unit of the basic image pixel information, the pixel unit of the dynamic image pixel information, and the pixel unit of the creative image pixel information, if any. The units may be chosen according to the purpose of the analysis. For example, in Experiment 1, imaging was performed using dynamic nuclear polarization (DNP-MRI) (Patent Document 1), which visualizes hydrogen nuclear spins subjected to dynamic nuclear polarization (DNP). The pixel unit in this embodiment may be a physical quantity representing dynamic nuclear polarization. In addition, as a substance that brings about dynamic nuclear polarization, Experiment 1 uses a paramagnetic substance Cmp (3-carbamoyl-PROXYL) free radical solution. Therefore, it is also possible to use the Cmp concentration as the unit of a pixel. Displaying units appropriate for the analytical purpose contributes to improving analytical efficiency. The indication of units can also be made to pixels of the generated image pixel information if possible.

[0099] (Optimization of total number of pixels for dynamic image pixel information) As described above in (The inventor's discovery), it has been discovered that, regardless of whether it is obtained by photography or artificial means, the image will be clearer if the operational image pixel information having a smaller total number of pixels is pixel-interpolated to match the total number of pixels of the basic image pixel information and then the Hadamard product is calculated. As mentioned above, it is also pointed out that the optimal total number of pixels may vary depending on factors such as the type of organ. The inventor also pointed out that because the optimal total number of pixels varies depending on the purpose of the analysis and the type of organ, the optimal value cannot be determined without creating creative image pixel information from operational image pixel information captured at various total pixel numbers.

[0100] Therefore, the inventor proposes an invention relating to an interface (display unit 6) that can contribute to creating creative image pixel information based on operational image pixel information with an optimal total number of pixels. FIG. 10 is an explanatory diagram illustrating the flow of data between the storage unit 3 and the calculation unit 4. As shown in FIG. The memory unit 3 is 16 x 16 CI op 31 (operating image pixel information), 32 x 32 CI op 31 (operating image pixel information), 64 x 64 CI op 31 (operating image pixel information) etc., multiple images of the same subject photographed with different total pixel counts, or images with a reduced total pixel count due to pixel count conversion CI op 31 (operating image pixel information).

[0101] The memory unit 3 includes: CI core At least one basic image pixel information 32 is stored in the memory unit 3. CI core 32 (basic image pixel information) is, as explained so far, CI op The image data may be the same as that of the sample 31 (operational image pixel information) and may be of the same analysis target area. CI core32 (basic image pixel information). The storage unit 3 may also store the above-mentioned creation basic image pixel information.

[0102] The calculation unit 4 calculates the total number of pixels stored. CI op 31 (operating image pixel information) to 256 x 256 CI core Pixel interpolation is performed to match the total number of pixels of 32 (basic image pixel information). CI op The resulting data 31 (differential image pixel information) is stored in the storage unit 3. The pixel interpolation performed here is a preparation for the Hadamard product calculation that is executed next. Figure 10 shows the result of pixel interpolation of the total number of pixels from 16x16 to 256x256. CI op FIG. 10 shows the process of obtaining the differential image pixel information 31. The resulting image pixel information is then converted from 16×16 to 256×256. CI op It is written as 31 (operating image pixel information).

[0103] Pixel interpolation completes preparation for Hadamard product calculation CI op 31 (operating image pixel information) is 256 x 256 by the calculation unit 4 CI core In this way, the calculation unit 4 performs Hadamard arithmetic on the basic image pixel information 32. CI op 31 (operating image pixel information) to multiple CI created A plurality of pieces of generated image pixel information 35 are created, and the results are stored in the memory unit 3. Furthermore, the calculation unit 4 displays the previously created multiple images on the display unit 6 in response to a user request. CI created 35 (created image pixel information) is displayed.

[0104] In the example of FIG. 10, the display unit 6 has a 16×16 CI op31 (operating image pixel information) was pixel-interpolated to 256 x 256, which is the same total number of pixels as the basic image pixel information. CI op 31 (operating image pixel information) and 256 x 256 CI core 16x16 → 256x256 obtained by Hadamard product of 32 (basic image pixel information) CI created 35 (original image pixel information) is displayed on the left. And the display unit 6 is 32x32 → 256x256 CI created 35 (created image pixel information) is shown on the right. The user compares both images and selects the best one. CI created 35 (created image pixel information). The selection may be performed automatically by the analysis device 1 judging the degree of improvement in SNR, ease of viewing, and the like. Displayed on display unit 6 CI created 35 (created image pixel information) can be 2 or more, and 3 CI created 35 (created image pixel information) may be displayed.

[0105] 11 is an explanatory diagram of a modified display format of the display unit 6. The display unit 6 is a 16×16→256×256 CI created 35 (created image pixel information) and 32x32 → 256x256 CI created Two images, 35 (created image pixel information), are displayed so that one can be selected. The user makes the selection by touching the screen. The selection may be made automatically by the analysis device 1, which determines the degree of improvement in SNR, ease of viewing, etc. The display unit 6 displays the two CI created 16x16 used to create 35 (created image pixel information) CI op 31 (operating image pixel information) and 32x32 CI op 31 (operating image pixel information) and 256 x 256 CI core32 (basic image pixel information) are displayed on the same screen. 16×16 CI op 31 (operating image pixel information) and 32x32 CI op 31 (operating image pixel information) is the raw data for analysis. The embodiment of FIG. 11 allows the user to take the raw data into account. CI created 35 (created image pixel information) can be selected.

[0106] (Summary of total pixel count optimization) The above embodiment utilizes the inventor's discovery that, regardless of whether it is by photography or artificial means, an image can be made clearer by pixel-interpolating operational image pixel information with a smaller total number of pixels, matching it with the total number of pixels of the basic image pixel information, and then calculating the Hadamard product. There is an optimum value for the total number of pixels of the operational image pixel information. The optimum value cannot be known until an original image is actually created. Therefore, by simultaneously displaying original image pixel information created from operational image pixel information with various total numbers of pixels on the display unit 6, the user can select the appropriate original image pixel information. This embodiment is a groundbreaking embodiment that brings about improvement in the quality of analysis and reduction in the analysis time for the user. The selection may be made automatically by the analysis device 1 judging the degree of improvement in SNR, ease of viewing, and the like.

[0107] This specification discloses an invention of an analysis system that displays multiple pieces of original image pixel information as follows. (Invention of an analysis system that displays multiple creative image pixel information) There is an image composition section and a display section. The image construction unit includes a storage unit and a calculation unit, The storage unit stores at least basic image pixel information and differential image pixel information, the foundation image pixel information and the operational image pixel information are of the same sample and include the same region of interest; The basic image pixel information includes a predetermined number of pixels captured by a different imaging method from that of the operational image pixel information, and does not include analysis information or, even if it does include analysis information, does not affect the analysis; the operational image pixel information is image pixel information including analytical information and having a lower brightness and / or resolution than the basic image pixel information; The calculation unit performs a calculation including a Hadamard product on at least one of the basic image pixel information and at least one of the differential image pixel information to construct new creative image pixel information, The storage unit stores a plurality of pieces of operational image pixel information having different total pixel numbers, The calculation unit generates a plurality of generated image pixel information derived from the operational image pixel information having different total pixel numbers, The analysis system is characterized in that the display unit simultaneously displays the plurality of pieces of generated image pixel information.

[0108] In addition to this embodiment, the original image pixel information displayed on the display unit may be selectable. The selection may be performed automatically by the analysis device 1, judging the degree of improvement in SNR, ease of viewing, etc.

[0109] This specification discloses an invention of a program that operates an analysis system that displays a plurality of pieces of original image pixel information as follows. A program for operating an analytical system, comprising: There is an image composition section and a display section. The image construction unit includes a storage unit and a calculation unit, The storage unit stores at least basic image pixel information and differential image pixel information, the foundation image pixel information and the operational image pixel information are of the same sample and include the same region of interest; the foundation image pixel information and the operational image pixel information are of the same sample and include the same region of interest; The basic image pixel information includes a predetermined number of pixels captured by a different imaging method from that of the operational image pixel information, and does not include analysis information or, even if it does include analysis information, does not affect the analysis; the operational image pixel information is image pixel information including analytical information and having a lower brightness and / or resolution than the basic image pixel information; The calculation unit performs a calculation including a Hadamard product on at least one of the basic image pixel information and at least one of the differential image pixel information to construct new creative image pixel information, The storage unit stores a plurality of pieces of operational image pixel information having different total pixel numbers, The calculation unit generates a plurality of generated image pixel information derived from the operational image pixel information having different total pixel numbers, A program for operating an analysis system, characterized in that the display unit simultaneously displays the plurality of pieces of generated image pixel information.

[0110] In addition to this embodiment, the original image pixel information displayed on the display unit may be a program that operates the analysis system to allow selection. The selection may be performed automatically by the analysis device 1, judging the degree of improvement in SNR, ease of viewing, etc.

[0111] (Optimization of basic image pixel information) The inventors have discovered that there is an optimum value for the width of the window of pixel values ​​contained in the basic image pixel information. The inventors will explain using a case where a CT image is used as a base image. CT images are displayed in HU units (air is -1000, water is 0). Users perform air and water calibration in HU units before taking images using air and water phantoms. The range of HU units is very large, from 4000 for metal to about -2000. CT images are generally displayed in 256 gray levels. If the HU value range of 4000 to -2000 is assigned to 256 gray levels, the resulting image will clearly show objects with extremely high HU values, such as bones, but will be virtually indistinguishable from organs, which contain more or less water and are concentrated around 0 HU values.

[0112] Therefore, all pixels with an HU value of -128 or less are displayed as -128, and all pixels with an HU value of 128 or more are displayed as 128, thereby adjusting the WW (Window width) to a total of 256 gradations. In addition, the WL (Window level) is adjusted by measuring the HU value of the target organ and setting the average HU value as the center of the WW (Window width). For example, if the average HU value of the liver is 60, setting the WL (Window level) to a HU value of 60 will make the 0 gradation, which is the center of the 256 gradations (-128 to 128), the average value of the liver, so that subtle differences in HU values ​​that appear in the liver are clearly reflected in the image.

[0113] In this way, adjustment of WW (window width) and WL (window level) is commonly performed in general CT image capture. The inventor realized that when using a CT image as basic image pixel information, it is necessary to adjust the WW (Window width) and WL (Window level) from the viewpoint of obtaining clear generated image pixel information, not from the viewpoint of making the image clear. FIG. 12 is an explanatory diagram of a modified example of the display unit 6 that improves the efficiency of the selection of basic image pixel information. The embodiment of Fig. 12 has an increased number of basic image pixel information in addition to the embodiment of Fig. 11. CI core 32 (basic image pixel information) and right B 256 x 256 CI core 32 (basic image pixel information) is basic image pixel information with different WW and WL. The display unit 6 displays the 256×256 pixel data of the left A on the same screen or on a different screen. CIcore 32 (basic image pixel information) CI created 35 (created image pixel information) and right B 256 x 256 CI core 32 (basic image pixel information) CI created 35 (created image pixel information) can be displayed. By adopting this display format of display 6, the user can select the optimal WW and WL. CI core 32 (basic image pixel information) can be selected. In addition, the optimal WW and WL can be selected. CI core The selection of 32 (basic image pixel information) can also be configured so that the analysis device 1 automatically selects it taking into consideration the SNR (signal-to-noise ratio) of the analysis region, the difference in luminance between regions, the clarity of the image, and the like.

[0114] (Utilization of AI) In addition, the user can prepare a large number of creative images including analysis areas that utilize the many clinical photographs that have been collected to date, and then act as a teacher to select the most suitable creative image that is suitable for analysis, thereby constructing a learning model. The selection of the optimal original image can be left to the learning model. The analysis device 1 (analysis system) creates basic image pixel information of various WW and WL, and also creates operational image pixel information of various total pixel numbers by photographing or pixel number conversion. The analysis device 1 (analysis system) creates a plurality of pieces of original image pixel information from a plurality of pieces of basic image pixel information and a plurality of pieces of operational image pixel information. The analysis device 1 (analysis system) inputs the obtained plurality of pieces of original image pixel information into a learning model, and outputs optimal original image pixel information from among the input original image pixel information.

[0115] Figure 12 shows two images with different pixel counts. CI op 31 (operating image pixel information) is displayed, and 256 x 256 of the left A CI core 32 (basic image pixel information) and right B 256 x 256CI core The Hadamard product was performed for each of the 32 basic image pixel information. As necessary CI op The WW (Window width) and WL (Window level) of 31 (operational image pixel information) may also be adjusted. In addition, the display unit 6 CI op These are displayed under 31 (operating image pixel information).

[0116] In the above example, the display unit 6 has several CI core 32 (basic image pixel information) and the user can select the appropriate CI created 35 (created image pixel information) was selected. The analysis device 1 creates different basic image pixel information with various WWs and WLs, and by performing calculations including the Hadamard product with appropriate operational image pixel information, it is possible to create various generated image pixel information derived from the various basic image pixel information and operational image pixel information. The analysis device 1 may be configured to display only optimal original image pixel information on the display unit 6, taking into consideration the SNR (signal-to-noise ratio) and image clarity of the various pieces of original image pixel information obtained. Furthermore, the conditions for selecting original image pixel information can be freely determined by the user, and the analysis device 1 may be configured to allow the user to input the selection conditions determined by the user.

[0117] Furthermore, as described above, it is a preferable embodiment to display on the display unit 6 the basic image pixel information, the differential image pixel information and the differential image pixel information derived therefrom in terms of the unit of pixel value in order to increase the efficiency of analysis.

[0118] (Taking pictures of the operation) In the above-mentioned embodiments, the operational image pixel information has been mainly described as having low brightness, poor resolution, and a small total number of pixels due to the limitations of the imaging technique. The creative image pixel information of the present invention visualizes the analytical information contained in such low-quality operational image pixel information in an analytic form. Due to constraints other than those of the imaging technique, the dynamic image pixel information may have low brightness, poor resolution, and a small total number of pixels. For example, if there is a photographing method that can photograph an analysis area included in an image with high resolution, a large total number of pixels, and good brightness, the image can be analyzed without applying the present invention. However, even if the photographing method is excellent, when trying to photograph a moving image, it is necessary to photograph each image at high speed, so the photographed image has low brightness, poor resolution, and a small total number of pixels. Furthermore, there are many factors that force the imaging time to be shortened. Because PET requires the administration of a radioactive drug to the patient, it is necessary to reduce the dosage of the radioactive drug. If the amount of radioactive drug is reduced, the image obtained will have lower brightness, poorer resolution, and fewer total pixels.

[0119] In addition, imaging may be required to be performed in a short time for reasons other than technical reasons. For example, even if an excellent imaging method is available, the patient must be immobilized in the gantry for a long time in order to obtain clear images. This places a great burden on the patient, not only physically but also psychologically. In order to avoid putting a strain on the patient, it is necessary to take images in a short period of time, and the images taken are blurry (low resolution, low brightness, or a small total number of pixels).

[0120] In this way, even if there is an excellent imaging method, due to factors other than the imaging method, the image obtained may not be satisfactory in terms of resolution, total number of pixels, or brightness. Even in such cases, the present invention can create an original image by using the obtained image as a dynamic image, making analysis easier.

[0121] (Application to video) The inventors describe a frequently desired application for video. The basic images used in the videos can be obtained by using a high-speed imaging method. X-ray CT images and ultrasound images can be taken faster than general MRIs. In addition, MRI has technology that enables high-speed imaging by using multiple array-type antennas. The imaging method for the basic images used in the videos does not need to contain analytical information, so it only needs to be clear and have a large total number of pixels (high resolution and high brightness), and can be selected from a wide range of options. On the other hand, as explained in the above section (Capturing an operational image), the operational image needs to include analytical information, so there are few options for the imaging method that can be adopted. The selected imaging method is not necessarily suitable for high-speed imaging. An operational image captured using an imaging method that is not suitable for high-speed imaging will have the imaging time forcibly shortened in order to obtain a moving image, resulting in an operational image with insufficient resolution, total number of pixels, and brightness, and it becomes necessary to create an artificial image.

[0122] The inventors will now explain the case of taking moving pictures of the heart. Since the heart beats periodically, the basic image of the heart is taken by taking one cycle of images in accordance with the heart beat cycle, which is used as the basic image. This shortens the time required to take the basic image and reduces the burden of taking the image on the patient. The operational images are captured in accordance with the pulsation cycle captured in the basic images. The multiple operational images required for the video are obtained by capturing multiple images over multiple pulsation cycles by shortening the capture time. The user prepares a base image and an operational image that match the pulsation cycle, and obtains one generated image by performing a calculation including the Hadamard product. By repeating this process, the user can obtain all generated images for the first pulsation cycle. The base image for the next pulsation cycle reuses the base image from the previous pulsation cycle. The generated image used for the next pulsation video is created by reusing the base image from the first pulsation cycle, and is created by performing a calculation including the Hadamard product on the reused base image and the operational image for the next pulsation cycle. As explained above, even if the working image does not have sufficient resolution, brightness, and total number of pixels, a clear original image can be created. By using the analysis device 1 of the present invention in this way, a moving image using a clear original image can be created very efficiently.

[0123] (program) Furthermore, although this specification discloses the analytical system as described above, it goes without saying that it also discloses a program for operating the analytical system. Even if only the analytical system is mentioned and there is no particular mention of the program, this specification discloses the program for operating the analytical system.

[0124] Although the inventor has described in detail the embodiment examples and experiments according to the present technology with reference to the drawings, the specific configuration is not limited to these embodiments. The present invention is included in the present invention even if the design is changed within the scope of the gist of the present invention. Furthermore, the above-described embodiments and experiments can be reused and combined as long as there are no particular contradictions or problems in their purpose, configuration, and the like. [Explanation of symbols]

[0125] 1. Analytical equipment (analytical system) 2. Photography Department 3 Data storage unit 31 CI OP (Dynamic image pixel information) 32 CI core (Basic image pixel information) 35 CI created (Generated image pixel information) 4 Operational Unit 6 Display unit 61 1st image display area (basic image display area) 611 First image SNR display area 62 Second image display area (operational image display area) 621 Second image SNR display area 63 Third image display area (creative image display area) 631 3rd image SNR display area 7. Created image (Hadamard product image) 8. Overlay Images

Claims

1. There is an image composition section, The image construction unit includes a storage unit and a calculation unit, The storage unit stores at least basic image pixel information and differential image pixel information, the foundation image pixel information and the operational image pixel information are of the same sample and include the same region of interest; The basic image pixel information includes a predetermined number of pixels captured by a different imaging method from that of the operational image pixel information, and does not include analysis information or, even if it does include analysis information, does not affect the analysis; the operational image pixel information is image pixel information including analytical information and having a lower brightness and / or resolution than the basic image pixel information; The calculation unit performs a calculation including a Hadamard product on at least one of the basic image pixel information and at least one of the differential image pixel information to construct new creative image pixel information, Furthermore, the calculation unit When the original image pixel information includes data that does not fit within a range of binary data having a predetermined number of bits, (A) performing scaling on the entire original image pixel information so as to fit the original image pixel information within a range of bit binary data having a predetermined number of bits, and outputting the original image pixel information corrected by performing the scaling to the storage unit; or (B) performing an operation to reduce the number of bits of the image pixel information so that the number of bits falls within a range of binary data having a predetermined number of bits, and outputting the modified original image pixel information that falls within the range of binary data having a predetermined number of bits to the storage unit; The present invention is characterized in that the method performs any one of the following processes: Analysis system.

2. The analysis system according to claim 1 , further comprising an imaging unit.

3. A program for operating the analysis system of claim 1.

Images