Blood oxygen image generation method and device, computer device, and storage medium

Abstract

The present application relates to the technical field of endoscopy, and discloses a blood oxygen image generation method and device, computer equipment and a storage medium. The method comprises: obtaining target reflected light data corresponding to a to-be-detected tissue site at a plurality of preset wavelengths, and a data mapping relationship, the data mapping relationship being used to represent the corresponding relationship between the reflected light data and blood oxygen parameters; determining a target blood oxygen parameter matched with the target reflected light data according to the data mapping relationship; and performing visualization processing based on the target blood oxygen parameter to generate a blood oxygen image corresponding to the to-be-detected tissue site. Through the technical solution of the present application, the reflected light data can be corresponded to the blood oxygen parameters through the data mapping relationship, so that the blood oxygen image is generated, and accurate blood oxygen level visualization is realized.

Description

Technical Field

[0001] This invention relates to the field of endoscopy technology, specifically to a method, apparatus, computer device, and storage medium for generating blood oxygenation images. Background Technology

[0002] In medical diagnosis, accurately understanding the blood oxygenation status of human tissues is crucial for early disease detection, disease assessment, and treatment planning. While current endoscopic techniques allow for deep visualization within the body, most only provide morphological information and lack quantitative analysis of blood oxygenation parameters, potentially leading to missed or misdiagnosed cases. Furthermore, although some blood oxygenation analysis technologies (such as multi-diameter single-fiber optic systems) can provide quantitative blood oxygenation analysis, it is often difficult to present the data visually. Doctors must rely on complex data interpretation to infer blood oxygenation status, which not only increases the difficulty and time cost of diagnosis but also risks affecting accuracy due to human error in interpretation. Summary of the Invention

[0003] In view of this, the present invention provides a method, apparatus, computer equipment and storage medium for generating blood oxygen images, in order to solve the problem of the lack of intuitive and accurate quantitative analysis and presentation of blood oxygen in current endoscopic technology.

[0004] In a first aspect, the present invention provides a method for generating a blood oxygenation image, comprising: acquiring target reflected light data corresponding to a tissue site to be detected at multiple preset wavelengths, and a data mapping relationship, wherein the data mapping relationship is used to characterize the correspondence between the reflected light data and blood oxygenation parameters; determining target blood oxygenation parameters that match the target reflected light data according to the data mapping relationship; and performing visualization processing based on the target blood oxygenation parameters to generate a blood oxygenation image corresponding to the tissue site to be detected.

[0005] The blood oxygenation image generation method provided in this invention analyzes the blood oxygenation status of the tissue under test from multiple angles by acquiring reflected light data at multiple preset wavelengths. Through data mapping, the target reflected light data is converted into blood oxygenation parameters. This mapping ensures a more accurate conversion between reflected light data and blood oxygenation parameters, reducing errors that may occur in direct measurement. This improves the matching accuracy of blood oxygenation parameters, enabling the final generated blood oxygenation image to more realistically and effectively reflect the tissue's blood oxygenation level. Visualization processing transforms blood oxygenation parameters into blood oxygenation images, making the originally abstract blood oxygenation data intuitive and easy to understand. Doctors and medical personnel can quickly identify blood oxygenation abnormalities through the images and intervene promptly, thereby improving the accuracy and efficiency of assessing blood oxygenation status.

[0006] In one optional implementation, the data mapping relationship is pre-constructed. The process of constructing the data mapping relationship includes: acquiring optical characteristic parameters and blood oxygen characteristic parameters of the sample tissue at different wavelengths; performing simulations based on the optical characteristic parameters and blood oxygen characteristic parameters for any wavelength to obtain tissue light transmission characteristic data of the sample tissue under different combinations of characteristic parameters; establishing a first mapping relationship between the combination of characteristic parameters and the tissue light transmission characteristic data; acquiring sample reflected light data of the sample tissue at different wavelengths and establishing a second mapping relationship between the sample reflected light data and the tissue light transmission characteristic data; and establishing a data mapping relationship between sample reflected light data and blood oxygen characteristic parameters at different wavelengths based on the first and second mapping relationships.

[0007] The blood oxygenation image generation method provided in this invention improves the accuracy and precision of blood oxygenation detection through a pre-constructed mapping relationship. This method comprehensively considers optical characteristic parameters at different wavelengths, blood oxygenation characteristic parameters, and tissue light transmission characteristic data, thus capturing the optical properties and blood oxygenation status of the tissue more comprehensively. By combining simulation and experimental data to establish a first mapping relationship and a second mapping relationship, the sample reflected light data can be accurately mapped to blood oxygenation characteristic parameters, reducing errors and uncertainties in direct measurement. Therefore, this method not only improves the conversion accuracy between blood oxygenation parameters and reflected light data but also ensures the consistency and reliability of data at different wavelengths.

[0008] In one optional implementation, determining the target blood oxygen parameter that matches the target reflected light data according to the data mapping relationship includes: comparing the target reflected light data with the reflected light data, determining the first reflected light data with the smallest difference from the target reflected light data based on the comparison result, and determining the blood oxygen parameter corresponding to the first reflected light data as the target blood oxygen parameter.

[0009] The blood oxygenation image generation method provided in this invention can effectively find the closest reflected light data by accurately comparing the target reflected light data with existing reflected light data, thereby determining the matching target blood oxygenation parameters. This comparison method based on minimizing the difference can significantly improve the accuracy of blood oxygenation parameter estimation and reduce the uncertainty caused by measurement errors or data deviations. At the same time, the data mapping relationship simplifies the blood oxygenation parameter inference process, making the entire blood oxygenation detection more efficient, accurate, and stable.

[0010] In one optional implementation, a correlation coefficient is obtained based on the linear correlation between the target reflected light data and the reflected light data; if the correlation coefficient is greater than a preset threshold, the second reflected light data corresponding to the correlation coefficient is determined, and the blood oxygen parameter corresponding to the second reflected light data is determined as the target blood oxygen parameter.

[0011] The blood oxygenation image generation method provided in this invention determines the correlation coefficient based on the linear correlation characteristics between target reflected light data and existing reflected light data, thereby effectively evaluating the degree of data matching. When the correlation coefficient exceeds a preset threshold, it indicates that the two sets of data have a high degree of consistency. This allows for the reliable selection of second reflected light data, and the accurate calculation of the target blood oxygenation parameters based on its corresponding blood oxygenation parameters. This simplifies the data matching process and improves the efficiency and stability of blood oxygenation detection.

[0012] In one optional implementation, acquiring target reflected light data corresponding to the tissue site to be detected at multiple preset wavelengths includes: acquiring initial reflected light data corresponding to the tissue site to be detected at multiple preset wavelengths; performing noise correction on the initial reflected light data to obtain corrected reflected light data; acquiring a preset standard reference value, the standard reference value being obtained based on a diffuse reflection standard white board; and performing normalization correction on the corrected reflected light data based on the standard reference value to obtain target reflected light data.

[0013] The blood oxygenation image generation method provided in this invention effectively removes interference factors from the data by acquiring initial reflected light data at multiple wavelengths and performing noise correction. Normalization correction based on standard reference values ​​ensures the consistency and comparability of the data at different wavelengths, thereby improving the accuracy of the detection results.

[0014] In one optional implementation, visualization processing is performed based on the target blood oxygen parameters to generate a blood oxygen image corresponding to the tissue site to be detected, including: obtaining a visualization mapping relationship, which is used to characterize the correspondence between blood oxygen parameters and visualization parameters; converting the target blood oxygen parameters into corresponding visualization parameters based on the visualization mapping relationship; and rendering the image according to the visualization parameters to generate a first image corresponding to the tissue site to be detected, wherein the blood oxygen image includes the first image.

[0015] The blood oxygenation image generation method provided in this invention establishes a mapping relationship between blood oxygenation parameters and visualization parameters, ensuring the accuracy and consistency of blood oxygenation data when converted into images. Based on this mapping relationship, the target blood oxygenation parameters are converted into visualization parameters, and a blood oxygenation image is generated through image rendering. This makes blood oxygenation information more intuitive and easier to understand, not only improving the visualization effect of blood oxygenation detection results but also enabling doctors or users to quickly and accurately obtain and interpret blood oxygenation information.

[0016] In one optional implementation, scan images corresponding to the tissue site to be detected at multiple preset wavelengths are acquired; if the target blood oxygen parameter corresponding to any target pixel in the scan image is greater than a preset blood oxygen parameter value, the color channel parameters corresponding to the target pixel are superimposed to obtain a first processing result; if the target blood oxygen parameter corresponding to the target pixel is less than the preset blood oxygen parameter value, the color channel parameters corresponding to the target pixel are adjusted based on a preset color mapping rule to obtain a second processing result; the first processing result and the second processing result are integrated to generate a second image corresponding to the tissue site to be detected, and the blood oxygen image includes the second image.

[0017] The blood oxygenation image generation method provided in this invention flexibly adjusts the image processing method based on a comparison between the target blood oxygenation parameter and a preset blood oxygenation parameter value, thereby achieving more accurate visualization of blood oxygenation information. When the target blood oxygenation parameter is greater than the preset value, image details are enhanced by overlaying color channel parameters; conversely, when the target blood oxygenation parameter is less than the preset value, color channels are adjusted based on color mapping rules to ensure the image presents more appropriate blood oxygenation information. Integrating these two processing results generates a more recognizable and accurate blood oxygenation image, making the detection results more intuitive and easier to interpret, thus improving the effectiveness and reliability of blood oxygenation detection.

[0018] Secondly, the present invention provides a blood oxygenation image generation device, comprising: an acquisition module for acquiring target reflected light data corresponding to a tissue site to be detected at multiple preset wavelengths, and a data mapping relationship, wherein the data mapping relationship is used to characterize the correspondence between the reflected light data and blood oxygenation parameters; a matching module for determining target blood oxygenation parameters that match the target reflected light data according to the data mapping relationship; and a generation module for performing visualization processing based on the target blood oxygenation parameters to generate a blood oxygenation image corresponding to the tissue site to be detected.

[0019] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the blood oxygen image generation method of the first aspect or any corresponding embodiment described above.

[0020] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the blood oxygen image generation method of the first aspect or any corresponding embodiment described above.

[0021] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the blood oxygen image generation method of the first aspect or any corresponding embodiment described above. Attached Figure Description

[0022] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a schematic flowchart of a blood oxygenation image generation method according to an embodiment of the present invention;

[0024] Figure 2 This is a schematic diagram of an endoscope system according to an embodiment of the present invention;

[0025] Figure 3 This is a schematic diagram of another endoscope system according to an embodiment of the present invention;

[0026] Figure 4 This is a schematic diagram of a blood oxygen image processing unit according to an embodiment of the present invention;

[0027] Figure 5 This is a schematic flowchart of another blood oxygenation image generation method according to an embodiment of the present invention;

[0028] Figure 6 This is a schematic diagram of the absorption coefficient curves of oxyhemoglobin and deoxyhemoglobin according to an embodiment of the present invention;

[0029] Figure 7 This is a schematic flowchart illustrating the determination of blood oxygen parameter values ​​according to an embodiment of the present invention;

[0030] Figure 8 This is a flowchart illustrating another method for generating blood oxygen images according to an embodiment of the present invention;

[0031] Figure 9 This is a structural block diagram of a blood oxygen image generation device according to an embodiment of the present invention;

[0032] Figure 10 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] Endoscopy is a high-resolution technique for observing tissues, but it suffers from a lack of specificity in visually observing early cancerous lesions, other diseased tissues, and normal tissues. Therefore, various methods have been applied to obtain blood oxygen saturation information of tissues and lesions during endoscopy. Common methods include fluorescence contrast imaging and hemoglobin absorption-based techniques. These methods acquire the diffuse reflectance spectrum of the target tissue and process the spectral information to estimate the tissue's biological information. Fluorescence contrast imaging technology is mature and has good accuracy; while the Tissue Oximeter T-Stat (Spectros Co., Portola Valley), based on hemoglobin absorption, has been widely tested and used, becoming a representative product for detecting tissue blood oxygen saturation. This device uses a fiber optic probe adapted to a traditional endoscope to acquire blood oxygen saturation information, but its functionality is limited. In addition, there are some methods based on hyperspectral imaging, which have broad applicability and high accuracy.

[0035] Current endoscopic systems and computational methods can generally meet the requirements for acquiring blood oxygenation information, but some problems still exist due to hardware limitations and insufficient computational methods. For example, the system has poor compatibility with white light mode, the pseudo-color video has a low frame rate, and factors such as gastrointestinal peristalsis and jitter can cause image artifacts, resulting in a mismatch between blood oxygenation values ​​and tissue images. In addition, fluorescence contrast imaging is affected by contrast agents, and the examination time is limited; while T-Stat can only provide blood oxygenation information for a single point within a limited range. Current special light modes utilize the absorption characteristics of hemoglobin to emphasize blood vessels to a certain extent, but they have failed to further explore and process potential biological information.

[0036] Furthermore, current blood oxygenation detection methods often fail to present blood oxygenation data in an intuitive graphical form. Doctors need to interpret complex data to infer the blood oxygenation status of tissues, which not only increases the difficulty and time cost of diagnosis, but may also affect the accuracy of diagnostic results due to human interpretation errors.

[0037] In view of this, the technical solution of this invention utilizes multi-wavelength reflected light data and the mapping relationship between reflected light data and blood oxygenation parameters to accurately determine target blood oxygenation parameters, thereby achieving a quantitative assessment of tissue blood oxygenation status. Simultaneously, it transforms blood oxygenation parameters into intuitive blood oxygenation images, helping doctors more clearly observe the blood oxygenation distribution in the tissue being tested, reducing diagnostic difficulty and time costs, minimizing human error, and promoting more accurate disease diagnosis and treatment planning.

[0038] According to an embodiment of the present invention, a method for generating blood oxygen images is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0039] This embodiment provides a method for generating blood oxygen images, which can be used in computer devices such as desktop computers and laptops. Figure 1 This is a flowchart of a blood oxygen image generation method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps:

[0040] Step S101: Obtain target reflected light data corresponding to the tissue sites to be detected under multiple preset wavelengths, as well as data mapping relationships. The data mapping relationships are used to characterize the correspondence between reflected light data and blood oxygen parameters.

[0041] Multiple preset wavelengths refer to specific wavelength ranges selected based on the different absorption characteristics of oxyhemoglobin and deoxyhemoglobin at different wavelengths. For example, specific wavelengths in the range of 400nm-700nm can be selected, such as key wavelengths like 420nm, 465nm, and 630nm. These wavelengths can effectively reflect the blood oxygen information of tissues.

[0042] The tissue site to be tested refers to the part of the human body where blood oxygenation needs to be measured. For example, it can be tissue in the digestive tract, such as the mucous membrane tissue of the stomach and intestines.

[0043] Target reflected light data refers to the data on the light reflected from the tissue site under illumination at multiple preset wavelengths. This data can include information such as light intensity and spectral distribution. Specifically, a specially designed endoscopic system can be used to acquire target reflected light data corresponding to the tissue site (also known as the observation site) at multiple preset wavelengths. For example... Figure 2 As shown, the specially designed endoscope system includes: an electronic endoscope 11 for taking pictures of the tissue sites within the specimen; a computer device 12 (also known as an image processing device) for analyzing and processing the signals captured by the electronic endoscope 11, and its supporting input components such as keyboard and mouse 13; a cold light source device 14 for providing illumination for observation of the electronic endoscope 11; and a display 15 for displaying the observed images.

[0044] The electronic endoscope 11 includes a flexible insertion part 16 inserted into a cavity, an operating part 17 disposed at the base of the insertion part, and a cable 18 connecting the operating part to an image processor device and a cold light source device. The insertion part 16 of the electronic endoscope 11 is composed of a head end 16a, a bending part 16b, and a flexible tube part 16c.

[0045] like Figure 3 As shown, an illumination window 21 is provided at the front end of the head end 16a to provide illumination light to the tissue to be detected. A cold light source device conducts the illumination light to the head end 16a through a light guide 22 (such as a large-diameter optical fiber and fiber bundle). After being homogenized by an illumination lens 23, the illumination light exits from the illumination window 21 at a specified beam angle. The reflected light carrying tissue information passes through the imaging window 27 and the imaging lens 24 located at the front end of the head end 16a and is imaged onto the photosensitive surface 25a of the image sensor 25. The image sensor 25 is a monochrome black and white sensor (CD image sensor or CMOS image sensor). Each pixel on the photosensitive surface performs photoelectric conversion on the received light signal, and the signal charge is converted into a voltage signal by an amplifier and output as an imaging signal from the image sensor 25.

[0046] Figure 3 The computer device 12 shown includes a controller 40 for controlling the amount of light emitted from the cold light source device 14 and the parameters (e.g., exposure time) of the image sensor 25 in the electronic endoscope 11; an image processing unit 41 for analyzing and processing the images output by the image sensor 25, including a blood oxygen image processing unit 41a, a white light image processing unit 41b, and a special light image processing unit 41c; a storage unit 42 for storing relevant parameter configurations that the controller 40 needs to access; and a display control circuit 43 for outputting a signal containing the target image to the display 15. Among these, for example... Figure 4 As shown, the blood oxygen image processing unit 41a includes an image correction unit 44, a blood oxygen value matching unit 45, a blood oxygen value matching result storage unit 46, and a blood oxygen image generation unit 47.

[0047] The cold light source device 14 includes a broadband white light source 30, a light source driving circuit 31, a power supply 32, an acousto-optic tunable filter (AOTF) 33, and an AOTF control circuit 34.

[0048] The broadband white light source 30 consists of common endoscopic white light sources such as xenon lamps and halogen lamps, as well as matching lens groups (neither the white light source nor the matching lens group is included in the main text). Figure 3 (Drawn in the middle). The white light source has a full-band continuous spectrum with a color rendering index greater than 90; the matching lens group is used to focus and collimate the light emitted by the white light source so that the beam is finally injected into the acousto-optic tunable filter 33 in the form of approximately parallel light.

[0049] The acousto-optic tunable filter 33 is an optical element that operates using the acousto-optic effect. It can perform tunable wavelength filtering on an incident broadband light source, achieving selective adjustment of the operating wavelength of the outgoing light. Its basic principle is that the interaction between light and sound waves causes diffraction and scattering phenomena during light propagation. Therefore, when an external control signal (such as an electrical signal) changes the frequency of the sound wave within the transparent medium, the wavelength and direction of the diffracted light in the transparent medium also change accordingly, thereby achieving modulation and filtering of the outgoing light signal. In short, the acousto-optic tunable filter 33 can selectively allow light of a specific wavelength to pass through or separate different light waves by changing the frequency of the input sound wave.

[0050] The broadband white light source 30 provides a continuous visible light spectrum across the entire wavelength range for the acousto-optic tunable filter 33. The acousto-optic tunable filter 33, based on an electrical control signal, allows only narrowband light with a specific peak wavelength to be emitted at a given moment. The emitted light is focused by a focusing lens (located between the light outlet of the acousto-optic tunable filter 33 and the light guide 22, not in…). Figure 3 (Drawn in the middle) The light converges at the end face of the light guide 22. The narrow band light is transmitted to the head end 16a through the light guide 22. After being homogenized by the illumination lens 23, it shines on the tissue surface from the illumination window 21 at a specified light distribution angle. The reflected light information from the tissue surface is finally imaged on the photosensitive surface 25a of the image sensor 25 through the imaging window 27 and the imaging lens 24.

[0051] In practical applications, the process of acquiring target reflected light data corresponding to the tissue site under multiple preset wavelengths using a specially designed endoscope system can be exemplified by: a broadband white light source 30 in the cold light source device 14 emitting light with a full-band continuous spectrum having a color rendering index greater than 90; this light is then focused and collimated by a matching lens group to form approximately parallel light that enters the acousto-optic tunable filter 33. The acousto-optic tunable filter 33 selects the wavelength of the incident light according to an electrical control signal, allowing only narrowband light with a specific peak wavelength to exit at a given moment. Its wavelength resolution can be set to 10 nm, and the scanning range is 400 nm-700 nm. By changing the input acoustic wave frequency, modulation and filtering of different wavelengths of light are achieved, thereby scanning the visible light band in the time dimension.

[0052] The emitted light is focused by the focusing lens onto the end face of the light guide 22, and then transmitted through the light guide 22 to the tip 16a of the electronic endoscope 11. After being homogenized by the illumination lens 23, the light is irradiated onto the tissue to be examined through the illumination window 21 at a specified light distribution angle. The tissue reflects the light, and the reflected light carrying tissue information passes through the imaging window 27 at the front end of the tip 16a and the imaging lens 24, and is finally imaged on the photosensitive surface 25a of the image sensor 25.

[0053] Image sensor 25 is a monochrome sensor whose pixels are sensitive to the entire visible light spectrum. It can perform photoelectric conversion on the received light signal, and the signal charge is converted into a voltage signal by an amplifier as the image signal output. During this process, the controller 40 in computer device 12 controls the amount of light emitted from cold light source device 14 and the parameters of image sensor 25 to ensure the accuracy of data acquisition.

[0054] After the acousto-optic tunable filter 33 completes a scanning cycle from 400nm to 700nm (a total of 30 wavelength channels), the image sensor 25 receives the image data corresponding to the 30 channels. This data constitutes a hyperspectral data cube containing biological information (e.g., when the photosensitive array is 1440×1080, its size is 1440×1080×30). This data is the target reflected light data corresponding to the tissue site to be detected at multiple preset wavelengths acquired by the computer device, and will subsequently be transmitted to the computer device 12 for further analysis and processing. For example, the hyperspectral data cube of a single pixel can be a one-dimensional gray-level response matrix with a length of 30, I(λ i ) = 1, 2, 3...30; I(λ) i ) represents the center wavelength of the illumination light as λ i At that time, the grayscale response value of the pixel.

[0055] Furthermore, if the wavelength resolution of the acousto-optic tunable filter 33 can be set to 10nm, the scanning range to 400nm-700nm, and the minimum single exposure time to 62µs, then the shortest time required to complete one scanning cycle (from 400nm to 700nm, a total of 30 wavelength channels) is 1.86ms. The image sensor 25 used is a monochrome sensor with pixels sensitive to the entire visible light spectrum, a readout frame rate of no less than 200fps, and a readout time of no more than 5ms per channel. The signal reception duration within one scanning cycle is 0.15s. Therefore, the imaging rate of the endoscope system is limited by the single-frame exposure time of the image sensor, and the shortest time required to complete one cycle is 0.15s.

[0056] Data mapping relationships are mathematical or logical correspondences between reflected light data and blood oxygenation parameters, established through extensive experiments, simulations, or theoretical derivations. For example, a lookup table can be used. Blood oxygenation parameters include key indicators such as blood oxygen saturation in tissues (e.g., StO2) and blood volume fraction (e.g., Cb), which directly reflect the oxygenation status and blood supply of tissues. Specifically, data mapping relationships can be established using experimental data or known biological models. In experiments, reflected light data at different wavelengths is measured under known blood oxygenation levels, and the relationship between reflected light data and blood oxygenation levels is analyzed to establish a mapping model.

[0057] Step S102: Determine the target blood oxygen parameters that match the target reflected light data according to the data mapping relationship.

[0058] Target blood oxygenation parameters refer to key indicators that reflect the oxygenation status of blood in the tissue being tested, such as blood oxygen saturation and blood volume fraction. Specifically, target reflected light data reflects the scattering and absorption characteristics of different wavelengths of light in the tissue being tested, while the data mapping relationship, through experimental or theoretical analysis, correlates the intensity of reflected light at different wavelengths with blood oxygenation parameters. By comparing the target reflected light data with the data mapping relationship, the corresponding blood oxygenation parameters can be calculated.

[0059] Step S103: Visualize the target blood oxygenation parameters to generate a blood oxygenation image corresponding to the tissue site to be detected.

[0060] A blood oxygenation image is an image that can visually present the distribution of blood oxygen in a tissue being examined. Specifically, it involves converting a defined target blood oxygenation parameter into an image format, for example, by mapping different target blood oxygenation parameter values ​​to different colors or image details through color mapping or other graphic methods, thereby generating an image that visually reflects the distribution of blood oxygen.

[0061] The blood oxygenation image generation method provided in this invention analyzes the blood oxygenation status of the tissue under test from multiple angles by acquiring reflected light data at multiple preset wavelengths. Through data mapping, the target reflected light data is converted into blood oxygenation parameters. This mapping ensures a more accurate conversion between reflected light data and blood oxygenation parameters, reducing errors that may occur in direct measurement. This improves the matching accuracy of blood oxygenation parameters, enabling the final generated blood oxygenation image to more realistically and effectively reflect the tissue's blood oxygenation level. Visualization processing transforms blood oxygenation parameters into blood oxygenation images, making the originally abstract blood oxygenation data intuitive and easy to understand. Doctors and medical personnel can quickly identify blood oxygenation abnormalities through the images and intervene promptly, thereby improving the accuracy and efficiency of assessing blood oxygenation status.

[0062] This embodiment provides a method for generating blood oxygen images, which can be used in computer devices such as desktop computers and laptops. Figure 5 This is a flowchart of a blood oxygen image generation method according to an embodiment of the present invention, such as... Figure 5 As shown, the process includes the following steps:

[0063] Step S201: Obtain target reflected light data corresponding to the tissue sites to be detected under multiple preset wavelengths, as well as data mapping relationships. The data mapping relationships are used to characterize the correspondence between reflected light data and blood oxygen parameters.

[0064] Specifically, step S201 includes:

[0065] Step S2011: Obtain initial reflected light data corresponding to the tissue sites to be detected at multiple preset wavelengths.

[0066] Initial reflected light data refers to the data of light initially reflected from the tissue to be detected and received by the image sensor under illumination of multiple preset wavelengths, without any correction processing. Specifically, the initial reflected light data can be obtained using the specially designed endoscope system in step S101 above. A cold light source device emits light of a specific wavelength to illuminate the tissue to be detected, and the light reflected from the tissue is transmitted to the image sensor through a light guide, illumination lens, etc. The image sensor (such as a monochrome sensor) converts the light signal into an electrical signal, and the light information corresponding to these electrical signals is the initial reflected light data.

[0067] Step S2012: Noise correction is performed on the initial reflected light data to obtain corrected reflected light data.

[0068] Corrected reflected light data is obtained by noise correction of the initial reflected light data, resulting in data that more accurately reflects the true reflected light characteristics of the tissue, as it removes noise interference from the original data. Specifically, noise correction can be performed on the initial reflected light data using dead pixel correction and dark current correction to obtain corrected reflected light data. For dead pixel correction, since individual pixels on the image sensor may exhibit abnormal light responses (i.e., dead pixels), after acquiring the initial reflected light data, the location information of these dead pixels is identified based on the pre-detection and calibration results of the image sensor. Subsequently, the data at the location of the dead pixels is corrected using statistical methods based on the data of surrounding normal pixels. Dark current correction takes into account that the pixels of the image sensor generate a certain current (dark current) in the absence of light. Before acquiring reflected light data, the dark current output value of each pixel on the image sensor in the absence of light is measured and recorded. After obtaining the initial reflected light data, the pre-recorded dark current value is subtracted from the reflected light data corresponding to each pixel, thereby eliminating the interference of dark current on the reflected light data.

[0069] Step S2013: Obtain the preset standard reference value, which is based on the diffuse reflection standard whiteboard.

[0070] The preset standard reference value is a benchmark data used to normalize and correct the reflected light data. It reflects the light signal characteristics of the illumination light after passing through a specific path under ideal conditions, without the influence of tissue-specific absorption and scattering, providing a unified reference standard for subsequent accurate analysis of the reflected light data of the tissue to be tested. A diffuse reflection standard white board is a standard reference material with special optical properties in the visible light band. In the visible light band, its spectral reflectance for any light is greater than 99%, approximately considered to be 1. This means that after light shines on the diffuse reflection standard white board, the reflected light is almost unaffected by complex factors such as absorption and scattering of light by the white board's own material. The characteristics of the reflected light mainly depend on the illumination light itself and the transmission characteristics of the optical system. Specifically, an endoscope system is used to photograph and collect data from the diffuse reflection standard white board. The specific operation is similar to the illumination and data acquisition process for the tissue to be tested. The light emitted by the cold light source device is transmitted through a series of optical elements and then shines on the diffuse reflection standard white board. The light reflected from the white board is imaged onto an image sensor through the camera window and lens of the electronic endoscope. The image sensor converts the light signal into an electrical signal and outputs it. After a complete scan cycle (e.g., scanning within a specific wavelength range, such as 400nm-700nm), the data received by the image sensor becomes the standard reference values. Due to the high reflectivity of the whiteboard, these data primarily reflect the combined effect of the illumination spectrum and the image sensor's spectral sensitivity response, and are used in subsequent calibration processes.

[0071] Step S2014: Normalize the corrected reflected light data based on the standard reference value to obtain the target reflected light data.

[0072] For each pixel's corrected reflected light data, divide it by the corresponding standard reference value to obtain the target reflected light data for each pixel. For example, for a certain pixel at a certain wavelength λ... i The corrected reflected light data is I c (λ i The corresponding standard reference value is I0(λ). i (This standard reference value is data obtained by photographing a diffuse reflection standard white board at the same wavelength), then the target reflected light data of this pixel after normalization correction at this wavelength. and R m (λ i The data is sent to the blood oxygen matching unit 45. This method eliminates the influence of factors such as light source fluctuations and individual differences in image sensors on the reflected light data, making data from different measurements and different devices comparable, thus laying the foundation for accurate determination of blood oxygen parameters.

[0073] The blood oxygenation image generation method provided in this invention effectively removes interference factors from the data by acquiring initial reflected light data at multiple wavelengths and performing noise correction. Normalization correction based on standard reference values ​​ensures the consistency and comparability of the data at different wavelengths, thereby improving the accuracy of the detection results.

[0074] Step S2015: Obtain the data mapping relationship. The data mapping relationship is used to characterize the correspondence between reflected light data and blood oxygenation parameters. For details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0075] In some optional implementations, the data mapping relationships are pre-built, and the process of building the data mapping relationships includes:

[0076] Step a1: Obtain the optical characteristic parameters and blood oxygen characteristic parameters of the sample tissue at different wavelengths.

[0077] Optical characteristic parameters include the absorption coefficient, scattering coefficient, tissue thickness, and refractive index of the sample tissue. These parameters describe the influence on light propagation within the sample tissue, determining the light transmission path, absorption, and scattering within the tissue. Specifically, these parameters can be obtained through extensive experimental measurements and analyses of different types of tissues. For example, tissue thickness can be approximated through anatomical studies or imaging methods, while the refractive index can be measured using specialized optical measuring instruments. Simultaneously, by utilizing measuring equipment with known optical properties, the absorption and scattering of light by the tissue at specific wavelengths can be measured, thereby calculating parameters such as the absorption coefficient and scattering coefficient. For instance, the absorption function of tissues in the digestive tract is mainly undertaken by hemoglobin, which is primarily divided into oxyhemoglobin (HbO2) and deoxyhemoglobin (Hb). The total tissue absorption coefficient is expressed as a linear weighted average of the absorption coefficients of oxyhemoglobin and deoxyhemoglobin.

[0078]

[0079] Where C b Represented as the fraction of blood volume in a tissue; μ aHbO2 and μ aHb S represents the absorption coefficients of oxyhemoglobin and deoxyhemoglobin, respectively; t O2 is the oxygen saturation of blood in tissues.

[0080] Figure 6These are the absorption coefficient curves for oxyhemoglobin and deoxyhemoglobin. Clearly, these two curves differ at most wavelengths, overlapping only in a few wavelength ranges. Under otherwise constant conditions, the total absorption coefficient is unique for different tissue physiological parameters (blood volume fraction Cb and blood oxygen saturation value StO2).

[0081] Blood oxygenation characteristic parameters refer to key indicators such as blood oxygen saturation (e.g., StO2) and blood volume fraction (e.g., Cb) in a sample tissue. These directly reflect the tissue's oxygenation status and blood supply. Specifically, invasive and non-invasive detection methods can be used to obtain these parameters. Invasive methods include extracting blood samples for blood gas analysis, directly measuring blood oxygen content to determine oxygen saturation and blood volume fraction. Non-invasive methods utilize techniques such as near-infrared spectroscopy, measuring the reflection and absorption of specific wavelengths of near-infrared light by tissues, combined with specific algorithms to estimate blood oxygenation characteristic parameters. Additionally, relevant data from existing clinical cases can be collected as a reference.

[0082] Step a2: For any wavelength, simulation is performed based on optical characteristic parameters and blood oxygen characteristic parameters to obtain tissue light transmission characteristic data of sample tissue under different combinations of characteristic parameters.

[0083] Different combinations of characteristic parameters refer to the arrangement and combination of optical characteristic parameters with different values ​​(such as different absorption coefficients, scattering coefficients, tissue thicknesses, and refractive indices) with different blood oxygenation characteristic parameters (such as different blood oxygen saturation and blood volume fractions). Tissue light transmission characteristic data refers to the various properties exhibited by light as it propagates through a sample tissue under different combinations of characteristic parameters, including information such as light transmittance, reflectance, and light intensity distribution. These data reflect the results of the interaction between light and tissue. Specifically, advanced simulation methods such as Monte Carlo simulation algorithm (MCML) can be used to simulate and obtain tissue light transmission characteristic data of the sample tissue under different combinations of characteristic parameters. The obtained different optical characteristic parameters and blood oxygenation characteristic parameters are used as input data, and the initial conditions of light (such as light source intensity and wavelength) and the geometric model of the tissue are set. Then, the MCML algorithm is used to simulate the random propagation of photons in tissue, considering the interactions between photons and tissue molecules, such as absorption and scattering. Data such as transmitted and reflected light after passing through a certain thickness of tissue, including light intensity, light distribution, and reflectivity, are calculated. These calculation results represent the tissue light transmission characteristic data corresponding to different combinations of characteristic parameters at that wavelength. By repeating the above process for multiple wavelengths, comprehensive tissue light transmission characteristic data can be obtained.

[0084] Step a3: Establish the first mapping relationship between the combination of characteristic parameters and the tissue optical transmission characteristic data.

[0085] The first mapping relationship refers to the correspondence established between different combinations of characteristic parameters (including various combinations of optical and blood oxygenation characteristic parameters) and tissue optical transmission characteristic data. Specifically, by simulating and calculating tissue optical transmission characteristic data for a large number of different combinations of characteristic parameters, each combination of characteristic parameters and its corresponding tissue optical transmission characteristic data are recorded and organized. For example, using tables or databases, different combinations of absorption coefficients, scattering coefficients, blood oxygen saturation, and blood volume fraction are used as indexes, and the corresponding tissue optical transmission characteristic data such as light transmittance and reflectance are stored as values, thereby establishing a clear correspondence between the two, i.e., the first mapping relationship.

[0086] Step a4: Obtain sample reflected light data of the sample tissue at different wavelengths, and establish a second mapping relationship between the sample reflected light data and the tissue light transmission characteristic data.

[0087] Sample reflected light data refers to the data obtained by illuminating a sample tissue and collecting its reflected light. This data records information such as the intensity and spectral distribution of the reflected light at different wavelengths. The second mapping relationship is the correspondence between sample reflected light data and tissue light transmission characteristic data. Specifically, using equipment and methods similar to those used to acquire reflected light data of the tissue to be tested, such as the light source and image sensor in an endoscope system, the sample tissue is illuminated and its reflected light is collected. Multiple measurements are performed at different wavelengths, and the reflected light data obtained from each measurement is recorded. Then, this sample reflected light data is compared and analyzed with tissue light transmission characteristic data obtained through simulation. Through statistical analysis methods, such as finding patterns in the sample reflected light data corresponding to the same or similar tissue light transmission characteristic data, a correlation is established between the two, i.e., the second mapping relationship.

[0088] Step a5: Based on the first and second mapping relationships, establish a data mapping relationship between sample reflected light data and blood oxygen characteristic parameters at different wavelengths.

[0089] Using the first mapping relationship, the corresponding combination of feature parameters can be derived from the tissue phototransmission feature data. Then, combining this with the second mapping relationship, the sample reflected light data is linked to the blood oxygen feature parameter in the feature parameter combination. For example, when the sample reflected light data at a certain wavelength is known, the corresponding tissue phototransmission feature data is found through the second mapping relationship, and then the corresponding feature parameter combination is found through the first mapping relationship, thereby extracting the blood oxygen feature parameter. Finally, a direct correspondence between the sample reflected light data and the blood oxygen feature parameter is established, i.e., a data mapping relationship, and this established data mapping relationship is stored in the lookup table storage unit 46.

[0090] Furthermore, due to the limited available resources of the storage unit, MCML simulation cannot establish the relationship between all reflected light data and blood oxygen parameters. Therefore, appropriate linear interpolation can be performed on blood oxygen parameters (such as blood oxygen saturation StO2 and blood volume fraction Cb).

[0091] In the above embodiments, the blood oxygenation image generation method provided by the present invention improves the accuracy and precision of blood oxygenation detection through a pre-constructed mapping relationship. This method comprehensively considers optical characteristic parameters at different wavelengths, blood oxygenation characteristic parameters, and tissue light transmission characteristic data, thus capturing the optical properties and blood oxygenation status of the tissue more comprehensively. By combining simulation and experimental data to establish a first mapping relationship and a second mapping relationship, the sample reflected light data can be accurately mapped to blood oxygenation characteristic parameters, reducing errors and uncertainties in direct measurement. Therefore, this method not only improves the conversion accuracy between blood oxygenation parameters and reflected light data but also ensures the consistency and reliability of data at different wavelengths.

[0092] Step S202: Determine the target blood oxygen parameters that match the target reflected light data according to the data mapping relationship.

[0093] For example, when the data mapping relationship is a lookup table, the lookup table storage unit 46 pre-stores a lookup table between reflected light data and blood oxygen parameters. For a single pixel, the blood oxygen value matching unit 45 will receive the R value from the image correction unit 44. m (λ i Add it to the lookup table and search for it in R. m (λ i The target blood oxygen parameter that best matches the spectral signal.

[0094] Specifically, step S202 includes:

[0095] Step S2021: Compare the target reflected light data with the reflected light data, and determine the first reflected light data with the smallest difference from the target reflected light data based on the comparison result; determine the blood oxygen parameter corresponding to the first reflected light data as the target blood oxygen parameter.

[0096] The first set of reflected light data is the set of reflected light data that has the smallest numerical difference from the target reflected light data at each wavelength within a pre-built data mapping relationship. Specifically, multiple sets of reflected light data are extracted from the stored data mapping relationship (such as a lookup table). These data are acquired and stored under different tissue conditions. For each set of reflected light data, the difference between it and the target reflected light data is calculated at each preset wavelength. By comparing the sum of these differences or some weighted sum, the set of reflected light data with the smallest difference is found, which is the first set of reflected light data.

[0097] For example, the target reflected light data R on a single pixel m (λ i The image correction unit 44 sends the data to the blood oxygen matching unit 45. This is done to find the optimal target reflected light data R across 30 wavelength channels. m (λ i Reflected light data R fitted with MCML S (λ i The objective function is introduced to find the blood oxygenation parameter (e.g., a combination of blood oxygen saturation StO2 and blood volume fraction Cb) that best approximates all elements in the dataset.

[0098]

[0099] This objective function represents the main function of the blood oxygen matching unit 45, namely, the target reflected light data R for each pixel. m (λ i Match a set of blood oxygen parameters with the smallest differences.

[0100] Each blood oxygen parameter can be mapped to a set of fitted reflected light data R in the lookup table storage unit 46. S (λ i ). Iterate through all blood oxygenation parameters and obtain R. S (λ i The objective function is input sequentially for calculation until a blood oxygen parameter value is found that minimizes the value of X. This is considered as finding the target reflected light data R. m (λ i The corresponding target blood oxygen parameters.

[0101] Step S2022: Based on the linear correlation characteristics between the target reflected light data and the reflected light data, the correlation coefficient is obtained.

[0102] The correlation coefficient is a statistical indicator used to measure the degree of linear correlation between two variables: target reflected light data and reflected light data. For example, it could be the Pearson correlation coefficient. Specifically, it is calculated using correlation analysis methods in statistics, such as the Pearson correlation coefficient formula. This formula calculates the ratio of the product of the covariance between two variables to the product of their respective standard deviations, thus quantifying the degree of linear relationship between them.

[0103] Step S2023: If the correlation coefficient is greater than the preset threshold, then determine the second reflected light data corresponding to the correlation coefficient, and determine the blood oxygen parameter corresponding to the second reflected light data as the target blood oxygen parameter.

[0104] The preset threshold is a pre-defined numerical threshold used to determine whether the correlation coefficient meets the conditions for determining the second reflected light data and the target blood oxygen parameter. The second reflected light data is the reflected light data corresponding to the first reflected light data with the smallest difference from the target reflected light data, provided the correlation coefficient is greater than the preset threshold. Specifically, when the calculated correlation coefficient is greater than the preset threshold, the second reflected light data corresponding to that correlation coefficient is found from the existing reflected light data. Because the second reflected light data and the target reflected light data have a high correlation, the blood oxygen parameter corresponding to the second reflected light data is determined as the target blood oxygen parameter.

[0105] For example, such as Figure 7 As shown, when the relevant parameter is the Pearson correlation coefficient, if the Pearson coefficient is greater than 90, the judgment stops, and the corresponding blood oxygen parameter value is taken as the final result. Furthermore, the above operation is performed pixel-by-pixel. After judging each pixel, its corresponding blood oxygen parameter result is saved and used as the initial value for calculating the next nearby pixel, which effectively improves the judgment speed.

[0106] Step S203: Visualize the target blood oxygenation parameters to generate a blood oxygenation image corresponding to the tissue site to be detected. For details, please refer to [link to relevant documentation]. Figure 1 Step S103 of the illustrated embodiment will not be described again here.

[0107] The blood oxygenation image generation method provided in this invention can effectively find the closest reflected light data by accurately comparing target reflected light data with existing reflected light data, thereby determining the matching target blood oxygenation parameter. This comparison method based on minimizing the difference significantly improves the accuracy of blood oxygenation parameter estimation and reduces uncertainties caused by measurement errors or data deviations. Simultaneously, the data mapping relationship simplifies the blood oxygenation parameter inference process, making the entire blood oxygenation detection more efficient, accurate, and stable. The correlation coefficient is determined based on the linear correlation characteristics between target reflected light data and existing reflected light data, effectively evaluating the degree of data matching. When the correlation coefficient exceeds a preset threshold, it indicates a high degree of consistency between the two sets of data, allowing for the reliable selection of second reflected light data and accurate estimation of the target blood oxygenation parameter based on its corresponding blood oxygenation parameter. This simplifies the data matching process and improves the efficiency and stability of blood oxygenation detection.

[0108] This embodiment provides a method for generating blood oxygen images, which can be used in computer devices such as desktop computers and laptops. Figure 8 This is a flowchart of a blood oxygen image generation method according to an embodiment of the present invention, such as... Figure 8 As shown, the process includes the following steps:

[0109] Step S301: Acquire target reflected light data corresponding to the tissue sites to be detected at multiple preset wavelengths, as well as the data mapping relationship. The data mapping relationship is used to characterize the correspondence between reflected light data and blood oxygenation parameters. For details, please refer to... Figure 5 Step S201 of the illustrated embodiment will not be described again here.

[0110] Step S302: Determine the target blood oxygen parameters that match the target reflected light data according to the data mapping relationship. For details, please refer to [link to relevant documentation]. Figure 5 Step S202 of the illustrated embodiment will not be described again here.

[0111] Step S303: Visualize the target blood oxygen parameters to generate a blood oxygen image corresponding to the tissue site to be detected.

[0112] Specifically, step S303 includes:

[0113] Step S3031: Obtain the visualization mapping relationship. The visualization mapping relationship is used to characterize the correspondence between blood oxygen parameters and visualization parameters.

[0114] Visualization mapping is a pre-defined mapping rule used to characterize the correspondence between blood oxygenation parameters and visualization parameters. It determines how blood oxygenation parameters are presented visually so that doctors can intuitively understand the blood oxygenation status of tissues. Visualization parameters are parameters used to describe the visual characteristics of each pixel in an image; these parameters ultimately determine the image's display effect. Specifically, based on clinical experience and general understanding of changes in blood oxygenation in human tissues, doctors or researchers can pre-define some visualization mapping relationships. For example, one color or shape might be chosen to represent the normal blood oxygenation range, while another color or shape might be chosen to represent hypoxic conditions.

[0115] Step S3032: Based on the visual mapping relationship, the target blood oxygen parameters are converted into corresponding visual parameters.

[0116] Specifically, a lookup table can be created based on the visualization mapping relationship. This lookup table details the correspondence between target blood oxygen saturation parameters in different ranges and their corresponding visualization parameters. For example, blood oxygen saturation between 90% and 100% might correspond to red, 80% to 89% to orange, and 70% to 79% to yellow, etc. Once the target blood oxygen parameter is determined, the corresponding visualization parameter is looked up in the lookup table. This process can be achieved through simple indexing operations, quickly and accurately obtaining the visualization parameter that matches the target blood oxygen parameter.

[0117] Step S3033: Render the image according to the visualization parameters to generate a first image corresponding to the tissue area to be detected. The blood oxygenation image includes the first image.

[0118] The first image is a pseudo-color image generated by rendering the image according to visualization parameters. Specifically, the color of each pixel in the image can be set according to the color information in the converted visualization parameters. If the visualization parameter is specified as red, then the corresponding pixel will be rendered as red, which can be achieved using image editing software or graphics libraries in programming languages. For continuous ranges of blood oxygen parameters, color gradients can be used to display them, making the image clearer and more intuitive. For example, a gradual transition from red to orange to yellow represents a change in blood oxygen saturation from high to low. Alternatively, corresponding shapes can be drawn in the image according to the shape information specified in the visualization parameters. If the visualization parameters require a circle to represent a region of blood oxygen range, then a circle will be drawn at the corresponding pixel position in the image. The size and position of the shape can be adjusted according to the specific value of the target blood oxygen parameter. For example, the higher the blood oxygen saturation, the larger the circle may be, and the closer it may be to the center of the image.

[0119] By integrating the visual parameters such as color and shape of each pixel to form a complete image, it means that after all pixels are rendered according to the requirements of the visual parameters, they together constitute the first image of the tissue to be detected.

[0120] For example, after calculating the blood oxygen parameters for all pixels and obtaining a 1080×1440 blood oxygen parameter matrix (the specific image size depends on the actual image sensor specifications), the blood oxygen value matching unit 45 sends this matrix to the blood oxygen image generation unit 47. The blood oxygen image generation unit 47 includes a pseudo-color image generation unit 47a and an overlay image generation unit 47b. The principle of the pseudo-color image generation unit 47a is similar to that of thermal imaging and infrared imaging. It maps blood oxygen saturation values ​​from high to low to color changes, such as from red to purple, by providing a color mapping table (i.e., a visual mapping relationship). Other color mapping rules can also be set according to specific needs.

[0121] Step S3034: Obtain scan images of the tissue sites to be detected at multiple preset wavelengths.

[0122] A scanned image is an image acquired after scanning the tissue area to be examined at multiple preset wavelengths. Specifically, an optical system similar to that used to acquire target reflected light data is used, such as the light source and image sensor in an endoscope system. A cold light source device emits light of a specific wavelength to illuminate the tissue area to be examined. The light reflected from the tissue is transmitted through a series of optical elements, received by the image sensor, and converted into electrical signals. These electrical signals are then processed to form a scanned image.

[0123] Step S3035: If the target blood oxygen parameter corresponding to any target pixel in the scanned image is greater than the preset blood oxygen parameter value, then the color channel parameters corresponding to the target pixel are superimposed to obtain the first processing result.

[0124] The preset blood oxygen parameter value is a pre-defined threshold value used to distinguish the blood oxygenation status of tissues. Specifically, the color channel parameters (such as the red, green, and blue channels) corresponding to the target pixel are added or otherwise superimposed. This means that the color at that pixel location will be more vibrant or prominent, indicating an area with higher blood oxygen content. Through this superposition operation, the pixel's display effect in the image is enhanced, making it easier to observe and identify, helping doctors quickly find areas with sufficient blood oxygenation.

[0125] Step S3036: If the target blood oxygen parameter corresponding to the target pixel is less than the preset blood oxygen parameter value, then the color channel parameter corresponding to the target pixel is adjusted based on the preset color mapping rule to obtain the second processing result.

[0126] Preset color mapping rules are a set of pre-defined rules used to adjust the color channel parameters of a target pixel based on whether the target blood oxygenation parameter is lower than a preset blood oxygenation parameter value. For example, when the blood oxygenation parameter is lower than the preset value, the value of the red channel is decreased, while the values ​​of the green and blue channels are increased, causing the pixel to display a specific color combination, such as blue or cyan, to represent an area with insufficient blood oxygenation. Alternatively, the brightness value of the pixel can be adjusted to make it darker or grayer to highlight areas with low blood oxygenation. Specifically, the color channel parameters of the target pixel are modified according to the preset color mapping rules. These preset color mapping rules can be based on clinical experience or experimental data and are designed to visually display areas with insufficient blood oxygenation. By adjusting the color channel parameters, the pixel can present different characteristics in the image compared to areas with sufficient blood oxygenation, thereby alerting doctors to potential lesions or abnormalities.

[0127] Step S3037: Integrate the first processing results and the second processing results to generate a second image corresponding to the tissue site to be detected. The blood oxygenation image includes the second image.

[0128] The second image integrates the processing results of regions where the target blood oxygen parameter is greater than or less than the preset blood oxygen parameter value, providing a comprehensive view of the tissue's blood oxygen distribution. Specifically, the first and second processing results of each target pixel are merged to form a complete image dataset. Using image processing software or algorithms, these processing results are synthesized according to certain rules to generate the second image corresponding to the tissue area to be detected.

[0129] In addition to the blood oxygen image processing unit 41a, the image processing unit 41 in the endoscope system also includes a white light image processing unit 41b and a special light image processing unit 41c.

[0130] The white light image processing unit 41b, corresponding to the white light illumination mode, uses Principal Component Analysis (PCA) to compress a cube of hyperspectral data containing 30 visible light bands into three principal component wavelengths, thereby synthesizing a white light image. First, the data is standardized to ensure that PCA can effectively extract the main variation information. Then, PCA maps the high-dimensional spectral data to a low-dimensional space, selecting the first three principal components as the most important spectral information. Finally, the extracted first three principal components are mapped to the RGB channels respectively to generate the white light image.

[0131] Similarly, the special light image processing unit 41c is used to synthesize other feature-emphasized spectral information and map it to the RGB channels to generate corresponding special light images. For example, blue-violet light has weak penetrating power and is difficult to reach the deep layers of the mucosa, thus reducing the interference of deep mucosal substances on the imaging of superficial capillaries and highlighting the capillaries on the surface of the mucosa.

[0132] After all pixels have been assigned or modified for color, the image signal is output to the display 15 via the display control circuit 43. To achieve different display effects, output can also be achieved through picture-in-picture or other methods. Taking normal white light mode and blood oxygen distribution imaging mode as examples, when switching modes, the image data of the other mode is stored in the frame memory and read and output to the display 15 during the switch.

[0133] The blood oxygenation image generation method provided in this invention establishes a mapping relationship between blood oxygenation parameters and visualization parameters, ensuring the accuracy and consistency of blood oxygenation data conversion into images. Based on this mapping relationship, the target blood oxygenation parameter is converted into a visualization parameter, and a blood oxygenation image is generated through image rendering. This makes blood oxygenation information more intuitive and easier to understand, improving not only the visualization effect of blood oxygenation detection results but also enabling doctors or users to quickly and accurately obtain and interpret blood oxygenation information. By flexibly adjusting the image processing method based on the comparison between the target blood oxygenation parameter and the preset blood oxygenation parameter value, more accurate blood oxygenation information visualization is achieved. When the target blood oxygenation parameter is greater than the preset value, image details are enhanced by overlaying color channel parameters; when the target blood oxygenation parameter is less than the preset value, color channels are adjusted based on color mapping rules to ensure that the image presents more appropriate blood oxygenation information. Integrating these two processing results generates a more recognizable and accurate blood oxygenation image, making the detection results more intuitive and easier to interpret, thereby improving the effect and reliability of blood oxygenation detection.

[0134] This embodiment also provides a blood oxygenation image generation device, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0135] This embodiment provides a blood oxygenation image generation device, such as... Figure 9 As shown, it includes:

[0136] The acquisition module 401 is used to acquire target reflected light data corresponding to the tissue sites to be detected under multiple preset wavelengths, as well as data mapping relationships. The data mapping relationships are used to characterize the correspondence between reflected light data and blood oxygen parameters.

[0137] Matching module 402 is used to determine the target blood oxygen parameters that match the target reflected light data according to the data mapping relationship;

[0138] The generation module 403 is used to perform visualization processing based on the target blood oxygen parameters to generate a blood oxygen image corresponding to the tissue site to be detected.

[0139] In some optional implementations, the acquisition module 401 includes:

[0140] The first acquisition submodule is used to acquire optical characteristic parameters and blood oxygen characteristic parameters of sample tissue sites at different wavelengths.

[0141] The simulation submodule is used to perform simulations for any wavelength based on optical characteristic parameters and blood oxygen characteristic parameters, and to obtain tissue light transmission characteristic data of sample tissue sites under different combinations of characteristic parameters.

[0142] The first submodule is used to establish a first mapping relationship between the combination of feature parameters and the tissue optical transmission feature data.

[0143] The second submodule is used to acquire sample reflected light data of sample tissue at different wavelengths and establish a second mapping relationship between sample reflected light data and tissue light transmission characteristic data.

[0144] The third submodule is used to establish a data mapping relationship between sample reflected light data and blood oxygen characteristic parameters at different wavelengths, based on the first and second mapping relationships.

[0145] In some alternative implementations, the matching module 402 includes:

[0146] The comparison submodule is used to compare the target reflected light data with the reflected light data, and determine the first reflected light data with the smallest difference from the target reflected light data based on the comparison result;

[0147] The first determining submodule is used to determine the blood oxygen parameter corresponding to the first reflected light data as the target blood oxygen parameter.

[0148] In some alternative implementations, the matching module 402 further includes:

[0149] The second determination submodule is used to derive the correlation coefficient based on the linear correlation characteristics between the target reflected light data and the reflected light data.

[0150] The third determination submodule is used to determine the second reflected light data corresponding to the correlation coefficient if the correlation coefficient is greater than a preset threshold, and to determine the blood oxygen parameter corresponding to the second reflected light data as the target blood oxygen parameter.

[0151] In some optional implementations, the acquisition module 401 further includes:

[0152] The second acquisition submodule is used to acquire the initial reflected light data corresponding to the tissue sites to be detected under multiple preset wavelengths;

[0153] The first correction submodule is used to perform noise correction on the initial reflected light data to obtain corrected reflected light data.

[0154] The third acquisition submodule is used to acquire preset standard reference values, which are obtained based on a diffuse reflection standard whiteboard.

[0155] The second correction submodule is used to normalize the corrected reflected light data based on the standard reference value to obtain the target reflected light data.

[0156] In some alternative implementations, the generation module 403 includes:

[0157] The fourth acquisition submodule is used to acquire the visualization mapping relationship, which is used to characterize the correspondence between blood oxygen parameters and visualization parameters;

[0158] The conversion submodule is used to convert target blood oxygen parameters into corresponding visual parameters based on visual mapping relationships;

[0159] The first generation submodule is used to render images according to visualization parameters and generate a first image corresponding to the tissue area to be detected. The blood oxygenation image includes the first image.

[0160] In some alternative implementations, the generation module 403 further includes:

[0161] The fifth acquisition submodule is used to acquire scan images of the tissue sites to be detected at multiple preset wavelengths;

[0162] The first processing submodule is used to superimpose the color channel parameters corresponding to the target pixel if the target blood oxygen parameter corresponding to any target pixel in the scanned image is greater than the preset blood oxygen parameter value, so as to obtain the first processing result.

[0163] The second processing submodule is used to adjust the color channel parameters corresponding to the target pixel based on the preset color mapping rules if the target blood oxygen parameter corresponding to the target pixel is less than the preset blood oxygen parameter value, so as to obtain the second processing result.

[0164] The second generation submodule is used to integrate the various first processing results and second processing results to generate a second image corresponding to the tissue site to be detected, including the blood oxygenation image.

[0165] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0166] In this embodiment, the blood oxygen image generation device is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0167] The blood oxygenation image generation device provided in this invention analyzes the blood oxygenation status of the tissue to be tested from multiple angles by acquiring reflected light data at multiple preset wavelengths. Through data mapping, the target reflected light data is converted into blood oxygenation parameters. This mapping ensures a more accurate conversion between reflected light data and blood oxygenation parameters, reducing errors that may occur in direct measurement. This improves the matching accuracy of blood oxygenation parameters, enabling the final generated blood oxygenation image to more realistically and effectively reflect the tissue's blood oxygenation level. Through visualization processing, blood oxygenation parameters are converted into blood oxygenation images, making the originally abstract blood oxygenation data intuitive and easy to understand. Doctors and medical personnel can quickly identify blood oxygenation abnormalities through the images and intervene in a timely manner, thereby improving the accuracy of blood oxygenation assessment and diagnostic efficiency.

[0168] This invention also provides a computer device having the above-described features. Figure 9 The blood oxygen image generation device shown.

[0169] Please see Figure 10 , Figure 10 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 10As shown, the computer device includes one or more processors 100, memory 200, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 10 Take a processor 100 as an example.

[0170] Processor 100 may be a central processing unit, a network processor, or a combination thereof. Processor 100 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GPA), or any combination thereof.

[0171] The memory 200 stores instructions executable by at least one processor 100 to cause the at least one processor 100 to perform the method shown in the above embodiments.

[0172] The memory 200 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 200 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 200 may optionally include memory remotely located relative to the processor 100, which can be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0173] The memory 200 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 200 may also include a combination of the above types of memory.

[0174] The computer device also includes an input device 300 and an output device 400. The processor 100, memory 200, input device 300, and output device 400 can be connected via a bus or other means. Figure 10 Taking the example of a connection between China and Israel via a bus.

[0175] Input device 300 can receive input numerical or character information, and generate key signal inputs related to user settings and function control of the computer device, such as a touchscreen, keypad, mouse, trackpad, touchpad, joystick, one or more mouse buttons, trackball, joystick, etc. Output device 400 may include display devices, auxiliary lighting devices (e.g., LEDs), and haptic feedback devices (e.g., vibration motors). The aforementioned display devices include, but are not limited to, liquid crystal displays, light-emitting diodes, displays, and plasma displays. In some alternative embodiments, the display device may be a touchscreen.

[0176] The computer device also includes a communication interface for communicating with other devices or communication networks.

[0177] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0178] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0179] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for generating blood oxygenation images, characterized in that, The method includes: Acquire target reflected light data corresponding to the tissue sites to be detected at multiple preset wavelengths, as well as data mapping relationships. The data mapping relationships are used to characterize the correspondence between reflected light data and blood oxygenation parameters. Based on the data mapping relationship, determine the target blood oxygen parameters that match the target reflected light data; Visualization processing is performed based on the target blood oxygen parameters to generate a blood oxygen image corresponding to the tissue site to be detected.

2. The method according to claim 1, characterized in that, The data mapping relationship is pre-built and generated. The process of building the data mapping relationship includes: Obtain optical and blood oxygenation parameters of the sample tissue at different wavelengths. For any wavelength, simulations are performed based on the optical characteristic parameters and the blood oxygen characteristic parameters to obtain tissue light transmission characteristic data of the sample tissue under different combinations of characteristic parameters. Establish a first mapping relationship between the combination of feature parameters and the tissue optical transmission feature data; Acquire sample reflected light data of the sample tissue at different wavelengths, and establish a second mapping relationship between the sample reflected light data and the tissue light transmission characteristic data; Based on the first mapping relationship and the second mapping relationship, a data mapping relationship is established between the sample reflected light data and blood oxygen characteristic parameters at different wavelengths.

3. The method according to claim 1, characterized in that, The step of determining the target blood oxygen parameters that match the target reflected light data according to the data mapping relationship includes: The target reflected light data is compared with the reflected light data, and the first reflected light data with the smallest difference from the target reflected light data is determined based on the comparison result; The blood oxygen parameter corresponding to the first reflected light data is determined as the target blood oxygen parameter.

4. The method according to claim 3, characterized in that, Also includes: Based on the linear correlation between the target reflected light data and the reflected light data, a correlation coefficient is derived; If the correlation coefficient is greater than a preset threshold, then the second reflected light data corresponding to the correlation coefficient is determined, and the blood oxygen parameter corresponding to the second reflected light data is determined as the target blood oxygen parameter.

5. The method according to claim 1, characterized in that, The acquisition of target reflected light data corresponding to the tissue sites to be detected at multiple preset wavelengths includes: Acquire the initial reflected light data corresponding to the tissue site to be detected at the multiple preset wavelengths; The initial reflected light data is subjected to noise correction to obtain corrected reflected light data; Obtain a preset standard reference value, which is based on a diffuse reflection standard whiteboard; The target reflected light data is obtained by normalizing the corrected reflected light data based on the standard reference value.

6. The method according to claim 1, characterized in that, The step of visualizing the target blood oxygenation parameters to generate a blood oxygenation image corresponding to the tissue site to be detected includes: Obtain a visual mapping relationship, which is used to characterize the correspondence between the blood oxygen parameter and the visual parameter; Based on the visualization mapping relationship, the target blood oxygen parameters are converted into corresponding visualization parameters; Image rendering is performed according to the visualization parameters to generate a first image corresponding to the tissue area to be detected, and the blood oxygen image includes the first image.

7. The method according to claim 6, characterized in that, Also includes; Acquire scan images of the tissue sites to be detected at the multiple preset wavelengths; If the target blood oxygen parameter corresponding to any target pixel in the scanned image is greater than the preset blood oxygen parameter value, then the color channel parameters corresponding to the target pixel are superimposed to obtain the first processing result; If the target blood oxygen parameter corresponding to the target pixel is less than the preset blood oxygen parameter value, the color channel parameter corresponding to the target pixel is adjusted based on the preset color mapping rule to obtain the second processing result; The first processing results and the second processing results are integrated to generate a second image corresponding to the tissue site to be detected, and the blood oxygenation image includes the second image.

8. A blood oxygenation image generation device, characterized in that, The device includes: The acquisition module is used to acquire target reflected light data corresponding to the tissue sites to be detected under multiple preset wavelengths, as well as data mapping relationships. The data mapping relationships are used to characterize the correspondence between reflected light data and blood oxygen parameters. A matching module is used to determine the target blood oxygen parameters that match the target reflected light data according to the data mapping relationship; The generation module is used to perform visualization processing based on the target blood oxygen parameters to generate a blood oxygen image corresponding to the tissue site to be detected.

9. A computer device, characterized in that, include: A memory and a processor are communicatively connected, the memory stores computer instructions, and the processor executes the blood oxygen image generation method according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to perform the blood oxygenation image generation method according to any one of claims 1 to 7.

Images