A precise photoacoustic imaging method for deep tissue blood oxygen saturation

Abstract

The application belongs to the technical field of medical imaging, and discloses a deep tissue blood oxygen saturation photoacoustic precise imaging method, which comprises the following steps: S1, obtaining a light flux distribution preliminary model of a to-be-measured region; S2, photoacoustic imaging is performed on the to-be-measured tissue to obtain original photoacoustic imaging data of the to-be-measured tissue; S3, calculating arterial photoacoustic signal intensity; S4, calculating the relative light flux distribution of each arterial region to form an arterial light flux space correction network; S5, introducing the arterial light flux space correction network into the light flux preliminary model to correct the light flux distribution preliminary model, and obtaining the corrected relative light flux distribution; and S6, calculating the blood oxygen saturation of the to-be-measured tissue in the region other than the artery according to the relative light flux of the artery and the to-be-measured tissue imaging region other than the artery, and obtaining the blood oxygen saturation distribution of the entire to-be-measured tissue imaging region. The application calibrates the light flux distribution in the deep tissue through the arterial photoacoustic signal, and realizes the precise measurement of the blood oxygen saturation of the deep tissue.

Description

Technical Field

[0001] This invention belongs to the field of medical imaging technology, and in particular relates to a precise photoacoustic imaging method for deep tissue blood oxygen saturation. Background Technology

[0002] The development of optical imaging technology has made it possible to measure blood oxygen saturation in deep tissues. Diffused light tomography has enabled non-invasive measurement of breast blood oxygenation and has correlated the hypoxic characteristics of breast tumors with tumor benignity / malignancy and response to neoadjuvant therapy. However, the low resolution of this imaging technique limits its effective clinical application. Photoacoustic tomography has overcome this limitation and fully leveraged its advantages in high-speed, high-resolution imaging of biological structures, functions, and molecular information in small animal imaging. In human tissues, hemoglobin is the main absorbing chromophore in the visible and near-infrared spectra. Therefore, photoacoustic functional imaging can measure the concentrations of oxyhemoglobin and deoxyhemoglobin separately by detecting the absorption of hemoglobin under different wavelengths of illumination, thereby obtaining the blood oxygen saturation within the tissue. Since the intensity of the photoacoustic signal is proportional to the product of the molar absorption coefficient (known), concentration (to be measured) of the tissue component, and the luminous flux (energy of pulsed light per unit area), accurate quantitative photoacoustic imaging depends on the accurate estimation of the luminous flux in the area to be measured. When using traditional linear fitting methods to calculate the content of each component, the measurement error usually increases with depth due to the uncertainty of light flux distribution (for example, the blood oxygen measurement error reaches 15% at an imaging depth of 0.5 cm). Most current photoacoustic quantitative imaging methods rely on the similarity between the absorption characteristics of the tissue under test and the simulation model. For example: ① By applying principal component analysis (PCA) to the simulated light flux spectrum, multiple characteristic spectra are selected and combined to predict the light flux at different wavelengths in biological tissues. This method shows higher blood oxygen measurement accuracy than traditional linear fitting methods in small animal imaging (<1 cm imaging depth); ② Introducing diffused light tomography or Monte Carlo simulation into the photoacoustic imaging system to estimate the light distribution in the tissue; ③ Using convolutional neural networks trained with unsupervised "simulation to experimental domain" conversion data to quantitatively estimate the tissue's light absorption. This method has not been fully validated in in vivo imaging, and its stability and universality need to be verified.

[0003] In practical applications, no follow-up studies have been conducted after the publication of the above methods, and there are no reports of other projects continuing to use these methods. In fact, most current photoacoustic quantitative imaging methods rely on the similarity between the absorption characteristics of the tissue being tested and the simulation model. Therefore, the uncertainty of tissue differences between biological individuals and within individuals affects the stability of the above measurement methods. Currently, there is no accurate, clear, and stable non-invasive imaging method for testing the local blood oxygen content of deep tissues (below the body surface > 0.5 cm). Therefore, the measurement accuracy of quantitative photoacoustic imaging (functional / molecular imaging) still faces challenges. Summary of the Invention

[0004] The purpose of this invention is to provide a precise photoacoustic imaging method for deep tissue blood oxygen saturation to solve the above-mentioned technical problems.

[0005] To address the aforementioned technical problems, the specific technical solution of the photoacoustic precision imaging method for deep tissue blood oxygen saturation according to the present invention is as follows:

[0006] A precise photoacoustic imaging method for deep tissue blood oxygen saturation includes the following steps:

[0007] S1. Before testing, measure the surface illumination range of the tissue to be tested and calculate a preliminary model of the light flux distribution in the test area.

[0008] S2. Perform photoacoustic imaging on the tissue to be tested to obtain the original photoacoustic imaging data of the tissue to be tested; or perform photoacoustic imaging and ultrafast ultrasound imaging on the tissue to be tested to obtain the original photoacoustic imaging and high-resolution ultrasound imaging data of the tissue to be tested.

[0009] S3. Based on the original photoacoustic imaging data of the tissue under test or based on the original photoacoustic imaging and high-resolution ultrasound imaging data of the tissue under test, segment the arterial blood vessel image and calculate the arterial photoacoustic signal intensity.

[0010] S4. Calculate the relative luminous flux distribution of each arterial region based on the intensity of the arterial photoacoustic signal, and construct an arterial luminous flux spatial correction network U.

[0011] S5. Introduce the arterial light flux spatial correction network into the preliminary light flux model in step S1, and use the arterial light flux spatial correction network as a reference to correct the preliminary light flux distribution model to obtain the corrected relative light flux distribution.

[0012] S6. In the corrected relative light flux distribution, the blood oxygen saturation of the arterial region is defined as the normal blood oxygen saturation value of the arterial region. The blood oxygen saturation of the tissue to be tested in the region outside the artery is calculated based on the relative light flux of the artery and the imaging region of the tissue to be tested outside the artery, and the blood oxygen saturation distribution of the entire imaging region of the tissue to be tested is obtained.

[0013] Furthermore, S4 calculates the relative luminous flux distribution of each arterial region based on prior information using the following equation, thus forming an arterial luminous flux spatial correction network U;

[0014]

[0015]

[0016] Among them, photoacoustic signals @波长1It is the luminous flux of light with wavelength 1 shining on the location to be measured, which can be measured; photoacoustic signal @波长2 It is the luminous flux of light with wavelength 2 shining on the same location to be measured, and can be obtained by measurement; This refers to the relative concentration of oxyhemoglobin. HbR It is the relative concentration of reduced hemoglobin; ε HbO2@波长1 It is the molar absorptivity of monooxyhemoglobin under wavelength irradiation; ε HbO2@波长2 It is the molar absorptivity of oxyhemoglobin under irradiation at wavelength 2; ε HbR@波长1 It is the molar absorptivity of reduced hemoglobin under irradiation at wavelength 1; ε HbR@波长2 This is the molar absorption coefficient of reduced hemoglobin under wavelength 2 irradiation; the above molar absorption coefficient is a known value; the photoacoustic signal coefficient is a constant related to ultrasound detection, so the signal coefficients at the same test location under different wavelengths can cancel each other out. The relative concentration of oxyhemoglobin HbO2 in arterial vessels is approximately 0.98, and the concentration of hemoglobin HbO2+HbR in each blood vessel of the breast is approximately equal, denoted as 1.

[0017] Furthermore, S4, in conjunction with the optical flux spatial correction network U, calculates the blood oxygen saturation in the region outside the artery according to the following formula, and then calculates the blood oxygen saturation distribution of the entire test area:

[0018]

[0019] Furthermore, the specific method for correcting the preliminary luminous flux model V in step S5 is as follows:

[0020] S5.1, Normalized relative luminous flux U / V in the arterial region;

[0021] S5.2. A continuous relative luminous flux distribution is generated by normalizing the relative luminous flux through the radial basis function (RBF) kernel difference.

[0022] S5.3, the continuous relative luminous flux distribution Anti-normalization

[0023] Furthermore, a dual-wavelength light source was used to measure the relative concentrations of oxyhemoglobin and deoxyhemoglobin, thus offsetting the influence of the boundary effect of photoacoustic imaging and the diffraction effect of ultrasonic detection on the measurement accuracy.

[0024] Furthermore, in addition to using a dual-wavelength light source, more near-infrared wavelengths are added to improve the accuracy of linear fitting.

[0025] Furthermore, the high-resolution ultrasound imaging module integrated into the photoacoustic imaging platform is used to measure blood flow velocity in blood vessels. Taking advantage of the higher blood flow velocity in arteries compared to veins, the sensitivity and accuracy of the imaging system in identifying small arteries are improved.

[0026] Furthermore, S3 uses the photon diffusion equation to calculate the intensity of the arterial photoacoustic signal.

[0027] Furthermore, the photoacoustic imaging platform includes a dual-wavelength light source, an imaging shaping bowl, and an ultrafast ultrasonic imaging module.

[0028] The photoacoustic precision imaging method for deep tissue oxygen saturation of this invention has the following advantages: This invention quantifies the vascular proliferation and blood oxygen distribution characteristics of the imaging, providing effective photoacoustic imaging information, including the average and variance of blood oxygen saturation, the lowest blood oxygen saturation in the internal and surrounding tissues, the distribution range and dispersion of low-oxygen tissues, the gradient of blood oxygen change at the boundary, and the vascular proliferation characteristics of the microenvironment (density, curvature, and distribution irregularity). This invention achieves precise measurement of blood oxygen saturation in deep tissues by calibrating the light flux distribution in deep tissues using arterial photoacoustic signals. Attached Figure Description

[0029] Figure 1 This is a step diagram of the first embodiment of the present invention;

[0030] Figure 2 This is a step diagram of the preliminary model (V) for correcting luminous flux in the first embodiment of the present invention;

[0031] Figure 3 This is a step diagram of the second embodiment of the present invention;

[0032] Figure 4 This is a schematic diagram illustrating the operation of a specific embodiment of the present invention. Detailed Implementation

[0033] To better understand the purpose, structure, and function of this invention, the following detailed description of a photoacoustic precision imaging method for deep tissue blood oxygen saturation, in conjunction with the accompanying drawings, is provided. Specific Implementation Example 1:

[0035] like Figure 1 As shown in the figure, this specific embodiment provides a method for precise photoacoustic imaging of deep tissue blood oxygen saturation, characterized by the following steps:

[0036] S1. Before testing, measure the surface illumination range of the tissue to be tested, based on the average optical parameters of the breast (μ). a ,μ′ s A preliminary model of the luminous flux distribution in the area to be measured (V) is calculated by modeling the luminous flux distribution of the relative intensity.

[0037] S2. Perform photoacoustic imaging on the tissue to be tested to obtain the raw photoacoustic imaging data of the tissue to be tested;

[0038] S3. Based on the original photoacoustic imaging data of the tissue to be tested, segment the arterial blood vessel image and calculate the photoacoustic signal intensity of the artery;

[0039] S4. Based on the intensity of the photoacoustic signal of the artery, calculate the relative luminous flux distribution of each arterial region to form an arterial luminous flux spatial correction network (U);

[0040] S5. Introduce the arterial light flux spatial correction network (U) into the light flux preliminary model (V) in step S1, and correct the light flux distribution preliminary model (V) based on the arterial light flux spatial correction network (U) to obtain the corrected relative light flux distribution (X).

[0041] S6. In the corrected relative light flux distribution (X), the blood oxygen saturation of the arterial region is defined as 97.5%. The blood oxygen saturation of the tissue outside the arterial region is calculated based on the relative light flux of the imaging region of the tissue outside the arterial region, and the blood oxygen saturation distribution of the entire imaging region of the tissue is obtained.

[0042] Preferably, step S4 uses the photon diffusion equation to model the luminous flux distribution of relative intensity.

[0043] Preferably, step S4 calculates the relative luminous flux distribution of each arterial region based on prior information using the following equation, thus forming an arterial luminous flux spatial correction network (U);

[0044]

[0045]

[0046] Among them, photoacoustic signals @1064 It is the luminous flux of light with a wavelength of 1064nm illuminating the location (pixel) to be measured, which can be obtained; photoacoustic signal. @755 It is the luminous flux of 755nm wavelength light shining on the same measurement location (pixel), which can be measured; This refers to the relative concentration of oxyhemoglobin. HbR It is the relative concentration of reduced hemoglobin; ε HbO2@1064 It is the molar absorptivity of oxyhemoglobin under irradiation at a wavelength of 1064 nm; ε HbO2@755 It is the molar absorptivity of oxyhemoglobin under 755nm wavelength irradiation; ε HbR@1064 It is the molar absorptivity of reduced hemoglobin under irradiation at a wavelength of 1064 nm; ε HbR@755This is the molar absorption coefficient of reduced hemoglobin under 755nm wavelength irradiation; the above molar absorption coefficient is a known value; the photoacoustic signal coefficient is a constant related to ultrasound detection, so the signal coefficients at the same test location (pixel) under different wavelengths can cancel each other out. The relative concentration of oxyhemoglobin (HbO2) in arterial blood vessels is approximately 0.98, and the concentration of hemoglobin (HbO2+HbR) in each blood vessel of the breast can be approximately equal, denoted as 1.

[0047] In this embodiment, the oxygen saturation of the region outside the artery is calculated according to the following formula by combining the optical flux spatial correction network (U), and then the oxygen saturation distribution of the entire test area is calculated.

[0048]

[0049] like Figure 2 As shown in this embodiment, the specific method for correcting the preliminary luminous flux model (V) in step S5 is as follows:

[0050] S5.1, Normalized relative luminous flux (U / V) in the arterial region;

[0051] S5.2. A continuous relative luminous flux distribution is generated by normalizing the relative luminous flux using radial basis function (RBF) kernel difference values.

[0052] S5.3, the continuous relative luminous flux distribution Anti-normalization

[0053] Preferably, a dual-wavelength light source of 1064nm and 755nm is used to measure the relative concentrations of oxyhemoglobin and deoxyhemoglobin to offset the influence of the boundary effect of photoacoustic image and the diffraction effect of ultrasonic detection on the measurement accuracy.

[0054] Preferably, to further improve the accuracy or stability of deep tissue oxygen saturation measurement, more wavelengths can be added to improve the accuracy of linear fitting, based on the use of a dual-wavelength light source of 1064nm and 755nm. Preferably, adding more near-infrared wavelengths further improves the accuracy of linear fitting.

[0055] One photoacoustic imaging platform in this embodiment includes an imaging shaping bowl and a dual-wavelength light source of 1064nm and 755nm. Specific Implementation Example 2:

[0057] like Figure 3As shown, in order to further improve the accuracy or stability of deep tissue blood oxygen saturation measurement, a high-resolution ultrasound imaging module integrated into the photoacoustic imaging platform is used to measure the blood flow velocity in blood vessels. Taking advantage of the fact that arterial blood flow velocity is higher than that of veins, the sensitivity and accuracy of the imaging system in identifying small arteries are improved, thereby extracting arterial vessels more accurately.

[0058] S1. Before testing, measure the surface illumination range of the tissue to be tested, based on the average optical parameters of the breast (μ). a ,μ′ s A preliminary model of the luminous flux distribution in the area to be measured (V) is calculated by modeling the luminous flux distribution of the relative intensity.

[0059] S2. Perform photoacoustic imaging and ultrafast ultrasound imaging on the tissue to be tested to obtain the original photoacoustic imaging and high-resolution ultrasound imaging data of the tissue to be tested.

[0060] S3. Based on the original photoacoustic imaging and high-resolution ultrasound imaging data of the tissue under test, the arterial blood vessel image is segmented more accurately, and the intensity of the arterial photoacoustic signal is calculated.

[0061] S4. Based on the intensity of the photoacoustic signal of the artery, calculate the relative luminous flux distribution of each arterial region to form an arterial luminous flux spatial correction network (U);

[0062] S5. Introduce the arterial light flux spatial correction network (U) into the light flux preliminary model (V) in step S1, and correct the light flux distribution preliminary model (V) based on the arterial light flux spatial correction network (U) to obtain the corrected relative light flux distribution (X).

[0063] S6. In the corrected relative light flux distribution (X), the blood oxygen saturation of the arterial region is defined as 97.5%. The blood oxygen saturation of the tissue outside the arterial region is calculated based on the relative light flux of the imaging region of the tissue outside the arterial region, and the blood oxygen saturation distribution of the entire imaging region of the tissue is obtained.

[0064] In this embodiment, step S3 uses the photon diffusion equation to calculate the intensity of the arterial photoacoustic signal.

[0065] In this embodiment, step S4 calculates the relative luminous flux distribution of each arterial region based on prior information using the following equation, thus forming an arterial luminous flux spatial correction network (U).

[0066]

[0067]

[0068] Among them, photoacoustic signals @1064It is the luminous flux of light with a wavelength of 1064nm illuminating the location (pixel) to be measured, which can be obtained; photoacoustic signal. @755 It is the luminous flux of 755nm wavelength light shining on the same measurement location (pixel), which can be measured; This refers to the relative concentration of oxyhemoglobin. HbR It is the relative concentration of reduced hemoglobin; ε HbO2@1064 It is the molar absorptivity of oxyhemoglobin under irradiation at a wavelength of 1064 nm; ε HbO2@755 It is the molar absorptivity of oxyhemoglobin under 755nm wavelength irradiation; ε HbR@1064 It is the molar absorptivity of reduced hemoglobin under irradiation at a wavelength of 1064 nm; ε HbR@755 This is the molar absorption coefficient of reduced hemoglobin under 755nm wavelength irradiation; the above molar absorption coefficient is a known value; the photoacoustic signal coefficient is a constant related to ultrasound detection, so the signal coefficients at the same test location (pixel) under different wavelengths can cancel each other out. The relative concentration of oxyhemoglobin (HbO2) in arterial blood vessels is approximately 0.98, and the concentration of hemoglobin (HbO2+HbR) in each blood vessel of the breast can be approximately equal, denoted as 1.

[0069] In this embodiment, the oxygen saturation of the region outside the artery is calculated according to the following formula by combining the optical flux spatial correction network (U), and then the oxygen saturation distribution of the entire test area is calculated.

[0070]

[0071] like Figure 2 As shown in this embodiment, the specific method for correcting the preliminary luminous flux model (V) in step S5 is as follows:

[0072] S5.1, Normalized relative luminous flux (U / V) in the arterial region;

[0073] S5.2. A continuous relative luminous flux distribution is generated by normalizing the relative luminous flux using radial basis function (RBF) kernel difference values.

[0074] S5.3, the continuous relative luminous flux distribution Anti-normalization

[0075] Preferably, a dual-wavelength light source of 1064nm and 755nm is used to measure the relative concentrations of oxyhemoglobin and deoxyhemoglobin to offset the influence of the boundary effect of photoacoustic image and the diffraction effect of ultrasonic detection on the measurement accuracy.

[0076] Preferably, to further improve the accuracy or stability of deep tissue oxygen saturation measurement, more wavelengths can be added to improve the accuracy of linear fitting, based on the use of a dual-wavelength light source of 1064nm and 755nm. Preferably, adding more near-infrared wavelengths further improves the accuracy of linear fitting.

[0077] Preferably, to further improve the accuracy or stability of deep tissue blood oxygen saturation measurement, a high-resolution ultrasound imaging module integrated into a photoacoustic imaging platform is used to measure blood flow velocity in blood vessels. Taking advantage of the higher arterial blood flow velocity compared to veins, the sensitivity and accuracy of the imaging system in identifying small arteries are improved.

[0078] This embodiment of a photoacoustic imaging platform includes an imaging shaping bowl, a multi-wavelength light source, and an ultrafast ultrasonic imaging module.

[0079] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A method for photoacoustic precise imaging of deep tissue oxygen saturation, characterized in that, The method comprises the following steps: S1, measure the surface light range of the tissue to be measured before testing, and calculate a preliminary model of the light flux distribution of the region to be measured ; S2, photoacoustic imaging is performed on the to-be-tested tissue to obtain original photoacoustic imaging data of the to-be-tested tissue; or photoacoustic imaging and ultrafast ultrasonic imaging are performed on the to-be-tested tissue to obtain original photoacoustic imaging and high-resolution ultrasonic imaging data of the to-be-tested tissue; S3, according to the original photoacoustic imaging data of the to-be-tested tissue or according to the original photoacoustic imaging and high-resolution ultrasonic imaging data of the to-be-tested tissue, an arterial blood vessel image is segmented out, and an arterial photoacoustic signal intensity is calculated; S4、According to the arterial photoacoustic signal intensity, the relative light flux distribution of each arterial region is calculated to form an arterial light flux space correction network ; The S4 calculates the relative light flux distribution of each arterial region according to prior information by the following equation to form an arterial light flux space correction network ; = ， ， wherein the photoacoustic signal @波长1 is the light flux of wavelength 1 at the measured position, which can be measured; the photoacoustic signal @波长2 is the light flux of wavelength 2 at the same measured position, which can be measured; is the relative concentration of oxyhemoglobin, is the relative concentration of reduced hemoglobin; is the molar absorption coefficient of oxyhemoglobin under wavelength 1 irradiation; is the molar absorption coefficient of oxyhemoglobin under wavelength 2 irradiation; is the molar absorption coefficient of reduced hemoglobin under wavelength 1 irradiation; is the molar absorption coefficient of reduced hemoglobin under wavelength 2 irradiation; the above molar absorption coefficients are known values; the photoacoustic signal coefficient is a constant related to ultrasonic detection, so the signal coefficients at different wavelengths at the same measured position can be offset, the relative concentration of oxyhemoglobin HbO2 in the arterial blood vessels is approximately 0.98, and the concentrations of hemoglobin HbO2+HbR in each blood vessel of the breast are approximately equal, denoted as 1; S5, introducing the arterial light flux spatial correction network into the light flux distribution preliminary model in step S1 In the method, the light flux distribution preliminary model is corrected based on the arterial light flux spatial correction network to obtain a corrected relative light flux distribution. S6, in the corrected relative light flux distribution, a blood oxygen saturation of an arterial region is defined as an arterial region normal blood oxygen saturation value, a blood oxygen saturation of a region other than the artery in the to-be-tested tissue is calculated according to the relative light flux of the arterial and the region other than the artery in the to-be-tested tissue imaging region, and a blood oxygen saturation distribution of the entire to-be-tested tissue imaging region is obtained.

2. The deep tissue blood oxygen saturation photoacoustic precision imaging method of claim 1, wherein, The s4 combines light flux spatial correction network The blood oxygen saturation of the region other than the artery is calculated according to the following formula, and then the blood oxygen saturation distribution of the whole region to be measured is calculated: Oxygen saturation .

3. The deep tissue blood oxygen saturation photoacoustic precision imaging method of claim 1, wherein, The S5 preliminary model of light flux distribution The specific method of correction is as follows: S5.

1. Normalizing the relative luminous flux of the arterial region ; S5.2, generating a continuous relative luminous flux distribution by radial basis function, RBF, kernel difference normalization of the relative luminous flux ; S5.3, said continuous relative luminous flux distribution is divided into a number of sub- ranges, each sub-range having a sub-range width, and un-normalized .

4. The deep tissue blood oxygen saturation photoacoustic precision imaging method of claim 1, wherein, The relative concentrations of oxyhemoglobin and reduced hemoglobin are measured by using a dual-wavelength light source, so as to offset the influence of the boundary effect of the photoacoustic image and the diffraction effect of the ultrasonic detection on the measurement accuracy.

5. The deep tissue blood oxygen saturation photoacoustic precision imaging method of claim 3, wherein, On the basis of using the dual-wavelength light source, more near-infrared wavelengths are added to improve the accuracy of linear fitting.

6. The deep tissue blood oxygen saturation photoacoustic precision imaging method of claim 1, wherein, The blood flow rate in the blood vessel is measured by using a high-resolution ultrasonic imaging module integrated in a photoacoustic imaging platform, the sensitivity and accuracy of the imaging system in identifying small arteries are improved by using the fact that the arterial blood flow rate is higher than the venous blood flow rate.

7. The deep tissue blood oxygen saturation photoacoustic precision imaging method of claim 1, wherein, The S3 uses a photon diffusion equation to calculate the arterial photoacoustic signal intensity.

8. The deep tissue blood oxygen saturation photoacoustic precision imaging method of claim 6, wherein, The photoacoustic imaging platform comprises a dual-wavelength light source, an imaging shaping bowl and an ultrafast ultrasonic imaging module.

Images