LU102578B1

LU102578B1 - Application of three-dimensional (3d) photoacoustic imaging in breast tumor scoring system and device

Info

Publication number: LU102578B1
Application number: LU102578A
Authority: LU
Inventors: Sirui Liu; Meng Yang; Rui Zhang; Zhenhong Qi; Na Su; Yuxin Jiang; Jianchu Li; Tianhong Tang; Yao Wei; Chenyang Zhao; Yang Gui; Ming Wang
Original assignee: Peking Union Medical College Hospital
Priority date: 2019-08-01
Filing date: 2020-01-10
Publication date: 2021-06-30
Also published as: LU102578A1; WO2021017418A1

Abstract

The scoring system includes the following steps: (1) acquiring image information of a breast tumor in vitro through photoacoustic/ultrasound dual-modality imaging; (2) analyzing acquired image information, and conducting morphological and functional scoring, separately; (3) obtaining a comprehensive score, based on the results of the morphological and functional scoring, to determine whether the breast tumor has the malignant potential. For example, if one or all of the parameters of either morphological scoring or functional scoring are regarded as the malignant potential, the breast tumor is considered to be a malignant tumor. The application of 3D tumor imaging can provide more stable quantitative results. Compared with single breast 3D photoacoustic imaging, in the case where ultrasonic imaging is conducted to describe a tumor region, the external and internal characteristics of the tumor can be analyzed separately to improve the diagnostic sensitivity and specificity.

Description

Description LU102578 APPLICATION OF THREE-DIMENSIONAL (3D) PHOTOACOUSTIC IMAGING IN

BREAST TUMOR SCORING SYSTEM AND DEVICE

TECHNICAL FIELD The present invention relates to the technical field of medical diagnosis, and in particular to the application of three-dimensional (3D) photoacoustic imaging in the breast tumor scoring system and device.

BACKGROUND Breast cancer is the most common malignant tumor among women, and it shows an increasing high incidence globally in the 21st century. In China, the occurrence of breast cancer in women, ranks first among all other tumors. Breast cancer has become a major public health problem threatening human health.

Currently, mammography and breast ultrasonography (US) are the two most widely used methods for breast cancer screening. Mammography lacks the ability to provide morphological information and it does not work well for dense breasts, which limits its popularization. US is more sensitive in providing information about the morphology and boundaries of breast lesions, regardless of the breast density. In addition, Color Doppler Ultrasound (CDUS) and Power Doppler Imaging (PDI) can provide detailed vascular characteristics of a lesion, which enhances the diagnostic confidence in differentiating breast cancer. Although a lot of work has been done to make an algorithm for quantitative analysis of the above imaging modalities, a reliable quantitative diagnosis method has not been developed, and the diagnosis still highly depends on a doctor's personal experience. These conventional imaging modes still show limited accuracy in the diagnosis of early breast cancer, especially those who do not have typical morphological characteristics. Due to this disadvantage, many patients, especially those with advanced breast cancer, must undergo invasive examinations to acquire more diagnostic information and treatment information.

In a novel fusion imaging technology, photoacoustic ultrasonic dual-modality imaging (PA/US) combines the high contrast of optical imaging and an extensive depth of an ultrasound. PA breaks through the deep barriers of high-resolution optical imaging in biological tissues, making it suitable for breast imaging. With the rapid development of PA, a number of clinical research reports of breast PA have been reported, but most of them are based on two- dimensional (2D) PA/US imaging. The 2D PA breast tumor diagnosis mainly depends on the subjective judgment of a radiologist on a lesion image, including: (1) selecting slices for image evaluation and analysis and (2) image semi-quantitative scoring, and the above process completely relies on the experience of the radiologist. Therefore, an objective image quantitative evaluation method is of significant value for improving the diagnostic accuracy of LU102578 breast lesions.

At present, in the clinical field, there are no mature 3D devices that can be used in a tumor scoring system of photoacoustic/ultrasound imaging technology.

SUMMARY The present invention is intended to provide a scoring system and device using breast tumor 3D photoacoustic imaging, which can use quantitative parameters to distinguish malignant and benign tumors. In addition, the present invention uses a 3D imaging method, which can provide more stable and objective quantitative results than 2D imaging.

One aspect of the present invention is to provide the application of 3D photoacoustic imaging in a breast tumor scoring system, including the following steps: (1) acquiring image information of a breast tumor in vitro through photoacoustic/ultrasound dual-modality imaging; (2) analyzing acquired image information, and conducting morphological and functional scoring, separately; (3) obtaining a comprehensive score, based on the results of the morphological and functional scoring, to determine whether the breast tumor has the malignant potential. For example, if one or all the parameters of either morphological scoring or functional scoring are regarded as the malignant potential, the breast tumor is considered to be a malignant tumor.

For the above use, preferably, the functional scoring may use tumor oxygen saturation values which are quantitatively calculated through the acquired image information as evaluation criteria; the oxygen saturation values may include an oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of a tumor; and a condition that both the oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of the tumor is less than 0.75 to 0.80 may be used as an evaluation criterion for hypoxic malignant potential.

The oxygen saturation value inside a tumor and the oxygen saturation (SO2) value in the periphery of a tumor may be calculated by the following formula: SO2(r)= CH. CHe(r)+ CaeHb(r))(PA(1,1)* EdeHb(A2)- PA(a2,1)* EdeHb(A1))(PA(1,1)*( Edenb(A2)- EHb(A2))+ PA(A2,1)*( EHb(A1)- EdeHb(A1)) where, Hb represents endogenous oxygenated hemoglobin and deHb represents deoxygenated hemoglobin, PA(A1r) *= pa(di,r)= Cho(r)Enb(41)+CdeHb(T)EdeHb(A1) PA(A2,r) *= pa(A2,1)= Cup(r)enp(A2)+Caenn(r)Edenn(A2) M=750 nm, A2= 830 nm.

For the above use, preferably, an evaluation criterion for the functional scoring may further LU102578 include vascular density, namely, the vascular density of a tumor region and a surrounding area of the tumor, which is acquired by dividing the calculated number of pixels with SO2 > 40% by the total number of pixels in a corresponding region.

For the above use, preferably, the tumor may include invasive breast cancer (IBC) and breast intraductal tumors.

For the above use, preferably, the breast tumor may be T1 stage IBC.

For the above use, preferably, the morphological scoring may adopt the following acquired image information as evaluation criteria: ~~ abundance of blood flow inside a tumor, volume of blood flow in the periphery of a tumor, spatial running of blood vessels in a tumor, and infiltration of blood vessels into a tumor.

For the above use, preferably, the abundance of blood flow inside a tumor may be scored as follows: O-none, 1-little, and 2-rich; the volume of blood flow in the periphery of a tumor may be scored as follows: 0-none, 1-little, and 2-rich; the spatial running of blood vessels in a tumor may be scored as follows: O-regular and 1-irregular; and the infiltration of blood vessels into a tumor may be scored as follows: 0-No and 1-Yes. Specifically, if the abundance of blood flow inside a tumor gets a score indicating little, a ratio of the volume of internal blood vessels to the nodule volume is lower than 50%, and if the abundance of blood flow inside a tumor gets a score indicating rich, a ratio of the volume of internal blood vessels to the nodule volume is higher than 50%; if the volume of blood flow in the periphery of a tumor gets a score indicating little, a distribution range of blood vessels in the periphery of the tumor is less than 50% of the nodule volume, and if the volume of blood flow in the periphery of a tumor gets a score indicating rich, a distribution range of blood vessels in the periphery of the tumor is larger than 50% of the nodule volume; if the spatial running of blood vessels in a tumor gets a score indicating regular, blood vessels run smoothly into strips, with uniform diameters, and branches gradually change from thick to thin, and if the spatial running of blood vessels in a tumor gets a score indicating irregular, blood vessels run distortedly and show irregular dilation, and anastomotic branches are disordered. More specifically, if the abundance of blood flow inside a tumor gets a score indicating rich, it is determined to have malignant potential; if the volume of blood flow in the periphery of a tumor gets a score indicating rich, it is determined to have malignant potential; if the spatial running of blood vessels in a tumor gets a score indicating irregular, it is determined to have malignant potential; and if the infiltration of blood vessels into a tumor gets a score indicating Yes, it is determined to have malignant potential.

For the above use, preferably, the morphological scoring may further involve the following acquired image information: volume of tumor vessels and spatial distribution of tumor vessels,

where; a ratio of the volume of tumor vessels to a nodule volume may be scored as follows: 0- LU102578 none, 1-low, and 2-high; and the spatial distribution of tumor vessels may be scored as follows: O-uniform and 1-nonuniform. Specifically, if the ratio of the volume of tumor vessels to a nodule volume gets a score indicating low, the volume of vessels accounts for less than 50% of the nodule volume, and if the ratio of the volume of tumor vessels to a nodule volume gets a score indicating high, the volume of vessels accounts for more than 50% of the nodule volume; and if the spatial distribution of tumor vessels gets a score indicating uniform, blood vessels are evenly distributed at symmetrical parts of the nodule, with consistent numbers and diameters, and if the spatial distribution of tumor vessels gets a score indicating nonuniform, blood vessels are unevenly distributed at symmetrical parts of the nodule, with inconsistent numbers and diameters.

In another aspect of the present invention, a breast tumor scoring system based on photoacoustic/ultrasonic dual-modality imaging technology is provided, and the system includes an information acquisition module, an information analysis module, a calculation and output module, and a determination module. The information acquisition module is connected to a photoacoustic/ultrasonic imaging device to acquire image information characteristic parameters of a breast tumor tissue and surrounding tissues thereof; the information analysis module is configured to analyze and provide a morphological score and a functional score based on the acquired image information characteristic parameters; the calculation and output module is configured to calculate the morphological score and the functional score, separately; and the determination module is configured to determine the nature of a tumor according to the morphological score and the functional score.

For the above scoring system, preferably, the morphological scoring may adopt the following acquired image information as evaluation criteria: abundance of blood flow inside a tumor, volume of blood flow in the periphery of a tumor, spatial running of blood vessels in a tumor, infiltration of blood vessels into a tumor, oxygen saturation inside a tumor, and oxygen saturation in the periphery of a tumor. The functional scoring may use tumor oxygen saturation values that are quantitatively calculated through the acquired image information as evaluation criteria; the oxygen saturation values may include an oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of a tumor; and a condition that both the oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of the tumor are less than 0.75 to 0.80 may be used as an evaluation criterion for hypoxic malignant potential.

For the above scoring system, preferably, the abundance of blood flow inside a tumor may be scored as follows: O-none, 1-little, and 2-rich; the volume of blood flow in the periphery of a tumor may be scored as follows: O-none, 1-little, and 2-rich; the spatial running of blood vessels in a tumor may be scored as follows: O-regular and 1-irregular; and the infiltration of LU102578 blood vessels into a tumor may be scored as follows: 0-No and 1-Yes.

For the above scoring system, preferably, the image information may further include volume of tumor vessels and spatial distribution of tumor vessels, where a ratio of the volume of tumor vessels to a nodule volume may be scored as follows: O-none, 1-low, and 2-high; and the spatial distribution of tumor vessels may be scored as follows: O-uniform and 1-asymmetric.

In another aspect of the present invention, a 2D and 3D breast tumor scoring system and device based on photoacoustic/ultrasonic imaging technology is provided, which can use quantitative parameters to distinguish malignant and benign tumors.

In one aspect of the present invention, a breast tumor scoring system based on photoacoustic/ultrasonic imaging technology is provided, and the system includes an information acquisition module, an information analysis module, and an output module.

The information acquisition module is configured to perform 2D or 3D image information acquisition through a photoacoustic/ultrasonic imaging device to acquire image information characteristic parameters of a breast tumor tissue and surrounding tissues thereof; the information analysis module is configured to perform classification and operation processing on image information to acquire various characteristic parameters of an acquired image; and the output module is configured to determine according to the acquired characteristic parameters.

Preferably, the information analysis module may include a functional calculation module, and the functional calculation module may be configured to quantitatively calculate a characteristic parameter of tumor oxygen saturation values based on the acquired image information for determination.

Preferably, the oxygen saturation values may include an oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of a tumor; and a condition that both the oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of the tumor are less than 0.75 to 0.80 may be used as an evaluation criterion for hypoxic malignant potential.

Preferably, the information analysis module may include a morphological determination module, and the morphological determination module may be configured to perform calculation or software operation processing on the following characteristic parameters in the acquired image information according to specific criteria: abundance of blood flow inside a tumor, volume of blood flow in the periphery of a tumor, spatial running of blood vessels in a tumor, infiltration of blood vessels into a tumor, and the like.

Preferably, the abundance of blood flow inside a tumor may be scored as follows: O-none, LU102578 1-little, and 2-rich; the volume of blood flow in the periphery of a tumor may be scored as follows: O-none, 1-little, and 2-rich; the spatial running of blood vessels in a tumor may be scored as follows: O-regular and 1-irregular; and the infiltration of blood vessels into a tumor may be scored as follows: 0-No and 1-Yes.

Preferably, the morphological determination module may further involve the following acquired image information: volume of tumor vessels and spatial distribution of tumor vessels, where a ratio of the volume of tumor vessels to a nodule volume may be scored as follows: 0- none, 1-low, and 2-high; and the spatial distribution of tumor vessels may be scored as follows: O-uniform and 1-asymmetric.

For the above use, preferably, the morphological scoring may further involve the following acquired image information: volume of tumor vessels and spatial distribution of tumor vessels, where a ratio of the volume of tumor vessels to a nodule volume may be scored as follows: 0-

none, 1-low, and 2-high; and the spatial distribution of tumor vessels may be scored as follows: LU102578 O-uniform and 1-nonuniform. Specifically, if the ratio of the volume of tumor vessels to a nodule volume gets a score indicating low, the volume of vessels accounts for less than 50% of the nodule volume, and if the ratio of the volume of tumor vessels to a nodule volume gets a score indicating high, the volume of vessels accounts for more than 50% of the nodule volume; and if the spatial distribution of tumor vessels gets a score indicating uniform, blood vessels are evenly distributed at symmetrical parts of the nodule, with consistent numbers and diameters, and if the spatial distribution of tumor vessels gets a score indicating nonuniform, blood vessels are unevenly distributed at symmetrical parts of the nodule, with inconsistent numbers and diameters.

Further, the present invention also provides a device including the scoring system described above, including: an ultrasonic probe configured to acquire sound and image information; a host connected to a light emission and transmission module and an ultrasonic phased array transmitting and receiving module via two-core cables, which is configured to drive the emission of laser and ultrasonic signals and to receive a photoacoustic signal and a reflected ultrasonic signal for imaging; a processor configured to convert specific parameters of an imaging signal into specific values; and an output unit configured to output specific images and values.

Preferably, the ultrasonic probe may include a phased array probe, a convex array probe, and a linear array probe.

Preferably, the processor may be any commercially available charge-coupled device (CCD) capable of converting photoacoustic videos into digital signals.

Preferably, the output unit may be a printer.

The present invention has the following beneficial effects.

Compared with the previous 2D photoacoustic/ultrasonic dual-modality imaging, the 3D photoacoustic/ultrasonic dual-modality imaging of the present invention has the advantage of being able to use quantitative parameters to distinguish malignant and benign tumors. Moreover, 3D imaging can also provide more robust quantitative results than 2D imaging. It can be seen from FIG. 7 that the oxygen saturation (SO2) values of different 2D slices in the malignant tumor region varies greatly among slices. It is not accurate to use an oxygen saturation (SO2) value calculated for a single slice to represent the oxygen saturation of the entire tumor. Therefore, 3D tumor scanning can be conducted to provide more stable quantitative results. In addition, compared with single breast 3D photoacoustic imaging, in the case where ultrasonic imaging is conducted to describe a tumor region, the external and internal characteristics of the tumor can be analyzed separately to improve the diagnostic sensitivity and specificity for malignant tumors. In addition, compared with the scoring system and device used LU102578 in previous studies, the present invention distinguishes malignant and benign tumors based on a critical value of oxygen saturation (SO2), which is more convenient, repeatable, and objective in diagnosis.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 shows representations defining a tumor region and a tumor surrounding area in an example of the present invention; FIG. 2 shows the mean oxygen saturation of the tumor region and tumor surrounding area in benign, malignant and normal groups; FIG. 3 shows receiver operating characteristic (ROC) curves plotted by changing the SO2 thresholds in the tumor region (FIG. 3(a)) and the tumor surrounding area (FIG. 3(b)) to distinguish malignant tumors from benign tumors; FIG. 4 shows the PA/US fusion imaging results of a malignant tumor (IBC) and a benign tumor (fibroadenoma), where, more abundant and irregular blood vessels with low-SOz can be observed in the malignant tumor region and the tumor surrounding area (FIG. 4(a)), which is different from the vascular pattern of the benign tumor (FIG. 4(b)); FIG. 5 shows mammography and CD31 immunohistochemistry (IHC) blood vessel staining results, where, it can be seen from the mammography result in FIG. 5(a) that because there is no obvious calcification and boundaries, it is difficult to detect a malignant tumor; it can be seen from the mammography result in FIG. 5(b) that a benign tumor cannot be identified by mammography as easily as by ultrasound; it can be seen from the IHC blood vessel staining results that more CD31 blood vessel staining appears in the malignant tumor region and the tumor surrounding area (FIG. 5(c)) and a different result is acquired for the benign tumor (FIG. 5(d)), which is consistent with the 2D PA/US imaging results; FIG. 6 shows the 3D blood vessel images of the same tumor shown in FIG. 4(a, b), where, abundant blood vessels can be seen in the surrounding area of the malignant tumor, while relatively few blood vessels are seen in the surrounding area of the benign tumor; the SO» distributions in the tumor region and the tumor surrounding area are shown (FIG. 6(c, d)); and compared with the benign tumor, a significantly reduced SO» distribution can be seen in the malignant tumor; FIG. 7 shows the average oxygen saturation (SO2) values of different 2D slices in the malignant tumor region; FIG. 8 is a structural block diagram of the breast tumor scoring system of the present invention; and FIG. 9 is a structural diagram of the breast tumor scoring system of the present invention.

DETAILED DESCRIPTION OF THE EMBODIMENTS LU102578 The following examples are provided to illustrate the present invention, rather than to limit the scope of the present invention. Unless otherwise specified, the technical means used in the examples are conventional means well known to those skilled in the art. The present invention is further described below with reference to the accompanying drawings and examples.

In this example, "3D" refers to three-dimensional, "PA" refers to photoacoustic, "US" refers to ultrasonic, "IBC" refers to invasive breast cancer, "SO" refers to oxygen saturation, "Hb" refers to oxygenated hemoglobin, "deHb" refers to deoxygenated hemoglobin, and "IHC" refers to immunohistochemistry.

PA/US dual-modality 3D imaging system The dual-modality system in this study is based on a high-end clinical ultrasonic machine (Resona 7, Mindray Bio-Medical Electronics Co., Ltd.), which can perform PA imaging and acquire data required by PA imaging. The delay-and-sum algorithm is used to reconstruct a PA imaging result online. A clinical linear probe (L9-3U, Mindray Bio-Medical Electronics Co., Ltd.) has 192 elements, with a size of 0.2 mm for each element, and a center frequency of 5.8 MHz. A laser source is an OPO tunable laser (Spitlight 600-OPO, Innolas laser GmbH), which generates 700 nm to 850 nm laser pulses at 10 Hz. In the study of the present invention, 750 nm and 830 nm are used for PA functional imaging. The time-division multiplexing (TDM) method is adopted to realize PA/US real-time imaging with two wavelengths and SO2 mapping at a frame rate of 5 Hz. By scanning the breast skin surface using the probe, the system can perform local 3D dual-modality functional imaging. During 3D image acquisition, a motor moves at a steady speed (1 mm/s), and at the same time, a group of 2D US images and two- wavelength PA images are acquired at a step interval of 0.2 mm, with a total scan length of 4 cm and a total scan time of 200 s. 3D imaging results are downloaded for further data analysis.

In order to acquire a 3D PA/US image, a 2D SO» image is introduced into Amira (version

6.0, Visage Imaging) and a blood vessel image is acquired by extracting the surface of the SO2 image. Then the surface of the tumor region identified in B mode and the blood vessel image are imaged together with a certain degree of transparency.

Patients From November 2017 to January 2018, 46 patients with a breast tumor smaller than 2 cm were consecutively recruited from the breast surgery and inpatient departments, who were received BIRADS scoring and got a score of 3 to 5. All patients were initially subjected to ultrasonic diagnosis, mammography diagnosis, and/or MRI diagnosis by an experienced imaging doctor. Three imaging physicians performed routine ultrasonography on all patients, who had more than 10 years of diagnostic experience in ultrasonic diagnosis of breast diseases.

After the routine ultrasonography, 2D and 3D PA/US dual-modality imaging was performed. LU102578 All patients underwent resection and biopsy of the lesion to acquire pathological results.

Among the 46 patients, two patients failed to produce images due to system failure; and in two other patients, ultrasonic diagnosis showed that the distal end of a breast lesion was more than 3.5 cm away from the skin layer, which exceeded the effective imaging depth of the current system due to the strong light attenuation in the tissue. Among the remaining 42 patients, 18 patients had intraductal lesions or distant metastases, 16 patients had T1 stage IBC without distant metastases, and 8 patients had fibroadenoma or adenosis of breast. In this study, the imaging results of 16 patients with T1 stage IBC and 8 patients with benign lesions (6 with fibroadenoma and 2 with adenosis of breast) were studied. Because intraductal carcinoma in situ (DCIS) has different internal pathogenesis from IBC and this study is focused on early breast cancer detection, the present invention selected 16 T1 stage IBC cases without distant metastases and 8 fibroadenoma or adenosis of breast cases for later data analysis.

Example 1, Construction of a breast tumor scoring system

1. Data analysis and 3D image acquisition In this application, the main two optical absorbents in the breast tissue are endogenous oxygenated hemoglobin (Hb) and deoxygenated hemoglobin (deHb). An optical absorption coefficient of blood was calculated according to the following equation: Ha(d,r)=Cur(r)eno(A)+CdeHr(T)EdeHb(A) (1) where, ja(À,r) represents the optical absorption coefficient of blood, €x»(A) represents the molar extinction of endogenous oxygenated hemoglobin (Hb), Cus(r) represents the concentration of endogenous oxygenated hemoglobin (Hb), &£dens(Ad) represents the molar extinction of deoxygenated hemoglobin (deHb), Caenn(r) represents the concentration of deoxygenated hemoglobin (deHb).

The PA signal is proportional to the product of the light absorption coefficient pa(A,r) and the luminous flux ®(A,r). The luminous flux depends on the wavelength (A) and the spatial position (r). Since the absorption coefficient ua (À) of background breast tissue at 750 nm and 830 nm is very close to the reduced scattering coefficient us’, in the research of the present invention, the luminous fluxes are roughly the same after the laser irradiation power at each wavelength is normalized. Then, the SO» at each pixel can be calculated with the following formula.

2) The above formula is acquired by the following method: if wavelengths used are À; = 750 nm and À = 830 nm, and the luminous flux is ©, then 0102578 PA(A1,0)=0(A1,0)* ualdi,r)= Cup(r)enp(A1)+Caenn(r)edenn(A1) (3) PA(A2,1)=0(A2,1)* pa(A2,7)= Cup(r)enp(A2)+Caenn(r)Edenn(A2) (4) If the difference between @(A1,r) and @(A2,r) can be ignored because it is small under the same laser energy, then (2) can be acquired through simultaneous equations (3) and (4).

SO2(r)= CH. CHe(r)+ CaeHb(r))(PA(1,1)* EdeHb(A2)- PA(a2,1)* EdeHb(A1))(PA(1,1)*( Edenb(A2)- EHb(A2))+ PA(A2,1)*( EHb(A1)- EdeHb(A1)) where, PA(M,r)* is the PA that ignores @(A1,r); PA(Xa.r)* is the PA that ignores @(A2,1); and a PA value can be directly acquired by an ultrasonic probe. Any pixels with negative SO2 values were removed in the subsequent analysis.

For 3D-PA/US quantitative calculation, the tumor boundary of each ultrasonic slice was first marked by an experienced doctor. Then, the least volume ellipse (LVE region) surrounding the 3D tumor region (tumor region) was calculated. By increasing each of the three axis lengths of the LVE by 1.2 times, an extension ellipse (extension ellipse region) was acquired. A region within the extension ellipse other than the tumor region was defined as the tumor surrounding area, as shown in FIG. 1.

After the tumor region and the tumor surrounding areas thereof were marked, the average oxygen saturation value inside the tumor and the average oxygen saturation value in the periphery of the tumor in the two regions were calculated, and 40% was set as a threshold to reduce the influence of artifacts. Similarly, the vascular densities (vas den) in the tumor region and the tumor surrounding area thereof were also calculated by dividing the calculated number of pixels with SO2 > 40% by the total number of pixels in a corresponding region.

2. Statistical analysis The non-parametric two-tailed Mann-Whitney U-test was used to calculate the statistical significance between two groups. Bonferroni correction for multiple comparisons (number of trials: n = 3) was conducted, and for the tumor region and the tumor surrounding area, P = 0.017 indicates 95% statistical significance. The Hodges-Lehmann estimator was used to give the difference between two groups and 95% statistical significance. The Matlab (Mathworks, Inc.) was used for statistical analysis.

3. Results The results of 24 patients (16 T1 stage IBC cases and 8 benign lesion cases) were included in the statistical analysis. These cases were divided into three groups: benign group (fibroadenoma or adenosis of breast, n = 8), malignant group (T1 stage IBC, n = 16), and normal

. . LU102578 group (contralateral healthy breast, n = 22). Among the 24 patients, 2 patients were excluded from the normal group due to the presence of a lesion in the contralateral breast. The average SO: values of the tumor region and the tumor surrounding area in the malignant group and the benign group are shown in Table 1 below: Table 1 SO» in the periphery of a tumor

0.624 0.656

0.662 0.663

0.664 0.688

0.678 0.697

0.685 0.698 | 6 | 0.692 0.717

0.695 0.722 Malignant tumor L 8 | 0.697 0.727 ghant tumor mg 0.711 0.731

0.712 0.747

0.718 0.750

0.740 0.750

0.748 0.754

0.748 0.757

0.761 0.772

0.768 0.772

0.683 0.667

0.809 0.743

0.825 0.777 Benign tumor 0.745 0.781 8 0.802 0.785 LL 6 | 0.842 0.790

0.794 0.803 | 8 | 0.711 0.806 As shown in FIG. 2(a), in terms of the tumor region, the average SO» value of the malignant group (Malignant) was 7.7% lower than that of the benign group (Benign) (95% confidence interval: 2.1%, 12.4%) (P = 0.016) and was 3.9% lower than that of the normal group (Normal) (95% confidence interval: 2.2%, 5.5% (P = 0.010)). In terms of the tumor surrounding area (FIG. 2(b)), the average SO2 value of the malignant group was 4.9% lower than that of the benign group (95% confidence interval: 1.6%, 8.4%) (P = 0.009). There was no significant difference between the average SO2 values of the benign group and the normal group at the 95% level. ROC curves were plotted by changing the SO» thresholds in the tumor region (FIG. 3(a)) and the tumor surrounding area (FIG. 3(b)) to distinguish malignant tumors from benign tumors. In this example, the SO» threshold of the tumor region was set to 0.769 to 0.794, a sensitivity for diagnosing malignant tumors was 100%, a specificity was 62.5%, and an area LU102578 under the ROC curve was 0.81. In this example, the SO» threshold of the tumor surrounding area was set to 0.776 to 0.781, a sensitivity for diagnosing malignant tumors was 100%, a specificity was 75%, and an area under the ROC curve (AUC) was 0.84. Based on experience, the inventors set the condition that both the oxygen saturation (SO2) value inside a tumor and an oxygen saturation (SO2) value in the periphery of the tumor are less than 0.75 to 0.80 as an evaluation criterion for hypoxic malignant potential, and the above value was acquired by comprehensively considering ROC curves of benign and malignant tumors, which ensured the optimal sensitivity and specificity. To better determine benign and malignant tumors, the ROC curve threshold range was adjusted according to actual needs, for example, the SOz range can be extended to improve the diagnostic specificity of malignant tumors, and the SO2 range can be narrowed to improve the diagnostic sensitivity of malignant tumors.

FIG. 4(a) shows the PA/US fusion imaging results of a malignant tumor (IBC), and FIG. 4(b) shows the PA/US fusion imaging results of a benign tumor (fibroadenoma). Abundant and irregularly-shaped blood vessels with low SO» could be observed inside and in the periphery of the malignant tumor (corresponding to the malignant sample 11 in Table 1, the SO; inside the tumor was 0.72 and the SO» in the periphery of the tumor was 0.75, which were lower than the set threshold) (FIG. 4(a)), which was different from the vascular pattern of a benign lesion (FIG. 4(b)) (corresponding to the benign sample 6 in Table 1, the SO: inside the tumor was 0.84 and the SO» in the periphery of the tumor was 0.79, which were higher than the set threshold). As shown in FIG. 4(a), there were multiple blood vessels (> 3, or forming vascular network) inside the tumor, the abundance of blood flow got a score of 2-rich; multiple blood vessels appeared in multiple regions in the periphery of the tumor, and the volume of blood flow got a score of 2-rich; the spatial running of blood vessels in the tumor got a score of 1-irregular; and the infiltration of blood vessels into a tumor got a score of 1-Yes.

For malignant cases, it is difficult to accurately grade malignant tumors based on the results of conventional ultrasound, because malignant tumors may have a regular shape similar to the benign tumors. In addition, results of mammography and CD31 IHC blood vessel staining are shown in FIG. 5. Results of mammography are shown in FIG. 5(a). With no obvious calcification and clear boundaries, malignant tumors are difficult to detect by mammography. The mammography results in FIG. 5(b) show that benign tumors cannot be identified by mammography as easily as by conventional ultrasound. Results of IHC blood vessel staining showed that more CD31 stained blood vessels appeared inside and in the periphery of the malignant tumor (FIG. 5(c)), but there was no such situation inside and in the periphery of the benign tumor (FIG. 5(d)), which was consistent with the 2D PA/US imaging result.

Example 2 Scoring Application Example LU102578 In May 2018, one patient with a breast tumor smaller than 2 cm was recruited from the breast surgery department, who was received BIRADS scoring and got a score of 4. The blood vessel images of the patient are shown in FIG. 6(a, b).

In FIG. 6(a), the ratio of blood vessels volume inside the tumor to the nodule volume was higher than 50%, indicating rich blood flow, with a score of 2; the distribution range of blood vessels in the periphery of the tumor was greater than 50% of the nodule volume, indicating high volume of blood flow, with a score of 2; the spatial running of blood vessels in the tumor was distorted with irregular dilation, and anastomotic branches were disordered, indicating irregular spatial running, with a score of 1; and the blood vessels infiltrated into tumor, with a score of 1. FIG. 6(a) got a total score of 6. According to the morphological score, it was determined to have malignant potential. In FIG. 6(a), the SO: inside the tumor was 0.72 and the SO in the periphery of the tumor was 0.75, which were lower than the set threshold. According to the functional score, it was determined to have malignant potential. In summary, the tumor in FIG. 6(a) was determined to have malignant potential.

In FIG. 6(b), the ratio of blood vessels volume inside the tumor to the nodule volume was lower than 50%, indicating little blood flow, with a score of 1; the distribution range of blood vessels in the periphery of the tumor was smaller than 50% of the nodule volume, indicating low volume of blood flow, with a score of 1; the spatial running of blood vessels in the tumor showed no distortion, irregular dilation, and disordered anastomotic branches, indicating regular spatial running, with a score of 0; and the tumor was not infiltrated with blood vessels, with a score of 0. FIG. 6(b) got a total score of 2. According to the morphological score, it was determined to have benign potential. In FIG. 6(b), the SO» inside the tumor was 0.84 and the SO: in the periphery of the tumor was 0.79, which were higher than the set threshold. According to the functional score, it was determined to have benign potential. In summary, the tumor in FIG. 6(b) was determined to have benign potential.

Through pathological diagnosis and clinical diagnosis, the patient was determined to indeed have a benign tumor.

Example 3, A tumor scoring system and device based on photoacoustic/ultrasonic imaging technology As shown in FIG. 8, a tumor scoring system based on photoacoustic/ultrasonic imaging technology is provided, and the system includes an information acquisition module, an information analysis module, and an output module.

The information acquisition module is configured to perform 2D or 3D image information acquisition through a photoacoustic/ultrasonic imaging device to acquire image information LU102578 characteristic parameters of a tumor tissue and surrounding tissues thereof; the information analysis module is configured to perform classification and operation processing on image information to acquire various characteristic parameters of an acquired image; and the output module is configured to determine according to the acquired characteristic parameters.

The information analysis module includes a functional calculation module and a morphological determination module. The functional calculation module is configured to quantitatively calculate a characteristic parameter of tumor oxygen saturation values based on the acquired image information for determination. The oxygen saturation values include an oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of a tumor, and a condition that both the oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of the tumor are less than 0.75 to 0.80 is used as an evaluation criterion for hypoxic malignant potential.

The morphological determination module is configured to perform calculation or software operation processing on the following characteristic parameters in the acquired image information according to specific criteria: abundance of blood flow inside a tumor, volume of blood flow in the periphery of a tumor, spatial running of blood vessels in a tumor, and the infiltration of blood vessels into a tumor.

Preferably, the abundance of blood flow inside a tumor may be scored as follows: O-none, 1-little, and 2-rich; the volume of blood flow in the periphery of a tumor may be scored as follows: O-none, 1-little, and 2-rich; the spatial running of blood vessels in a tumor may be scored as follows: O-regular and 1-irregular; and the infiltration of blood vessels into a tumor may be scored as follows: 0-No and 1-Yes. The morphological determination module may further involve the following acquired image information: volume of tumor vessels and spatial distribution of tumor vessels, where a ratio of the tumor vessels volume to a nodule volume may be scored as follows: O-none, 1-low, and 2-high; and the spatial distribution of tumor vessels may be scored as follows: O-uniform and 1-asymmetric.

As shown in FIG. 9, the present invention also provides a device including the scoring system described above, including: an ultrasonic probe 1 configured to acquire sound and image information; a host 2 connected to a light emission and transmission module and an ultrasonic phased array transmitting and receiving module via two-core cables, which is configured to drive the emission of laser and ultrasonic signals and to receive a photoacoustic signal and a reflected ultrasonic signal for imaging; a processor 3 configured to convert specific parameters of an imaging signal into specific values; and an output unit 4 configured to output specific LU102578 images and values.

Preferably, the ultrasonic probe 1 may include a phased array probe, a convex array probe, and a linear array probe.

Preferably, the processor 3 may be any commercially available CCD capable of converting photoacoustic videos into digital signals.

Preferably, the output unit 4 may be a printer.

The present invention has been described in detail above with reference to general descriptions and specific examples, but it will be apparent to those skilled in the art that some modifications or improvements can be made based on the present invention. Therefore, all these modifications or improvements made without departing from the spirit of the present invention fall within the scope of the present invention.

Industrial Applicability The 3D photoacoustic imaging of the present invention can be used industrially in a breast tumor scoring system and device, and thus has industrial applicability.

Claims

CLAIMS LU102578

1. Application of three-dimensional (3D) photoacoustic imaging in a breast tumor scoring system, comprising the following steps: (1) acquiring image information of a breast tumor in vitro through photoacoustic/ultrasound dual-modality imaging; (2) analyzing acquired image information, and conducting morphological and functional scoring, separately; (3) obtaining a comprehensive score, based on the results of the morphological and functional scoring, to determine whether the breast tumor has the malignant potential; for example, if one or all of the parameters of either morphological scoring or functional scoring are regarded as the malignant potential, the breast tumor is considered to be a malignant tumor.

2. The application according to claim 1, wherein, the functional scoring uses tumor oxygen saturation values that are quantitatively calculated through the acquired image information as evaluation criteria; the oxygen saturation values comprise an oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of a tumor; and a condition that both the oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of the tumor are less than 0.75 to 0.80 is used as an evaluation criterion for hypoxic malignant potential.

3. The application according to claim 2, wherein, the oxygen saturation value inside a tumor and the oxygen saturation (SO2) value in the periphery of a tumor are calculated by the following formula: SO2(r)= CH. CHe(r)+ CaeHb(r))(PA(1,1)* EdeHb(A2)- PA(a2,1)* EdeHb(A1))(PA(1,1)*( Edenb(A2)- EHb(A2))+ PA(A2,1)*( EHb(A1)- EdeHb(A1)) wherein, Hb represents endogenous oxygenated hemoglobin and deHb represents deoxygenated hemoglobin, PA(A1r) *= pa(di,r)= Cho(r)Enb(41)+CdeHb(T)EdeHb(A1) PAÇ(o,r) *= pua(do,r)= Cup(r)enp(A2)+Caern(r)den(A2) AM=750 nm, A= 830 nm.

4. The application according to claim 1, wherein, an evaluation criterion for the functional scoring further comprises vascular density.

5. The application according to claim 4, wherein, the vascular density is acquired by dividing the calculated number of pixels with SO2 > 40% by the total number of pixels in a corresponding region.

6. The application according to claim 1, wherein, the breast tumor comprises invasive breast cancer (IBC) and breast intraductal tumors. 0102578

7. The application according to claim 6, wherein, the breast tumor is T1 stage IBC.

8. The application according to claim 1, wherein, the morphological scoring adopts the following acquired image information as evaluation criteria: abundance of blood flow inside a tumor, volume of blood flow in the periphery of a tumor, spatial running of blood vessels in a tumor, and infiltration of blood vessels into a tumor.

9. The application according to claim 8, wherein, the abundance of blood flow inside a tumor is scored as follows: O-none, 1-little, and 2-rich; the volume of blood flow in the periphery of a tumor is scored as follows: O-none, 1-little, and 2-rich; the spatial running of blood vessels in a tumor is scored as follows: O-regular and 1-irregular; and the infiltration of blood vessels into a tumor is scored as follows: 0-No and 1-Yes.

10. The application according to claim 9, wherein, the morphological scoring further involves the following acquired image information: volume of tumor vessels and spatial distribution of tumor vessels, wherein, a ratio of the tumor vessels volume to a nodule volume is scored as follows: O-none, 1-low, and 2-high; and the spatial distribution of tumor vessels is scored as follows: O-uniform and 1-asymmetric.

11. A breast tumor scoring system based on photoacoustic/ultrasonic imaging technology, wherein, the system comprises an information acquisition module, an information analysis module, and an output module; the information acquisition module is configured to perform two-dimensional (2D) or 3D image information acquisition through a photoacoustic/ultrasonic imaging device to acquire image information characteristic parameters of a breast tumor tissue and surrounding tissues thereof; the information analysis module is configured to perform classification and operation processing on image information to acquire various characteristic parameters of an acquired image; and the output module is configured to determine according to the acquired characteristic parameters.

12. The scoring system according to claim 11, wherein, the information analysis module comprises a functional calculation module, and the functional calculation module is configured to quantitatively calculate a characteristic parameter of breast tumor oxygen saturation values based on the acquired image information for determination.

13. The scoring system according to claim 12, wherein, the oxygen saturation values comprise an oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of a tumor; and a condition that both the oxygen saturation value inside a tumor and an oxygen saturation value in the periphery of the tumor are less than 0.75 to 0.80 is used as an LU102578 evaluation criterion for hypoxic malignant potential.

14. The scoring system according to claim 11, wherein, the information analysis module further comprises a morphological determination module, and the morphological determination module is configured to perform calculation or software operation processing on the following characteristic parameters in the acquired image information according to specific criteria: abundance of blood flow inside a tumor, volume of blood flow in the periphery of a tumor, spatial running of blood vessels in a tumor, and the infiltration of blood vessels into a tumor.

15. The scoring system according to claim 14, wherein, the abundance of blood flow inside a tumor is scored as follows: O-none, 1-little, and 2-rich; the volume of blood flow in the periphery of a tumor is scored as follows: O-none, 1-little, and 2-rich; the spatial running of blood vessels in a tumor is scored as follows: O-regular and 1-irregular; and the infiltration of blood vessels into a tumor is scored as follows: 0-No and 1-Yes.

16. The scoring system according to claim 14 or 15, wherein, the morphological determination module further involves the following acquired image information: volume of tumor vessels and spatial distribution of tumor vessels, wherein, a ratio of the tumor vessels volume to a nodule volume is scored as follows: O-none, 1-low, and 2-high; and the spatial distribution of tumor vessels is scored as follows: O-uniform and 1-asymmetric.

17. A device comprising the scoring system according to any one of claims 11 to 16, comprising: an ultrasonic probe configured to acquire sound and image information; a host connected to a light emission and transmission module and an ultrasonic phased array transmitting and receiving module via two-core cables, which is configured to drive the emission of laser and ultrasonic signals and to receive a photoacoustic signal and a reflected ultrasonic signal for imaging; a processor configured to convert specific parameters of an imaging signal into specific values; and an output unit configured to output specific images and values.

18. The device according to claim 17, wherein, the ultrasonic probe comprises a phased array probe, a convex array probe, and a linear array probe.

19. The device according to claim 17, wherein, the processor is any commercially available charge-coupled device (CCD) capable of converting photoacoustic videos into digital signals.

20. The device according to claim 17, wherein, the output unit is a printer.