Method for calculating residual renal function of renal images based on fine renal anatomy

Abstract

The present invention discloses a method for calculating the renal function of the remaining kidney based on the fine anatomy of the kidney in renal imaging, including: S1. Automatically segmenting the kidney and surrounding tissues through the training of a large data model of kidney CT images; S2. Defining renal tumors, renal cortex, renal medulla, collecting system, renal artery, and renal vein based on the multi-phase dynamic changes of kidney CT images; S3. Integrating the segmentation information of each layer to construct a 3D map, and respectively calculating the tumor volume, renal cortex volume, and renal medulla volume in the diseased kidney and healthy kidney. The present invention obtains the dynamic information of the plain scan phase, arterial phase, and venous phase images of renal tumors, renal cortex, renal medulla, collecting system, renal artery, and renal vein based on kidney CT images, accurately segments each different tissue region of the kidney and calculates the volume, and designs an operation formula according to the physiological functions of each region of the kidney to more accurately estimate the glomerular filtration rate of the remaining kidney.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and specifically to a method for calculating the residual renal function of the kidney based on the fine anatomy of the kidney using renal imaging. Background Art

[0002] The incidence of renal cell carcinoma accounts for 2%-3% of adult malignancies. The widespread use of non-invasive radiological techniques, especially the application of ultrasound, has increased the diagnosis of renal incidentalomas, with smaller tumors and earlier stages. Now, major guidelines recommend that T1a-stage renal cell carcinoma is an indication for nephron-sparing surgery. Experienced units can perform nephron-sparing surgery for selected T1b-stage renal cell carcinomas. For renal benign tumors, nephron-sparing surgery is recommended regardless of size. The amount of remaining renal function is closely related to postoperative renal insufficiency and treatment effects, and is also one of the important indicators for evaluating the efficacy of nephron-sparing surgery.

[0003] Currently, the commonly used evaluation method is to use radionuclide renography to evaluate split renal function. This evaluation method is not only expensive, but also requires the injection of radioactive nuclides and the detection of the radionuclide concentration in both kidneys at multiple time points, with the disadvantages of complex examination and radioactive radiation.

[0004] The Chinese invention patent with the publication number CN112700875B discloses a system and a computer-readable storage medium for predicting renal function after kidney tumor surgery, which estimates the loss of renal volume and glomerular filtration rate based on preoperative imaging examinations, but ignores the fine anatomy of the kidney. The volumes of the renal cortex and the renal medulla are both the volume of the renal parenchyma, and their volumes can be equivalent, but their functions are not equivalent. In fact, they together can form the basic renal unit. The renal tumor treatment guidelines no longer have clear regulations on the resection margin of renal tumors, only requiring complete resection of the tumor. Due to the different sizes of renal tumors and different surgeons performing the main surgery, the methods and margins of removing renal tumors are different. For benign tumors, it is not necessary to remove normal renal parenchyma. Therefore, estimating the loss of renal volume by 1 cm around the renal tumor is inaccurate.

[0005] After partial nephrectomy, the renal parenchyma needs to be sutured, and there is also a loss of renal function in the sutured part of the renal parenchyma, which is not correspondingly reflected in the above patent. Summary of the Invention

[0006] Aiming at the above-mentioned existing technical deficiencies, the purpose of the present invention is to provide a method for calculating the residual renal function of the kidney based on the fine anatomy of the kidney using renal imaging, so as to solve the problem of inaccurate calculation of the existing residual renal volume.

[0007] To solve the above technical problems, the present invention adopts the following technical solutions:

[0008] The present invention provides a method for calculating the residual renal function of the kidney based on the fine anatomy of the kidney using renal imaging, including:

[0009] S1. Automatically segment the kidney and surrounding tissues through training with a large - data model of kidney CT images;

[0010] S2. Define renal tumors, renal cortex, renal medulla, collecting system, renal artery, and renal vein based on the multi - phase dynamic changes of kidney CT images;

[0011] S3. Synthesize the segmentation information of each layer, construct a 3D map, and calculate the tumor volume, renal cortex volume, and renal medulla volume in the diseased kidney and healthy kidney respectively;

[0012] S4. The renal cortex volume A of each individual corresponds to a corresponding proportion of the renal medulla volume B. Assume the proportionality coefficient is:

[0013] A / B = N;

[0014] Let the diseased renal cortex volume be represented by A_diseased, the diseased renal medulla volume be represented by B_diseased, the healthy renal cortex volume be represented by A_healthy, and the healthy renal medulla volume be represented by B_healthy. Then:

[0015] A = A 患 + A 健 ;

[0016] B = B 患 + B 健 ;

[0017] The total eGFR is obtained from preoperative renal artery imaging or through the renal filtration rate calculation formula:

[0018] MDRD (simplified) I: eGFR = 186×Scr -1.154 ×age -0.203 ×(0.742 if female)

[0019] The total eGFR = diseased eGFR + healthy eGFR = coefficient C 皮 *A = coefficient C 髓 *B;

[0020] S5. Calculate the remaining renal cortex volume A_residual and remaining renal medulla volume B_residual of the diseased kidney respectively. Resect a certain volume of the diseased renal cortex △a and diseased renal medulla volume △b. The proportionality coefficient is:

[0021] △a / △b = n;

[0022] S6. Calculate △a = A 患 - A 残 , △b = B 患 - B 残 , and compare the magnitudes of N and n;

[0023] If N < n, it means that the renal cortex has more loss than the renal medulla. According to the principle of the smallest renal unit composed of the corresponding proportions of the renal cortex and the renal medulla, taking the renal cortex with more loss as the calculation standard, then:

[0024] The diseased eGFR = coefficient C 皮 *(A 残 + A 健 ) - the healthy eGFR;

[0025] If n < N, it means that the renal medulla has more loss than the renal cortex. According to the principle of the smallest renal unit composed of the corresponding proportions of the renal cortex and the renal medulla, taking the renal medulla with more loss as the calculation standard, then:

[0026] The diseased eGFR = coefficient C 髓 *(B 残 + B 健 ) - the healthy eGFR;

[0027] If n = N, it means that the renal medulla and the renal cortex have the same amount of loss. According to the principle of the smallest renal unit composed of the corresponding proportions of the renal cortex and the renal medulla, when the renal cortex and the renal medulla have the same amount of loss, the results calculated based on the renal cortex and the renal medulla should be equal. Then:

[0028] The diseased eGFR = coefficient C 髓 *(B 残 + B 健 ) - the healthy eGFR = coefficient C 皮 *(A 残 + A 健 ) - the healthy eGFR.

[0029] The beneficial effects of the present invention are as follows:

[0030] 1. Based on the kidney CT images, dynamic information of the plain scan phase, arterial phase, and venous phase images of renal tumors, renal cortex, renal medulla, collecting system, renal artery, and renal vein is obtained, and each different tissue region of the kidney is accurately segmented and the volume is calculated. According to the physiological functions of each region of the kidney, an operation formula is designed to more accurately estimate the glomerular filtration rate of the remaining kidney;

[0031] 2. After each patient's operation, kidney CT needs to be followed up. According to the postoperative kidney CT images, the losses of the renal cortex and the renal medulla are accurately calculated, and the lost glomerular filtration rate is accurately calculated based on this.

[0032] 3. The renal parenchyma and renal medulla sutured by the suture may have ischemic atrophy and lose renal function. This part of the change is excluded because it does not conform to the renal cortex and renal medulla defined by the multi-phase changes of the kidney CT, making the estimation of postoperative renal function loss more accurate. Description of the Drawings

[0033] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0034] Figure 1 It is a schematic diagram of building a deep learning segmentation model for the kidney and renal cancer in S1 of the present invention;

[0035] Figure 2 It is to define and segment renal tumors, renal cortex, renal medulla, and renal arteries according to the multi-phase dynamic changes of renal CT in S2 of the present invention;

[0036] Figure 3 It is a schematic diagram of the steps of S3 of the present invention;

[0037] Figure 4 It is a calculation roadmap for S4 - S6 of the present invention.

[0038] Explanation of reference numerals: 1, renal tumor; 2, renal cortex; 3, renal artery; 4, renal medulla. Detailed implementation manners

[0039] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0040] Example 1, as Figures 1 to 4 shown, a method for calculating the residual renal function of renal images based on fine renal anatomy includes:

[0041] S1. Automatically segment the kidney and surrounding tissues through training with a large - data model of renal CT images;

[0042] S2. Based on the multi - phase dynamic changes of renal CT images, define renal tumors, renal cortex, renal medulla, collecting system, renal arteries, renal veins. The minimum renal unit structure is composed of renal corpuscles and renal tubules. The renal corpuscles are located in the renal cortex, and the renal tubules are located in the renal medulla;

[0043] S3. Integrate the segmentation information of each layer, construct a 3D map, and calculate the tumor volume, renal cortex volume, and renal medulla volume in the diseased kidney and healthy kidney respectively;

[0044] S4. The renal cortex volume A of each individual corresponds to a corresponding proportion of the renal medulla volume B. Assume the proportionality coefficient is:

[0045] A / B = N;

[0046] The volume of the affected renal cortex is represented by A_affected, the volume of the affected renal medulla is represented by B_affected, the volume of the healthy renal cortex is represented by A_healthy, and the volume of the healthy renal medulla is represented by B_healthy. Then there is:

[0047] A = A 患 + A 健 ;

[0048] B = B 患 + B 健 ;

[0049] The total eGFR is obtained from preoperative renal artery imaging or through the renal filtration rate calculation formula:

[0050] MDRD(simplified)I: eGFR = 186 × Scr -1.154 × age -0.203 × (0.742 if female)

[0051] The calculation formula MDRD(simplified)I adopted in the present invention was published in Ann Intern Med. 2006. PMID: 16908915. This is an equation for predicting renal function obtained by statistics in nephrology

[0052] Total eGFR = Affected eGFR + Healthy eGFR = Coefficient C 皮 * A = Coefficient C 髓 * B;

[0053] S5. Calculate the remaining cortex volume A_residual and the remaining medulla volume B_residual of the affected kidney respectively. Because the positions and sizes of tumors vary, and different surgeons perform the operation, after partial nephrectomy for renal tumors, a certain volume of the affected renal cortex △a and the affected renal medulla volume △b are removed. The proportionality coefficient is:

[0054] △a / △b = n;

[0055] S6. Calculate △a = A 患 - A 残 , △b = B 患 - B 残 , and compare the magnitudes of N and n;

[0056] If N < n, it means that more renal cortex is lost than renal medulla. According to the principle of the smallest renal unit composed of the corresponding proportions of the renal cortex and renal medulla, taking the renal cortex with more loss as the calculation standard, then:

[0057] Affected eGFR = Coefficient C 皮 *(A 残 + A 健 ) - Healthy eGFR;

[0058] If n < N, it means that the renal medulla is lost more than the renal cortex. According to the principle of the smallest renal unit composed of the corresponding proportions of the renal cortex and the renal medulla, taking the renal medulla with more loss as the calculation standard, then:

[0059] Patient eGFR = coefficient C 髓 *(B 残 +B 健 ) - healthy eGFR;

[0060] If n = N, it means that the renal medulla and the renal cortex are lost equally. According to the principle of the smallest renal unit composed of the corresponding proportions of the renal cortex and the renal medulla, when the renal cortex and the renal medulla are lost equally, the results calculated based on the renal cortex and the renal medulla should be equal. Then:

[0061] Patient eGFR = coefficient C 髓 *(B 残 +B 健 ) - healthy eGFR = coefficient C 皮 *(A 残 +A 健 ) - healthy eGFR.

[0062] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and modifications.

Claims

1. A method for calculating the residual renal function of renal images based on the fine anatomy of the kidney, characterized in that, Including: S1. Automatically segment the kidney and surrounding tissues through the training of the kidney CT image big data model; S2. Define renal tumors, renal cortex, renal medulla, collecting system, renal artery, and renal vein based on the multi-phase dynamic changes of kidney CT images; S3. Integrate the segmentation information of each layer, construct a 3D map, and calculate the tumor volume, renal cortex volume, and renal medulla volume in the diseased kidney and healthy kidney respectively; S4. The renal cortex volume A of each individual corresponds to the renal medulla volume B of the corresponding proportion. Assume the proportionality coefficient is: A / B = N; Let the diseased renal cortex volume be represented by A_diseased, the diseased renal medulla volume be represented by B_diseased, the healthy renal cortex volume be represented by A_healthy, and the healthy renal medulla volume be represented by B_healthy. Then there is: A = A 患 + A 健 ; B = B 患 + B 健 ; The total eGFR is obtained from preoperative renal artery imaging or through the renal filtration rate calculation formula: MDRD (simplified) I: eGFR = 186 × Scr -1.154 × age -0.203 × (0.742 if female) Total eGFR = Affected eGFR + Healthy eGFR = Coefficient C 皮 *A = Coefficient C 髓 *B; S5. Calculate the remaining renal cortex volume A_residual and the remaining renal medulla volume B_residual of the diseased kidney respectively. Resect a certain volume of the diseased renal cortex △a and the diseased renal medulla volume △b. The proportionality coefficient is: △a / △b = n; S6. Calculate △a = A 患 - A 残 , △b = B 患 - B 残 , compare the magnitudes of N and n; If N < n, it means that more renal cortex is lost than renal medulla. According to the principle of the smallest renal unit composed of the corresponding proportions of the renal cortex and renal medulla, taking the renal cortex with more loss as the calculation standard, then: Have eGFR = coefficient C 皮 *(A 残 +A 健 ) - healthy eGFR; If n < N, it means that more renal medulla is lost than renal cortex. According to the principle of the smallest renal unit composed of the corresponding proportions of the renal cortex and renal medulla, taking the renal medulla with more loss as the calculation standard, then: Subject to eGFR = coefficient C 髓 *(B 残 + B 健 ) - healthy eGFR; If n = N, it means that the renal medulla and renal cortex are lost equally. According to the principle of the smallest renal unit composed of the corresponding proportions of the renal cortex and renal medulla, the renal cortex and renal medulla are lost equally, and the results calculated based on the renal cortex and renal medulla should be equal. Then: Subject eGFR = coefficient C 髓 *(B 残 +B 健 ) - Healthy eGFR = coefficient C 皮 *(A 残 +A 健 ) - Healthy eGFR.

Images