Artificial neural network–based method for evaluating renal tumor and device for evaluating renal tumor using same

Abstract

The present invention provides a renal tumor evaluation method and an evaluation device and system using same. The method, which is implemented by a processor, comprises the steps of: receiving a medical image including a kidney and a renal tumor; generating three-dimensional image data from the medical image; segmenting, in the received three-dimensional image data, regions corresponding to the tumor, the renal parenchyma, the renal pelvis, and the renal hilum by using a prediction model trained to segment the regions of the tumor, the renal parenchyma, the renal pelvis, and the renal hilum using three-dimensional image data as input; determining, on the basis of the segmented regions, a first boundary surface between the tumor and the renal parenchyma or a second boundary surface between the tumor and the renal hilum; determining a first distance between the first boundary surface and the renal hilum or a second distance between the second boundary surface and the renal hilum; and performing renal tumor evaluation on the basis of the first distance or the second distance.

Description

An artificial neural network-based evaluation method for renal tumors and a device for evaluating renal tumors using the same

[0001] The present invention relates to an evaluation method for renal tumors based on an artificial neural network and a device for evaluating renal tumors using the same.

[0002] Renal cell carcinoma (RCC) is a malignant tumor that originates in the kidney. Recent advancements in imaging technology have led to an increase in early detection cases. Treatment options for renal cell carcinoma include radical nephrectomy and partial nephrectomy (PN). Partial nephrectomy is preferred, particularly for small renal cell carcinomas, due to the growing importance of preserving renal function. However, the surgical difficulty of partial nephrectomy varies significantly depending on the anatomical location and structural complexity of the tumor, which can affect the incidence of postoperative complications and renal function decline.

[0003] At this time, it may be very important to more precisely evaluate the impact of the location and structural characteristics of renal cancer on surgical difficulty and long-term preservation of renal function.

[0004] In other words, there is a continuous need for the development of a more quantitative and objective evaluation system for renal tumors.

[0005] The background technology of the invention has been prepared to facilitate a better understanding of the present invention. It should not be construed as an admission that the matters described in the background technology of the invention constitute prior art.

[0006] Various assessment systems have been proposed to evaluate renal tumors and predict surgical difficulty.

[0007] The proposed conventional evaluation systems for renal tumors include PADUA, RENAL, C-index, and CSA (Contact Surface Area), and these systems are used as major factors in evaluating renal tumors by considering the size, location, and relationship of the tumor to the renal sinus and renal hilum.

[0008] Meanwhile, although these evaluation systems help to reflect the anatomical characteristics of renal tumors, most of them are based on two-dimensional image information, which has the limitation of not sufficiently reflecting the three-dimensional structural complexity of renal tumors.

[0009] Due to this, the previously proposed assessment systems may be limited in the precise evaluation of renal tumors and may be dependent on the subjective experience of the medical practitioner.

[0010] To solve this problem, the inventors of the present invention focused on an evaluation index that can automatically segment renal tumor-related structures from medical images using a learned prediction model based on an artificial neural network and quantify three-dimensional anatomical information based on the segmented structures.

[0011] In particular, the inventors of the present invention recognized that existing evaluation methods only considered the simple size or location of the tumor, and sought to develop a quantitative evaluation system that reflects the three-dimensional spatial relationship between the tumor and the renal parenchyma and the renal pelvis and renal medulla.

[0012] As a result, the inventors of the present invention developed a system that automatically segments tumors, parenchyma, renal pelvis, and spleen using an artificial neural network-based predictive model, calculates distances from multiple points on the boundary of the segmented structures to the center of the spleen, and mathematically quantifies this to determine a renal tumor evaluation index.

[0013] The inventors of the present invention expected that the renal tumor evaluation index could reflect the mutual positional relationship between the tumor and the surrounding anatomical structures, thereby overcoming the structural limitations of the existing two-dimensional analysis method and more precisely reflecting the anatomical complexity of the tumor.

[0014] In addition, the inventors of the present invention expected that the evaluation index can objectively and quantitatively analyze the structural characteristics of renal tumors, and thus can be usefully utilized in evaluating the correlation with surgical difficulty and postoperative prognosis.

[0015] Accordingly, the problem to be solved by the present invention is to provide an artificial neural network-based renal tumor evaluation method capable of evaluating the anatomical complexity of a renal tumor by automatically segmenting three-dimensional data of a renal tumor extracted from a medical image using an artificial neural network-based prediction model and quantitatively calculating a renal tumor evaluation index based on the same, and a device and system using the same.

[0016] The subject matter of the present invention is not limited to the above-mentioned subject matter, and may include other subjects that can be clearly understood by those skilled in the art from the following description.

[0017] In order to solve the above-described problem, an evaluation method for a renal tumor according to one embodiment of the present invention is provided.

[0018] The method comprises the steps of receiving a medical image as an evaluation method for a renal tumor implemented by a processor, generating three-dimensional image data from the medical image, segmenting regions for a tumor, parenchyma, renal pelvis, and spleen in the three-dimensional image data using a prediction model learned to segment regions for a tumor, parenchyma, renal pelvis, and spleen by inputting the three-dimensional image data, determining a first boundary surface between the tumor and parenchyma or a second boundary surface between the tumor and spleen based on the segmented regions, determining a first distance between the first boundary surface and the spleen or a second distance between the second boundary surface and the spleen, and performing an evaluation for the renal tumor based on the first distance or the second distance.

[0019] According to a feature of the present invention, the step of dividing the region may include the step of assigning a class number to each region of the tumor, the substance, the renal pelvis, and the newspaper, and the step of predicting the location and boundary of each region based on the assigned class number to determine the region division result.

[0020] According to another feature of the present invention, the method may further include, after the dividing step, a step of applying majority voting to the region dividing result to determine the final dividing result.

[0021] According to another feature of the present invention, the method may further include, after the segmenting step, a step of applying a color map including a color corresponding to each region to the region segmentation result so as to obtain a visualized three-dimensional volume image, and a step of providing a visualized three-dimensional volume image.

[0022] According to another feature of the present invention, the step of determining the first boundary surface or the second boundary surface may include the step of determining the first boundary surface and the second boundary surface based on the segmented region, the step of determining the first distance or the second distance may include the step of determining the first distance and the second distance, and the step of performing an evaluation for the renal tumor may include the step of performing an evaluation for the renal tumor based on the first distance and the second distance.

[0023] According to another feature of the present invention, the step of determining the first distance and the second distance may include the step of determining the first distance between any point selected from the first boundary surface and the center of the newspaper, and the step of determining the second distance between any point selected from the second boundary surface and the center of the newspaper.

[0024] According to another feature of the present invention, the step of performing an evaluation for a renal tumor may include a step of reciprocating the values ​​of the calculated first distance and second distance, a step of determining an evaluation index for the renal tumor by adding the values ​​of the reciprocated first distance and second distance, and a step of quantitatively evaluating the anatomical complexity of the tumor based on the evaluation index for the renal tumor.

[0025] According to another feature of the present invention, the step of determining the first distance and the second distance may include the step of determining a plurality of first distances from each of a plurality of points selected from the first boundary surface to the center of the newspaper, and the step of determining a plurality of second distances from each of a plurality of points selected from the second boundary surface to the center of the newspaper, and the step of determining the renal tumor evaluation index may include the step of reciprocating each of the plurality of calculated first distance values, summing the plurality of reciprocated first distance values, reciprocating each of the plurality of calculated second distance values, summing the plurality of reciprocated second distance values, and further summing the summed plurality of reciprocated first distance values ​​and the summed plurality of reciprocated second distance values ​​to determine the renal tumor evaluation index.

[0026] According to another feature of the present invention, the step of generating three-dimensional image data may include a step of generating three-dimensional image data using 3D rendering software.

[0027] According to another feature of the present invention, the medical image may include at least one of a CT image, an MRI image, and an ultrasound image.

[0028] According to another feature of the present invention, the step of determining the first distance or the second distance may include the step of determining the first distance or the second distance based on a Euclidean distance.

[0029] According to another feature of the present invention, the method further comprises a step of receiving morphological characteristics of the tumor, and the step of performing an evaluation of the renal tumor may comprise a step of performing an evaluation of the renal tumor based on the first distance, the second distance, and the morphological characteristics of the tumor.

[0030] According to another feature of the present invention, the method may include a step of calculating a curvature of a first boundary surface and a curvature of a second boundary surface, and a step of performing an evaluation of a renal tumor based on the calculated curvature of the first boundary surface and the curvature of the second boundary surface.

[0031] In order to solve the above-described problem, a device for evaluating a renal tumor according to another embodiment of the present invention is provided.

[0032] The device may include a communication unit configured to receive a medical image including a kidney and a renal tumor, and a processor functionally connected to the communication unit. The processor is configured to generate three-dimensional image data from the medical image, and segment regions for the tumor, the parenchyma, the renal pelvis, and the spleen in the three-dimensional image data using a prediction model learned to segment regions for the tumor, the parenchyma, the renal pelvis, and the spleen as inputs of the three-dimensional image data, determine a first boundary between the tumor and the parenchyma or a second boundary between the tumor and the spleen based on the segmented regions, determine a first distance from the first boundary to the spleen or a second distance from the second boundary to the spleen, and perform an evaluation of the renal tumor based on the first distance or the second distance.

[0033] According to a feature of the present invention, the processor may be further configured to assign a class number to each region of the tumor, the substance, the kidney, and the newspaper, and to predict the location and boundary of each region according to the assigned class number to determine the region segmentation result.

[0034] According to another feature of the present invention, the processor may be further configured to determine the final division result by applying majority voting to the region division result.

[0035] According to another feature of the present invention, the processor may be further configured to apply a color map including a color corresponding to each region to the region segmentation result so as to obtain a visualized three-dimensional volume image and provide the visualized three-dimensional volume image.

[0036] According to another feature of the present invention, the processor may be configured to determine a first boundary surface and a second boundary surface based on the segmented region, determine a first distance and a second distance, and perform an evaluation of a renal tumor based on the first distance and the second distance.

[0037] According to another feature of the present invention, the processor may be configured to determine a first distance between any point selected from the first boundary surface and the center of the newspaper, and to determine a second distance between any point selected from the second boundary surface and the center of the newspaper.

[0038] According to another feature of the present invention, the processor may be configured to reciprocate the values ​​of the first distance and the second distance, respectively, and determine a renal tumor evaluation index by adding up the reciprocated values ​​of the first distance and the second distance, and quantitatively evaluate the anatomical complexity of the tumor based on the renal tumor evaluation index.

[0039] According to another feature of the present invention, the processor may be further configured to determine a plurality of first distances from each of a plurality of points selected from the first boundary surface to the center of the newspaper, determine a plurality of second distances from each of a plurality of points selected from the second boundary surface to the center of the newspaper, reciprocate each of the plurality of calculated first distance values, sum the plurality of reciprocated first distance values, reciprocate each of the plurality of calculated second distance values, sum the plurality of reciprocated second distance values, and further sum the summed reciprocated first distance values ​​and the summed reciprocated second distance values ​​to determine a renal tumor evaluation index.

[0040] According to another feature of the present invention, the processor may be configured to generate three-dimensional image data using 3D rendering software.

[0041] According to another feature of the present invention, the processor may be configured to determine the first distance or the second distance based on a Euclidean distance.

[0042] According to another feature of the present invention, the communication unit may be configured to receive morphological characteristics of the tumor, and the processor may be configured to perform a renal tumor evaluation based on the first distance, the second distance, and the morphological characteristics of the tumor.

[0043] According to another feature of the present invention, the processor may be configured to calculate a curvature of a first boundary surface and a curvature of a second boundary surface, and perform a renal tumor evaluation based on the calculated curvatures of the first boundary surface and the second boundary surface.

[0044] In order to solve the aforementioned problem, a system for evaluating a renal tumor according to another embodiment of the present invention is provided.

[0045] The system comprises an internal memory configured to store a prediction model learned to segment regions for tumor, parenchyma, renal pelvis, and spleen by inputting medical images and three-dimensional image data including a kidney and a renal tumor, and a processing unit configured to access the internal memory, generate three-dimensional image data from the medical images, segment regions for tumor, parenchyma, renal pelvis, and spleen in the three-dimensional image data using the prediction model, determine a first boundary between the tumor and parenchyma or a second boundary between the tumor and spleen based on the segmented regions, determine a first distance from the first boundary to the spleen or a second distance from the second boundary to the spleen, and perform a renal tumor evaluation based on the first distance or the second distance.

[0046] Specific details of other embodiments are included in the detailed description and drawings.

[0047] and can provide more objective and quantitative information.

[0048] More specifically, the present invention comprehensively analyzes the size of a tumor, the degree of invasion into the renal parenchyma, the distance from the renal pelvis and the spleen based on medical image data, and at the same time uses a learned prediction model to automatically segment the areas of the tumor, parenchyma, renal pelvis, and spleen, thereby providing quantitative evaluation values ​​and constructing a consistent and reliable evaluation system through automated analysis results, unlike the existing method that relies on the subjective experience of medical staff.

[0049] Through this, the present invention can improve the prognosis of renal cancer patients and support safer and more effective renal tumor treatment.

[0050] That is, the present invention can increase the efficiency of the diagnosis and treatment process of renal cancer and improve the predictability of surgical results by providing a prediction model and evaluation system that can more precisely and objectively evaluate the structural characteristics of renal tumors.

[0051] The effects according to the present invention are not limited to those exemplified above, and more diverse effects are included in this specification.

[0052] FIG. 1 is an exemplary diagram illustrating a device-based system for evaluating renal tumors according to one embodiment of the present invention.

[0053] FIG. 2A is a block diagram showing the configuration of a user device according to one embodiment of the present invention.

[0054] FIG. 2b is a block diagram showing the configuration of a server of an evaluation device according to one embodiment of the present invention.

[0055] Figures 3a to 3d illustrate the procedure of an evaluation method for a renal tumor according to one embodiment of the present invention.

[0056] FIG. 4 illustrates an evaluation procedure for a renal tumor according to one embodiment of the present invention.

[0057] Figures 5a to 5e illustrate evaluation results of an evaluation system for renal tumors according to various embodiments of the present invention.

[0058] The advantages of the invention and the methods for achieving them will become clearer with reference to the embodiments described below in detail with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms. These embodiments are provided solely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention. The present invention is defined solely by the scope of the claims.

[0059] The shapes, sizes, ratios, angles, numbers, etc. disclosed in the drawings for explaining embodiments of the present invention are exemplary, and therefore the present invention is not limited to the matters illustrated. In addition, in describing the present invention, if it is determined that a detailed description of a related known technology may unnecessarily obscure the gist of the present invention, the detailed description thereof will be omitted. When the terms “includes,” “has,” and “consists of” are used in this specification, other parts may be added unless “only” is used. When a component is expressed in the singular, it includes a case where the plural is included unless there is a specifically explicit description.

[0060] When interpreting components, it is interpreted as including the error range even if there is no separate explicit description.

[0061] The individual features of the various embodiments of the present invention can be partially or wholly combined or combined with each other, and as can be fully understood by those skilled in the art, various technical connections and operations are possible, and each embodiment can be implemented independently of each other or can be implemented together in a related relationship.

[0062] For clarity in the interpretation of this specification, the terms used in this specification are defined below.

[0063] As used herein, the term "medical image" may refer to image data taken of a body part, including a kidney or a kidney tumor. In various embodiments of the present invention, medical images may include, but are not limited to, computed tomography (CT) images, magnetic resonance imaging (MRI) images, ultrasound images, and the like.

[0064] As used herein, the term "three-dimensional image data" may refer to three-dimensional data generated based on medical images. In various embodiments of the present invention, the three-dimensional image data may be data converted to enable structural analysis of kidneys and renal tumors using 3D rendering software. However, the present invention is not limited thereto.

[0065] As used herein, the term "segmentation" may refer to the process of identifying and separating regions such as kidneys, tumors, parenchyma, renal pelvis, and kidney tissue in three-dimensional image data. In various embodiments of the present invention, segmentation may be performed using, but is not limited to, an automatic segmentation technique using an artificial neural network-based predictive model.

[0066] The term "prediction model" as used herein may mean an artificial neural network-based model trained to automatically identify specific anatomical structures such as renal tumors, renal parenchyma, renal pelvis, and renal pelvis using medical image data as input and segment the regions.

[0067] In various embodiments of the present invention, the prediction model may be constructed based on a deep learning algorithm, and in various embodiments of the present invention, the prediction model may be based on an artificial neural network structure optimized for medical image processing, such as nnU-Net, U-Net, 3D convolutional neural network (3D-CNN), or the like.

[0068] In various embodiments of the present invention, the predictive model is trained based on a large number of patient data to automatically recognize morphological features of renal tumors and surrounding structures, and assign a unique class number to each structure to generate a region segmentation result at the pixel or voxel level.

[0069] For example, renal parenchyma may be assigned class number 1, tumor may be assigned class number 2, renal pelvis may be assigned class number 3, and renal pelvis may be assigned class number 4.

[0070] In this way, when a unique identification number is assigned to each structure, a corresponding class number is assigned to each voxel in the medical image, allowing for systematic distinction of the location and boundaries of the segmented structures. The predictive model analyzes the image characteristics of each voxel according to learned criteria and assigns one of the above class numbers, thereby automatically classifying the input medical image data into multiple structures for renal tumor evaluation.

[0071] These class numbers can provide consistent structural identification information for subsequent quantitative analysis, visualization, distance calculation, and evaluation index calculation processes, and can contribute to improving the accuracy and reliability of the overall evaluation system.

[0072] In various embodiments of the present invention, the prediction model may be configured to improve generalization performance through, but is not limited to, majority voting based on a learning dataset, and various artificial intelligence techniques and learning methods may be applied.

[0073] Accordingly, the prediction model used in the present invention can be utilized as an element technology capable of generating accurate and repeatable automatic analysis results on the anatomical structure of renal tumors without manual segmentation or subjective judgment of a medical professional.

[0074] As used herein, the term "tumor" may refer to a lesion arising from abnormal cell proliferation within the kidney. For example, a renal tumor may include a malignant or benign tumor, and may include renal cell carcinoma (RCC).

[0075] As used herein, the term "parenchyma" refers to the tissue responsible for kidney function, and may refer to the area within the kidney that filters blood and produces urine. The renal parenchyma may include the renal cortex and renal medulla.

[0076] As used herein, the term "renal sinus" may refer to the internal space of the kidney, where the renal calyx and renal blood vessels are located. In this case, the renal sinus may also be the area where urine collects in the kidney and then moves to the ureter.

[0077] As used herein, the term "renal hilum" may refer to the area in the kidney where the renal artery, renal vein, and ureter enter and exit. The hilum is connected to the renal pelvis and may play a crucial role in the blood flow into the kidney and the route for urine excretion.

[0078] In a more diverse embodiment of the present invention, each of the segmented regions can be provided as a visualized three-dimensional volume image.

[0079] The term "visualized three-dimensional volume image" as used herein may refer to medical image data that visually expresses kidney and kidney tumor-related anatomical structures in three-dimensional space.

[0080] In various embodiments of the present invention, a visualized three-dimensional volume image is based on voxel-unit data generated from a region segmentation result, and can be expressed by applying a color map including a unique color corresponding to each region (e.g., tumor, renal parenchyma, renal pelvis, and nephron).

[0081] For example, tumors can be colored red, renal parenchyma blue, renal pelvis green, and renal spleen yellow, and this color coding allows users to intuitively identify each structure.

[0082] In various embodiments of the present invention, a visualized three-dimensional volume image can be generated through 3D rendering software, and can be provided as a visual information interface that allows medical staff to intuitively understand the location, size, direction of invasion, and spatial relationship with the renal pelvis and spleen of a renal tumor, and utilize it for diagnosis or surgical planning.

[0083] In more diverse embodiments, the visualized three-dimensional volume images may be implemented in a form that allows user manipulation such as transparency adjustment, rotation, and zooming, or may be provided together with parallel images based on 2D slices.

[0084] The term "Renal Tumor Evaluation Index" as used herein may mean a quantitative evaluation value calculated by adding the reciprocals of the first distance and the second distance.

[0085] At this time, the renal tumor evaluation index is an indicator that quantitatively evaluates the anatomical complexity of renal tumors, and can reflect the location of the renal tumor, the renal parenchyma, and the distance relationship to the kidney.

[0086] In more diverse embodiments of the present invention, the renal tumor evaluation index can be calculated by the following mathematical formula 1.

[0087] [Mathematical Formula 1]

[0088]

[0089] Here, CSAM,p may refer to the interface between the renal parenchyma and the tumor (mass) (first interface), Zp may refer to an arbitrary point selected at the interface between the renal parenchyma and the tumor, CSAM,s may refer to the interface between the renal pelvis and the tumor (second interface), Zs may refer to an arbitrary point selected at the interface between the renal pelvis and the tumor, Distance (Zp, Sinus) may refer to the distance between an arbitrary point Zp selected at the interface between the renal parenchyma and the tumor and the center of the renal pelvis, and Distance (Zs, Sinus) may refer to the distance between an arbitrary point Zs selected at the interface between the renal pelvis and the tumor and the center of the renal pelvis.

[0090] As used herein, the term "first interface" may refer to the interface where the tumor and renal parenchyma come into contact. In various embodiments of the present invention, the first interface may serve as a reference for assessing how deeply a renal tumor has invaded the renal parenchyma. However, this is not limited thereto.

[0091] As used herein, the term "second interface" may refer to the interface where the tumor and the lesion come into contact. In various embodiments of the present invention, the second interface may serve as a criterion for assessing the proximity of a renal tumor to the lesion, but is not limited thereto.

[0092] As used herein, the term "first distance" may refer to the distance from the first boundary surface to the center of the newspaper. In various embodiments of the present invention, the first distance may be a numerical value indicating the extent to which a renal tumor has infiltrated the renal parenchyma. However, the present invention is not limited thereto.

[0093] As used herein, the term "a plurality of first distances" may refer to a set of distance values ​​between each of a plurality of points forming the first boundary surface and the center of the newspaper. In various embodiments of the present invention, the plurality of first distances may be measured at a plurality of points sampled at regular intervals along the contact surface between the renal tumor and the renal parenchyma.

[0094] As used herein, the term "reciprocal multiple first distance values" may refer to a set of values ​​obtained by taking the reciprocal of each of the plurality of first distance values. In various embodiments of the present invention, the reciprocal multiple first distance values ​​are designed to have a greater effect the shorter each distance value is, and may be reflected as a factor that increases the difficulty of surgery when a renal tumor is in close contact with the renal parenchyma, but is not limited thereto.

[0095] As used herein, the term "second distance" may refer to the distance from the second boundary surface to the center of the newspaper. In various embodiments of the present invention, the second distance may be a value indicating how close a renal tumor is to the newspaper. However, this is not a limitation.

[0096] As used herein, the term "multiple second distances" may refer to a set of distance values ​​between each of a plurality of points forming the second boundary surface and the center of the newspaper. In various embodiments of the present invention, the plurality of second distances may be measured at a plurality of points sampled along the contact surface between the renal tumor and the newspaper.

[0097] As used herein, the term "reciprocal multiple second distance values" may refer to a set of values ​​obtained by taking the reciprocal of each of the plurality of second distance values. In various embodiments of the present invention, the reciprocal multiple second distance values ​​have higher values ​​the closer the renal tumor is to the kidney, which can be utilized to quantitatively analyze the impact of the location of the renal tumor on the difficulty of surgery and the possibility of preserving renal function.

[0098] As used herein, the term "Euclidean distance" may refer to the straight-line distance between two points in three-dimensional space. In various embodiments of the present invention, the Euclidean distance may be applied to, but is not limited to, measuring the distance between a specific point on the first boundary surface or the second boundary surface and the center of a newspaper.

[0099] For example, the distance between a specific point on each boundary and the center of the newspaper can be calculated using the Euclidean distance formula, which can quantitatively assess how far or close the tumor is located in spatial relationship to the renal parenchyma and newspaper.

[0100] In various embodiments of the present invention, the distance value can be utilized as an objective numerical value representing the anatomical complexity of a renal tumor and can be used to predict surgical difficulty and prognosis, but is not limited thereto.

[0101] The term "curvature" as used herein may refer to a value quantitatively representing the degree of curvature of a first boundary surface or a second boundary surface. In this case, curvature may be utilized to analyze the degree of tumor protrusion or depth of invasion, but is not limited thereto.

[0102] As used herein, the term "morphological characteristics of a tumor" may refer to characteristics that quantitatively analyze the shape, size, complexity of the border, and relationship with the surrounding tissue of a renal tumor. In various embodiments of the present invention, the morphological characteristics of a tumor may include, but are not limited to, at least one of the location of the tumor, the volume of the tumor, the surface area of ​​the tumor, the major and minor axis lengths of the tumor, the sphericity of the tumor, the irregularity of the margin of the tumor, the depth of invasion, and the compactness of the tumor.

[0103] Hereinafter, with reference to FIGS. 1 and 2a and 2b, a renal tumor evaluation system, a user device, and a renal tumor evaluation device using a renal tumor evaluation device according to one embodiment of the present invention will be described.

[0104] FIG. 1 illustrates an evaluation system for renal tumors using a device for evaluating renal tumors according to one embodiment of the present invention. FIG. 2a is a block diagram illustrating the configuration of a user device according to one embodiment of the present invention. FIG. 2b is a block diagram illustrating the configuration of a server for the evaluation device according to one embodiment of the present invention.

[0105] First, referring to FIG. 1, a renal tumor evaluation system (1000) may be a system configured to provide information on anatomical complexity by calculating a renal tumor evaluation index. At this time, the renal tumor evaluation system (1000) may be configured with a user device (100) that receives information on renal tumor evaluation, a medical imaging scanner (200) that provides medical images, and a renal tumor evaluation server (300) that calculates a renal tumor evaluation index based on the received medical images to determine information on anatomical complexity.

[0106] First, the user device (100) is an electronic device that provides a user interface for displaying evaluation results for a renal tumor, and may include at least one of a smartphone, a tablet PC (personal computer), a laptop, and / or a PC.

[0107] The user device (100) can receive evaluation results for a renal tumor from the evaluation server (300) and display the received results through a display unit.

[0108] The evaluation server (300) may include a general-purpose computer, laptop, and / or data server that performs various operations for evaluating a renal tumor from medical images provided from a medical imaging scanner (200). In this case, the evaluation server (300) may be, but is not limited to, a device for accessing a web server that provides web pages or a mobile web server that provides a mobile website.

[0109] In various embodiments of the present invention, the evaluation server (300) receives a CT image from a medical imaging scanner (200), generates three-dimensional image data from the received CT image, and then applies an artificial neural network-based prediction model learned to automatically segment structures for a renal tumor, renal parenchyma, renal pelvis, and renal spleen, thereby distinguishing boundaries between anatomical structures and producing quantitative data necessary for evaluation.

[0110] Next, the evaluation server (300) can provide the evaluation results for the renal tumor to the user device (100).

[0111] The evaluation results for renal tumors provided from the evaluation server (300) in this way may be provided as a web page via a web browser installed on the user device (100), or may be provided in the form of an application or program. In various embodiments, such data may be provided in a form included in the platform in a client-server environment.

[0112] Next, with reference to FIGS. 2a and 2b, the components of the evaluation server (300) of the present invention will be described in detail.

[0113] First, referring to FIG. 2A, a user device (100) may include a memory interface (110), one or more processors (120), and a peripheral interface (130). Various components within the user device (100) may be connected by one or more communication buses or signal lines.

[0114] The memory interface (110) is connected to the memory (150) and can transmit various data to the processor (120). Here, the memory (150) can include at least one type of storage medium among flash memory type, hard disk type, multimedia card micro type, card type memory (e.g., SD or XD memory, etc.), RAM, SRAM, ROM, EEPROM, PROM, network storage, cloud, and blockchain data.

[0115] In various embodiments, the memory (150) can store at least one of an operating system (151), a communication module (152), a graphical user interface (GUI) module (153), a sensor processing module (154), a telephone module (155), and an application module (156). Specifically, the operating system (151) can include instructions for processing basic system services and instructions for performing hardware operations. The communication module (152) can communicate with at least one of other devices, computers, and servers. The graphical user interface (GUI) module (153) can process a graphical user interface. The sensor processing module (154) can process sensor-related functions (e.g., processing voice input received using one or more microphones (192)). The telephone module (155) can process telephone-related functions. The application module (156) can perform various functions of the user application, such as electronic messaging, web browsing, media processing, navigation, imaging, and other processing functions. In addition, the user device (100) can store one or more software applications (156-1, 156-2) (e.g., an image enhancement application) associated with a type of service in the memory (150).

[0116] In various embodiments, the memory (150) can store a digital assistant client module (157) (hereinafter, DA client module), and accordingly, can store commands for performing client-side functions of the digital assistant and various user data (158).

[0117] Meanwhile, the DA client module (157) can obtain the user's voice input, text input, touch input, and / or gesture input through various user interfaces (e.g., I / O subsystem (140)) provided in the user device (100).

[0118] Additionally, the DA client module (157) can output data in audiovisual and tactile forms. For example, the DA client module (157) can output data consisting of a combination of at least two or more of voice, sound, notification, text message, menu, graphic, video, animation, and vibration. In addition, the DA client module (157) can communicate with a digital assistant server (not shown) using a communication subsystem (180).

[0119] In various embodiments, the DA client module (157) may collect additional information about the surroundings of the user device (100) from various sensors, subsystems, and peripheral devices to construct a context associated with the user input. For example, the DA client module (157) may provide context information along with the user input to a digital assistant server to infer the user's intent. Here, the context information that may accompany the user input may include sensor information, such as lighting, ambient noise, ambient temperature, images of the surrounding environment, video, etc. As another example, the context information may include the physical state of the user device (100) (e.g., device orientation, device position, device temperature, power level, speed, acceleration, motion pattern, cellular signal strength, etc.). As another example, context information may include information related to the software state of the user device (100) (e.g., processes running on the user device (100), installed programs, past and current network activity, background services, error logs, resource usage, etc.).

[0120] In various embodiments, the memory (150) may include additional or deleted instructions, and further, the user device (100) may include additional configurations other than those illustrated in FIG. 2A, or may exclude some configurations.

[0121] The processor (120) can control the overall operation of the user device (100) and execute various commands to implement an interface that provides converted CT images by driving an application or program stored in the memory (150).

[0122] The processor (120) may correspond to a computing device such as a CPU (Central Processing Unit) or an AP (Application Processor). In addition, the processor (120) may be implemented in the form of an integrated chip (Integrated Chip (IC)) such as a SoC (System on Chip) in which various computing devices such as an NPU (Neural Processing Unit) are integrated.

[0123] The peripheral interface (130) can be connected to various sensors, subsystems, and peripheral devices to provide data so that the user device (100) can perform various functions. Here, it can be understood that the user device (100) performs a certain function as being performed by the processor (120).

[0124] The peripheral interface (130) can receive data from a motion sensor (160), a light sensor (light sensor) (161), and a proximity sensor (162), through which the user device (100) can perform orientation, light, and proximity detection functions, etc. For another example, the peripheral interface (130) can receive data from other sensors (163) (positioning system - GPS receiver, temperature sensor, biometric sensor), through which the user device (100) can perform functions related to the other sensors (163).

[0125] In various embodiments, the user device (100) may include a camera subsystem (170) connected to a peripheral interface (130) and an optical sensor (171) connected thereto, which may enable the user device (100) to perform various photographing functions, such as taking pictures and recording video clips.

[0126] In various embodiments, the user device (100) may include a communication subsystem (180) connected to a peripheral interface (130). The communication subsystem (180) may be comprised of one or more wired / wireless networks and may include various communication ports, radio frequency transceivers, and optical transceivers.

[0127] In various embodiments, the user device (100) includes an audio subsystem (190) coupled to a peripheral interface (130), the audio subsystem (190) including one or more speakers (191) and one or more microphones (192), such that the user device (100) can perform voice-activated functions, such as voice recognition, voice replication, digital recording, and telephony functions.

[0128] In various embodiments, the user device (100) may include an I / O subsystem (140) coupled to a peripheral interface (130). For example, the I / O subsystem (140) may control a touch screen (143) included in the user device (100) via a touch screen controller (141). As an example, the touch screen controller (141) may detect a user's contact and movement or cessation of contact and movement using any one of a plurality of touch sensing technologies, such as capacitive, resistive, infrared, surface acoustic wave technology, proximity sensor array, etc. As another example, the I / O subsystem (140) may control other input / control devices (144) included in the user device (100) via other input controller(s) (142). As an example, the other input controller(s) (142) may control one or more buttons, rocker switches, thumb-wheels, infrared ports, USB ports, and pointer devices such as a stylus.

[0129] Next, referring to FIG. 2b, the evaluation server (300) may include a communication interface (310), a memory (320), an I / O interface (330), and a processor (340), each component of which may communicate with one another via one or more communication buses or signal lines.

[0130] The communication interface (310) can be connected to a user device (100) and a medical imaging scanner (200) via a wired / wireless network to exchange data. For example, the communication interface (310) can receive a medical image, such as a CT image, from the medical imaging scanner (200) and transmit the evaluation results for a renal tumor to the user device (100).

[0131] Meanwhile, the communication interface (310) that enables transmission and reception of such data includes a communication port (311) and a wireless circuit (312), wherein the wired communication port (311) may include one or more wired interfaces, for example, Ethernet, Universal Serial Bus (USB), FireWire, etc. In addition, the wireless circuit (312) may transmit and receive data with an external device via an RF signal or an optical signal. In addition, the wireless circuit may use at least one of a plurality of communication standards, protocols, and technologies, for example, GSM, EDGE, CDMA, TDMA, Bluetooth, Wi-Fi, VoIP, Wi-MAX, or any other suitable communication protocol.

[0132] The memory (320) can store various data used in the evaluation server (300). For example, the memory (320) can be configured to store medical images received from a medical imaging scanner (200), data generated during a renal tumor evaluation process, renal tumor evaluation results, etc.

[0133] In various embodiments, the memory (320) may include a volatile or non-volatile storage medium capable of storing various data, commands, and information. For example, the memory (320) may include at least one type of storage medium among flash memory type, hard disk type, multimedia card micro type, card type memory (e.g., SD or XD memory, etc.), RAM, SRAM, ROM, EEPROM, PROM, network storage storage, cloud, and blockchain data.

[0134] In various embodiments, the memory (320) may store a configuration of at least one of an operating system (321), a communication module (322), a user interface module (323), and one or more applications (324).

[0135] An operating system (321) (e.g., an embedded operating system such as LINUX, UNIX, MAC OS, WINDOWS, VxWorks, etc.) may include various software components and drivers to control and manage general system operations (e.g., memory management, storage device control, power management, etc.) and may support communication between various hardware, firmware, and software components.

[0136] The communication module (322) can support communication with other devices through the communication interface (310). The communication module (322) can include various software components for processing data received by the wired communication port (311) or wireless circuit (312) of the communication interface (310).

[0137] The user interface module (323) can receive a user's request or input from a keyboard, touch screen, microphone, etc. through an I / O interface (330) and provide a user interface on the display.

[0138] The application (324) may include a program or module configured to be executed by one or more processors (340). Here, the application for providing converted CT images may be implemented on a server farm.

[0139] The I / O interface (330) can connect at least one of an input / output device (not shown) of the evaluation server (300), such as a display, a keyboard, a touch screen, and a microphone, to the user interface module (323). The I / O interface (330) can receive user input (e.g., voice input, keyboard input, touch input, etc.) together with the user interface module (323) and process a command according to the received input.

[0140] The processor (340) is connected to a communication interface (310), a memory (320), and an I / O interface (330) to control the overall operation of the evaluation server (300), and can perform various commands for evaluating a renal tumor through an application or program stored in the memory (320).

[0141] Additionally, the processor (340) may be configured to execute an artificial neural network-based prediction model for automatically segmenting structures such as renal tumors, parenchyma, renal pelvis, and spleen from medical images, and may include a Neural Processing Unit (NPU) to execute the prediction model.

[0142] For example, the processor (340) can perform operations such as distance calculation, reciprocal calculation, and exponentiation for evaluating the anatomical complexity of a renal tumor, and can be linked with computational hardware (e.g., GPU, TPU, etc.) for high-speed parallel processing of such operations.

[0143] The processor (340) may correspond to a computing device such as a CPU (Central Processing Unit) or an AP (Application Processor). In addition, the processor (340) may be implemented in the form of an integrated chip (Integrated Chip (IC)) such as a SoC (System on Chip) in which various computing devices are integrated, and may also be implemented in the form of an integrated computing device including an NPU that performs artificial neural network-based computing.

[0144]

[0145] Hereinafter, with reference to FIGS. 3a to 3d and 4, an evaluation method for a renal tumor according to various embodiments of the present invention will be specifically described.

[0146] Figures 3a to 3d illustrate procedures for evaluating a renal tumor according to one embodiment of the present invention. Figure 4 exemplarily illustrates an evaluation procedure for a renal tumor according to one embodiment of the present invention.

[0147] First, referring to FIG. 3A, an evaluation procedure according to an embodiment of the present invention is as follows. First, a medical image including a kidney and a renal tumor is received (S310). Next, three-dimensional image data is generated from the received medical image (S320). Next, region segmentation of the tumor, parenchyma, renal pelvis, and spleen is performed in the three-dimensional image data using a prediction model (S330). Next, a first boundary surface between the tumor and parenchyma or a second boundary surface between the tumor and spleen is determined (S340), a first distance between the first boundary surface and the spleen or a second distance between the second boundary surface and the spleen is determined (S350), and a renal tumor evaluation is performed based on the first distance or the second distance (S360).

[0148] According to various embodiments of the present invention, in the step (S330) where region segmentation is performed, a class number is assigned to each region of the tumor, the substance, the renal pelvis, and the spleen by the prediction model, and the location and boundary of each region are predicted according to the assigned class number, thereby determining the region segmentation result.

[0149] For example, renal parenchyma may be assigned class number 1, tumor may be assigned class number 2, renal pelvis may be assigned class number 3, and renal pelvis may be assigned class number 4.

[0150] In this way, when a unique identification number is assigned to each structure, a corresponding class number is assigned to each voxel unit of the medical image, allowing for systematic distinction of the location and boundary of the segmented structure.

[0151] The predictive model can automatically classify the input medical image data into multiple structures for renal tumor evaluation by analyzing the image features of each voxel according to learned criteria and assigning one of the above class numbers.

[0152] These class numbers can provide consistent structural identification information for subsequent quantitative analysis, visualization, distance calculation, and evaluation index calculation processes, and can contribute to improving the accuracy and reliability of the overall evaluation system.

[0153] According to various embodiments of the present invention, after the step (S330) in which region division is performed, a majority voting method may be applied to the region division result to determine the final division result.

[0154] For example, according to various embodiments of the present invention, the predicted class values ​​for each voxel are compared for the segmentation results obtained from each of the cross-validation results for the learned multiple prediction models or the same prediction model, and the final label can be assigned to the class value predicted most frequently at that location.

[0155] This majority voting method can produce more reliable segmentation results than methods that rely on a single model or a single prediction result, reflecting the consistency of prediction results between models or between folds, and can contribute to minimizing the influence of noise or outliers.

[0156] Accordingly, in various embodiments of the present invention, subsequent quantitative analyses such as calculating a renal tumor evaluation index, calculating distances between anatomical structures, and predicting surgical difficulty can be performed based on the final segmentation results to which a majority voting method is applied.

[0157] According to various embodiments of the present invention, a step of applying a color map including a color corresponding to each region to a region segmentation result to obtain a visualized 3D volume image and providing a visualized 3D volume image may be further performed.

[0158] Referring to FIG. 3b, according to various embodiments of the present invention, three-dimensional image data (422) is generated from an original medical image (412) acquired from a medical imaging scanner, and the image data is input into a learned prediction model (430) to perform segmentation of a tumor, renal parenchyma, renal pelvis, and nephron.

[0159] Thereafter, a color map with a unique color assigned to each anatomical structure is applied to the segmentation result (432) corresponding to the output result of the prediction model (430), so that the segmentation result (432) can be expressed as a visualized 3D volume image.

[0160] The visualized 3D volume images generated in this way enable an intuitive understanding of the spatial relationship between the tumor and surrounding organs, and can help medical staff more clearly understand the location of the renal tumor, the degree of invasion, and the surgical margins.

[0161] That is, the present invention can implement a renal tumor evaluation system that can be usefully utilized in clinical settings by providing automatic segmentation results based on a prediction model as intuitive visual information.

[0162] Meanwhile, referring to FIG. 3c, in various embodiments of the present invention, a first distance between a selected arbitrary point on the first boundary surface and the center of the newspaper is determined (S3502), and a second distance between a selected arbitrary point on the second boundary surface and the center of the newspaper is determined (S3504). Then, the values ​​of the calculated first and second distances are each reciprocal and summed to determine a renal tumor evaluation index (S3602), and the anatomical complexity of the renal tumor is quantitatively evaluated based on the renal tumor evaluation index (S3604).

[0163] That is, the distance from each of the plurality of points constituting the first boundary surface and the second boundary surface to the center of the newspaper is calculated, and the calculated distance values ​​are reciprocally added up to determine a renal tumor evaluation index, and the anatomical complexity of the renal tumor can be quantitatively evaluated based on the index.

[0164] According to another embodiment of the present invention, in the step (S3602) where a renal tumor evaluation index is determined, the renal tumor evaluation index can be calculated by the following mathematical expression 1.

[0165] [Mathematical Formula 1]

[0166]

[0167] Here, CSAM,p may refer to the interface between the renal parenchyma and the tumor (mass) (first interface), Zp may refer to an arbitrary point selected at the interface between the renal parenchyma and the tumor, CSAM,s may refer to the interface between the renal pelvis and the tumor (second interface), Zs may refer to an arbitrary point selected at the interface between the renal pelvis and the tumor, Distance (Zp, Sinus) may refer to the distance between an arbitrary point Zp selected at the interface between the renal parenchyma and the tumor and the center of the renal pelvis, and Distance (Zs, Sinus) may refer to the distance between an arbitrary point Zs selected at the interface between the renal pelvis and the tumor and the center of the renal pelvis.

[0168] More specifically, referring to FIG. 3d and FIG. 4 (a) and (b) together, in various embodiments of the present invention, a plurality of first distances are determined from each of a plurality of points selected from the first boundary surface to the center of the newspaper (S3506), a plurality of second distances are determined from each of a plurality of points selected from the second boundary surface to the center of the newspaper (S3508), each of the plurality of first distance values ​​is reciprocated, and the plurality of reciprocated first distance values ​​are summed (S36022), each of the plurality of second distance values ​​is reciprocated, and the plurality of reciprocated second distance values ​​are summed (S36024), and finally, the summed reciprocated first distance values ​​and the summed reciprocated second distance values ​​are further summed, thereby determining a renal tumor evaluation index (S36026).

[0169] That is, the distances of each of the points constituting the first interface (the contact surface between the renal parenchyma and the tumor) and the second interface (the contact surface between the renal pelvis and the tumor) to the center of the renal pelvis are calculated, and the corresponding distance values ​​are reciprocally added and then added to finally determine the renal tumor evaluation index. In various embodiments of the present invention, the anatomical complexity of a renal tumor can be quantitatively evaluated by reciprocally adding the distance values ​​at all points (Zp∈CSAM,p) constituting the first interface and the distance values ​​at all points (Zs∈CSAM,s) constituting the second interface.

[0170] These evaluation indices can enable a highly precise evaluation of renal tumors by reflecting the locational characteristics of renal tumors and their spatial relationship with the renal parenchyma and renal pelvis.

[0171] Returning to FIG. 3a, in a more diverse embodiment, in the step (S320) of generating three-dimensional image data, three-dimensional image data may be generated using three-dimensional rendering software. Here, the three-dimensional rendering software may be at least one of 3D Slicer, ITK-SNAP, and InVesalius 3, but is not limited thereto.

[0172] According to a feature of the present invention, in the step (S350) where the first distance or the second distance is calculated, the first distance or the second distance may be determined based on the Euclidean distance, but is not limited thereto.

[0173] According to another feature of the present invention, a step of receiving morphological characteristics of the tumor and performing a renal tumor evaluation based on the first distance, the second distance and the morphological characteristics of the tumor may be further performed.

[0174] According to another feature of the present invention, the curvature of the first boundary surface and the curvature of the second boundary surface are calculated, and renal tumor evaluation may be performed based on the curvature of the first boundary surface and the curvature of the second boundary surface.

[0175]

[0176] Evaluation: Evaluation of the scoring system for renal tumors

[0177] In the following examples, evaluation results of a renal tumor evaluation system according to various embodiments of the present invention are described with reference to FIGS. 5a to 5e.

[0178] First, referring to Figure 5a, preoperative clinical information, surgical-related information, postoperative complications, and renal tumor evaluation indices collected from 228 renal cancer patients are illustrated.

[0179] More specifically, as renal tumor evaluation indices, the PADUA score was 8.38 ± 1.87 on average, the RENAL score was 6.92 ± 2.04, and the C-index was 3.27 ± 1.92 cm, and the renal tumor evaluation index according to various embodiments of the present invention was 114.47 ± 99.86 on average.

[0180] At this time, the renal tumor evaluation index can be provided as an indicator for quantitatively evaluating the anatomical complexity of the renal tumor by reciprocally adding up the distance values ​​from the center of the renal parenchyma to the center of the kidney.

[0181] Referring to FIG. 5b, a comparison result between an automatic renal tumor evaluation index according to various embodiments of the present invention and an evaluation index manually calculated by medical staff is illustrated.

[0182] More specifically, the renal tumor assessment index produced by the automatic method shows higher mean and median values ​​overall than the manual method, and the difference between the two methods is statistically significant.

[0183] These results may suggest that the automatic renal tumor assessment method based on a predictive model according to various embodiments of the present invention can provide a more consistent and reproducible renal tumor assessment index compared to manual calculation by medical staff.

[0184] That is, the present invention can reliably quantify the anatomical complexity of a renal tumor through automatic analysis using a predictive model, and can be applied to an evaluation system that does not depend on the subjective judgment of medical staff.

[0185] Next, referring to FIG. 5c, as an example visually showing the segmentation performance and evaluation index calculation results of the automatic renal tumor evaluation system according to various embodiments of the present invention, the manual segmentation results performed by medical staff and the automatic segmentation results performed by the prediction model are compared and shown.

[0186] More specifically, (a), (b) and (c) of FIG. 5c represent manual segmentation results, and (d), (e) and (f) of FIG. 5c represent automatic segmentation results using a prediction model according to various embodiments of the present invention.

[0187] At this time, the tumor is displayed in yellow and the renal parenchyma is displayed in green in each image, and the automatic segmentation results in both the 3D rendering image and the CT image are similar to the manual segmentation results in terms of anatomical structure and shape.

[0188] More specifically, the renal tumor evaluation index calculated from the manual segmentation method was 29.79, and the evaluation index calculated from the automatic segmentation method was 27.03. At this time, the two index values ​​are similar, and it appears that the automatic evaluation system of the present invention can provide results close to the manual method.

[0189] These results suggest that the predictive model-based automatic segmentation and evaluation method according to various embodiments of the present invention can provide reliable and highly reproducible renal tumor evaluation results and has high clinical applicability by reducing the burden of manual work on medical staff.

[0190] Next, referring to FIG. 5d, the results of Bland-Altman analysis are shown to evaluate the quantitative consistency between the renal tumor evaluation index calculated by the automatic segmentation method according to various embodiments of the present invention and the manual segmentation method of medical staff.

[0191] More specifically, at this time, the average of the manual evaluation index and the automatic evaluation index for each patient was displayed on the horizontal axis, and the difference between the two indices was displayed on the vertical axis, visualizing the distribution of deviations between the two methods.

[0192] The average difference between the two evaluation methods was approximately 12.1, indicating that most data were distributed within the mean ±1.96 standard deviations (SD). This suggests that the automated evaluation method yields quantitative results similar to those of manual evaluation, suggesting that the overall agreement between the segmentation methods is similar.

[0193] That is, the automatic segmentation-based renal tumor evaluation method according to various embodiments of the present invention has precision and reliability close to that of a manual method, and can be utilized as an automated evaluation system that can secure repeatability and consistency in a clinical environment.

[0194] Next, FIG. 5e illustrates the results of analyzing the effect of the automatic renal tumor assessment index according to various embodiments of the present invention on the occurrence of postoperative complications.

[0195] At this time, analyses were performed by dividing into overall postoperative complications and severe postoperative complications for both univariable logistic regression and multivariable logistic regression.

[0196] More specifically, according to the results of univariate analysis, the automated renal tumor assessment index according to various embodiments of the present invention showed a significant positive correlation with the occurrence of overall postoperative complications (OR = 1.004, 95% CI: 1.001-1.007, p = 0.014), and also showed a significant association with the occurrence of severe complications (OR = 1.005, 95% CI: 1.000-1.010, p = 0.046).

[0197] Additionally, in the multivariate analysis results, the automated renal tumor assessment index of the present invention appears as a statistically significant predictive factor for both overall complications (OR = 1.004, p = 0.006) and severe complications (OR = 1.005, p = 0.006).

[0198] These results suggest that the automatic evaluation system based on the renal tumor evaluation index according to various embodiments of the present invention can be utilized as a reliable indicator that can quantitatively predict postoperative prognosis, and may be useful for predicting surgical risk in advance in a more objective and consistent manner than existing evaluation indicators.

[0199] Accordingly, the present invention can more quantitatively evaluate the anatomical complexity of a renal tumor, and by reflecting the three-dimensional structural relationship between the tumor and the renal parenchyma and kidney, it has the effect of more accurately reflecting anatomical elements that were not considered in existing evaluation indicators.

[0200] In particular, the renal tumor evaluation index provided in various embodiments of the present invention quantitatively indicates the correlation with surgical difficulty, risk of postoperative complications, decline in renal function, and possibility of CKD progression, compared to existing evaluation indices (e.g., PADUA, RENAL, C-index, CSA), and thus has the effect of predicting various clinical conditions in a more reliable and consistent manner.

[0201] Therefore, the renal tumor evaluation index of the present invention can be utilized as a useful evaluation index for establishing a patient-tailored treatment plan and predicting surgical risk in advance in renal oncology.

[0202] Although the embodiments of the present invention have been described in more detail with reference to the attached drawings, the present invention is not necessarily limited to these embodiments, and various modifications may be implemented without departing from the technical spirit of the present invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical spirit of the present invention, but to explain it, and the scope of the technical spirit of the present invention is not limited by these embodiments. Therefore, it should be understood that the embodiments described above are exemplary in all aspects and not restrictive. The protection scope of the present invention should be interpreted by the claims below, and all technical ideas within a scope equivalent thereto should be interpreted as being included in the scope of the rights of the present invention.

[0203] 100: User Device

[0204] 110: Memory interface 120: Processor

[0205] 130: Peripheral Interface 140: I / O Subsystem

[0206] 141: Touch screen controller 142: Other input controllers

[0207] 143: Touch screen

[0208] 144: Other input control devices

[0209] 150: Memory 151: Operating System

[0210] 152: Communication Module 153: GUI Module

[0211] 154: Sensor processing module 155: Phone module

[0212] 156: Applications

[0213] 156-1, 156-2: Applications

[0214] 157: Digital Assistant Client Module

[0215] 158: User data

[0216] 160: Motion sensor 161: Light sensor

[0217] 162: Proximity sensor 163: Other sensors

[0218] 170: Camera subsystem 171: Optical sensor

[0219] 180: Communication Subsystem

[0220] 190: Audio subsystem

[0221] 191: Speaker 192: Microphone

[0222] 300: Evaluation server

[0223] 310: Communication Interface

[0224] 311: Wired communication port 312: Wireless circuit

[0225] 320: Memory

[0226] 321: Operating system 322: Communication module

[0227] 323: User Interface Module 324: Application

[0228] 330: I / O interface 340: Processor

[0229] [National Research and Development Project Supporting This Invention]

[0230] [Project ID] 1711171259

[0231] [Assignment Number] 2022R1F1A1065543

[0232] [Ministry Name] Ministry of Science and ICT

[0233] [Name of Project Management (Specialist) Institution] National Research Foundation of Korea

[0234] [Research Project Name] Individual Basic Research (Ministry of Science and ICT)

[0235] [Research Project Title] Optimal Treatment Decision System for Incidentally Detected Early Solid Tumors:

[0236] Predicting residual function after treatment based on a self-adaptive deep learning image segmentation model.

[0237] [Name of the project performing organization] Yonsei University

[0238] Research Period: June 1, 2022 - February 28, 2025

Claims

1. An evaluation method for renal tumors implemented by a processor, A step of receiving a medical image including a kidney and a kidney tumor; A step of generating three-dimensional image data from the above medical image; A step of segmenting regions based on the received three-dimensional image data using a prediction model learned to segment regions for tumor, substance, renal pelvis, and kidney using three-dimensional image data as input; A step of determining a first boundary surface between the tumor and the substance or a second boundary surface between the tumor and the substance based on the segmented region; A step of determining a first distance between the first boundary surface and the newspaper or a second distance between the second boundary surface and the newspaper, and A method for evaluating a renal tumor, comprising the step of performing an evaluation for a renal tumor based on the first distance or the second distance.

2. In paragraph 1, The step of dividing the above area is: A step of assigning a class number to each area of ​​the tumor, substance, renal pelvis and newspaper, and An evaluation method for a renal tumor, comprising a step of determining a region segmentation result by predicting the location and boundary of each region according to an assigned class number.

3. In paragraph 1, After the above dividing step, An evaluation method for a renal tumor, further comprising a step of determining a final segmentation result by applying majority voting to the region segmentation result.

4. In paragraph 1, After the above dividing step, A step of applying a color map including a color corresponding to each region to the region segmentation result to obtain a visualized 3D volume image, and A method for evaluating a renal tumor, further comprising the step of providing a visualized three-dimensional volume image.

5. In paragraph 1, The step of determining the first boundary surface or the second boundary surface is: A step of determining the first boundary surface and the second boundary surface based on the divided area, The step of determining the first distance or the second distance comprises: comprising a step of determining the first distance and the second distance, The steps for performing an evaluation of the above renal tumor are: A method for evaluating a renal tumor, comprising the step of performing an evaluation for a renal tumor based on the first distance and the second distance.

6. In paragraph 5, The step of determining the first distance and the second distance comprises: A step of determining a first distance between any point selected from the first boundary surface and the center of the newspaper, and An evaluation method for a renal tumor, comprising the step of determining a second distance between any point selected from the second boundary surface and the center of the newspaper.

7. In paragraph 5, The steps for performing an evaluation for the above renal tumor are: A step of reciprocally calculating the values ​​of the first distance and the second distance respectively; A step of determining an evaluation index for a renal tumor by adding the values ​​of the first distance and the second distance that have been reversed, and An evaluation method for a renal tumor, comprising a step of quantitatively evaluating the anatomical complexity of the tumor based on an evaluation index for the renal tumor.

8. In paragraph 7, The step of determining the first distance and the second distance comprises: A step of determining a plurality of first distances from each of a plurality of points selected from the first boundary surface to the center of the newspaper, and A step of determining a plurality of second distances from each of the plurality of points selected from the second boundary surface to the center of the newspaper, The step of determining the above renal tumor evaluation index is: A step of reciprocating each of the plurality of first distance values ​​produced; A step of summing the first distance values ​​of the plurality of reciprocal numbers; A step of reciprocating each of the plurality of second distance values ​​produced; a step of adding up the second distance values ​​of the reciprocal, and An evaluation method for a renal tumor, comprising the step of determining a renal tumor evaluation index by further adding the plurality of reciprocally quantified first distance values ​​and the plurality of reciprocally quantified second distance values.

9. In paragraph 1, The step of generating the above three-dimensional image data is: A method for evaluating a renal tumor, comprising the step of generating three-dimensional image data using three-dimensional rendering software.

10. In paragraph 1, The above medical images are, A method for evaluating a renal tumor, comprising at least one of CT imaging, MRI imaging, and ultrasound imaging.

11. A communication unit configured to receive medical images including kidneys and kidney tumors, and including a processor functionally connected to the above communication unit, The above processor, Generating three-dimensional image data from the above medical image, Using a prediction model learned to segment regions for tumor, substance, renal pelvis, and kidney using three-dimensional image data as input, regions are segmented based on the received three-dimensional image data, Based on the segmented area, the first boundary surface between the tumor and the substance or the second boundary surface between the tumor and the newspaper is determined, Determine the first distance between the first boundary surface and the newspaper or the second distance between the second boundary surface and the newspaper, A device for evaluating a renal tumor, configured to perform an evaluation of a renal tumor based on the first distance or the second distance.

12. In paragraph 11, The above processor, Assign a class number to each area of ​​the tumor, substance, renal pelvis and newspaper, A device for evaluating a renal tumor, further configured to determine a region segmentation result by predicting the location and boundary of each of the above regions according to the assigned class number.

13. In paragraph 11, The above processor, A device for evaluating renal tumors, further configured to determine the final segmentation result by applying majority voting to the region segmentation result.

14. In paragraph 11, The above processor, To obtain a visualized 3D volume image, a color map containing a color corresponding to each region is applied to the region segmentation result, A device for evaluating a renal tumor, further configured to provide the above visualized three-dimensional volume image.

15. In paragraph 11, The above processor, Determine the first boundary surface and the second boundary surface based on the divided area, Determine the first distance and the second distance, A device for evaluating a renal tumor, further configured to perform an evaluation of the renal tumor based on the first distance and the second distance.

16. In paragraph 15, The above processor, Determine a first distance between any point selected from the first boundary surface and the center of the newspaper, A device for evaluating a renal tumor, further configured to determine a second distance between any point selected from the second boundary surface and the center of the newspaper.

17. In paragraph 15, The above processor, The values ​​of the first distance and the second distance are each reciprocated, and the reciprocated values ​​of the first distance and the second distance are added to determine the renal tumor evaluation index. A device for evaluating a renal tumor, configured to quantitatively evaluate the anatomical complexity of the tumor based on the renal tumor evaluation index.

18. In paragraph 17, The above processor, Determine a plurality of first distances from each of the plurality of points selected from the first boundary surface to the center of the newspaper, Determine a plurality of second distances from each of the plurality of points selected from the second boundary surface to the center of the newspaper, Reciprocalize each of the plurality of first distance values ​​produced, Sum the first distance values ​​of the reciprocal plural, Reciprocalize each of the plurality of second distance values ​​produced, Add up the second distance values ​​of the reciprocal plural, A device for evaluating a renal tumor, further configured to determine a renal tumor evaluation index by further adding the plurality of reciprocally quantified first distance values ​​and the plurality of reciprocally quantified second distance values.

19. In paragraph 11, The above processor, A device for evaluating a renal tumor, further configured to generate the three-dimensional image data using three-dimensional rendering software.

20. In paragraph 11, The above medical images are, A device for evaluating renal tumors, comprising at least one of CT images, MRI images and ultrasound images.

21. An internal memory storing a prediction model learned to segment regions for tumors, renal parenchyma, renal pelvis, and kidney using medical images and three-dimensional image data including kidney and renal tumors as input. It is configured to access the above internal memory, Generating three-dimensional image data from the above medical image, Using the above prediction model, the three-dimensional image data is segmented into regions for tumor, substance, renal pelvis, and kidney, Based on the segmented area, the first boundary surface between the tumor and the substance or the second boundary surface between the tumor and the newspaper is determined, Determine the first distance between the first boundary surface and the newspaper or the second distance between the second boundary surface and the newspaper, A system for evaluating a renal tumor, comprising a processing unit configured to perform a renal tumor evaluation based on the first distance or the second distance.

Images