Apparatus for visualizing inferior alveolar nerve canal, operation method thereof, and recording medium

Abstract

An apparatus for visualizing an inferior alveolar nerve canal and an operation method thereof are disclosed. The operation method of the apparatus may comprise the operations of: setting a region of interest (ROI) on the basis of an inferior alveolar nerve canal detected from three-dimensional oral cavity data through an artificial intelligence model; generating, on the basis of Hounsfield unit (HU) values for pixels of sub-regions set for the region of interest, a histogram for each of the sub-regions in a two-dimensional cross-section including the region of interest; calculating entropy of each of the sub-regions on the basis of a probability distribution of the HU values calculated for each bin interval of the histogram; determining, on the basis of the calculated entropy, whether an inferior alveolar nerve canal is present in a pixel included in each of the sub-regions; and generating a three-dimensional heat map for the inferior alveolar nerve canal on the basis of whether the inferior alveolar nerve canal is present in the pixel included in each of the sub-regions.

Description

Device for visualizing the Hachijo nerve tube, method of operation thereof, and recording medium

[0001] The embodiments of the present disclosure relate to an apparatus for visualizing the Hachijo nerve tube, a method of operating the same, and a recording medium.

[0002] Conventionally, software was used to analyze a patient's 3D oral data, and an implant surgery plan was established by considering the anatomical structures identified in the analysis results. In particular, when implanting an implant in a patient's mandible, the inferior alveolar nerve canal area was identified in the 3D oral data, and an inferior alveolar nerve canal line was generated based on this to establish a surgical plan while predicting in advance the collision between the implant or other structures and the inferior alveolar nerve canal.

[0003] Existing software generated inferior alveolar nerve canal lines by connecting nerve canal region points entered by the user in 2D cross-sectional images generated based on 3D oral data. However, this method can result in the generation of inaccurate inferior alveolar nerve canal lines if the inferior alveolar nerve canal is not clearly visible in the 2D cross-sectional image, as accurate nerve canal region points are not entered by the user. Additionally, while the actual inferior alveolar nerve canal may not have a uniform thickness across its entire area, the lines generated by existing software were depicted with a uniform thickness, leading to a discrepancy with the actual shape.

[0004] The information described above may be provided as related art for the purpose of aiding understanding of the present disclosure. No claim or determination is made as to whether any of the foregoing may be applied as prior art related to the present disclosure.

[0005] The present disclosure may provide a method and apparatus for determining the inferior alveolar nerve canal by using HU (Hounsfield unit) values ​​extracted from three-dimensional oral data.

[0006] In addition, the present disclosure can provide a method and apparatus for visually distinguishing the upper and lower margins of the inferior alveolar nerve tube by generating a three-dimensional heatmap for the inferior alveolar nerve tube determined using HU values.

[0007] However, technical challenges are not limited to the technical challenges described above, and other technical challenges may exist.

[0008] A method of operation of a device according to one embodiment may include: setting a region of interest (ROI) based on the inferior alveolar nerve canal detected in three-dimensional oral data through an artificial intelligence model; generating a histogram for each of the detailed regions based on HU (Hounsfield Unit) values ​​for pixels of the detailed regions set for the region of interest in a two-dimensional cross-section including the region of interest; calculating the entropy of each of the detailed regions based on the probability distribution of the HU values ​​calculated for each bin interval of the histogram; determining whether a pixel included in each of the detailed regions is an inferior alveolar nerve canal based on the calculated entropy; and generating a three-dimensional heatmap for the inferior alveolar nerve canal based on whether a pixel included in each of the detailed regions is an inferior alveolar nerve canal.

[0009] The operation of determining whether the above-mentioned inferior neural tube is a pixel belonging to the above-mentioned detailed region is determined to be inside the above-mentioned inferior neural tube when the entropy calculated for the above-mentioned detailed region is within a preset first range; and the operation of determining the pixel belonging to the above-mentioned detailed region to be at least one of the upper or lower edge of the above-mentioned inferior neural tube when the entropy calculated for the above-mentioned detailed region is within a preset second range.

[0010] The above first range may have a lower entropy value than the above second range.

[0011] The method may further include: adjusting the bin interval of the histogram when the concordance rate between the region of the Hachijo neural tube determined based on the calculated entropy and the region of the Hachijo neural tube detected through the artificial intelligence model is less than a preset value; and recalculating the probability distribution and entropy of the HU value based on the adjusted bin interval of the histogram to re-determine whether the pixels included in each of the detailed regions are Hachijo neural tubes.

[0012] The operation of generating the above three-dimensional heatmap may include: an operation of generating a two-dimensional heatmap for the two-dimensional cross-section based on whether the pixels included in each of the detailed regions are Hachijo neural tubes; and an operation of aligning the generated two-dimensional heatmap to generate a three-dimensional heatmap.

[0013] The operation of setting the region of interest may include setting the region of interest as an area extended by a preset thickness compared to the region of the Hachijo neural tube detected through the artificial intelligence model.

[0014] The operation of normalizing the HU value based on the window width and window level of the imaging equipment used to extract the HU value may be further included.

[0015] The operation of generating the above three-dimensional heatmap may include an operation of displaying the areas corresponding to the upper and lower margins of the above-mentioned inferior alveolar nerve tube with gradient colors or visualizing them by blurring.

[0016] The above visualization operation may include an operation that displays a probability value of the pixel corresponding to the user operation being a Hachijo neural tube when the user operation is identified in the above-described gradient color or the above-described blurred area.

[0017] An apparatus according to one embodiment includes at least one processor comprising a processing circuit; and a memory comprising one or more storage media for storing instructions, wherein when the instructions are executed individually or collectively by the at least one processor, the apparatus may set a region of interest (ROI) based on the inferior alveolar nerve canal detected in three-dimensional oral data through an artificial intelligence model, generate a histogram for each of the detailed regions based on the Hounsfield Unit (HU) values ​​for pixels of the detailed regions set for the ROI in a two-dimensional cross-section including the ROI, calculate the entropy of each of the detailed regions based on the probability distribution of the HU values ​​calculated for each bin interval of the histogram, determine whether the pixels included in each of the detailed regions are the inferior alveolar nerve canal based on the calculated entropy, and generate a three-dimensional heatmap for the inferior alveolar nerve canal based on whether the pixels included in each of the detailed regions are the inferior alveolar nerve canal.

[0018] According to one embodiment of the present disclosure, the inferior alveolar nerve canal can be determined more accurately through analysis based on HU values ​​extracted from three-dimensional oral data.

[0019] In addition, according to one embodiment of the present disclosure, by generating a three-dimensional heatmap for the inferior alveolar nerve canal to visually distinguish the upper and lower margins of the inferior alveolar nerve canal, the user can intuitively determine the collision between the implant or other structures and the inferior alveolar nerve canal.

[0020] In relation to the description of the drawings, the same or similar reference numerals may be used for identical or similar components.

[0021] FIG. 1 is a drawing for illustrating a device for visualizing the Hachijo nerve tube according to one embodiment.

[0022] FIG. 2 is a diagram illustrating a method for visualizing the lower neural tube of a device according to one embodiment.

[0023] FIG. 3 is a diagram illustrating the configuration of three-dimensional oral data according to one embodiment.

[0024] FIG. 4 is a diagram illustrating an example of masking three-dimensional oral data using an artificial intelligence model according to one embodiment.

[0025] FIG. 5 is a diagram illustrating a method for setting a region of interest according to one embodiment.

[0026] FIG. 6 is a diagram illustrating the setting values ​​required for image normalization of HU values ​​according to one embodiment.

[0027] FIG. 7 is a diagram illustrating the organizational structure within the region of interest according to one embodiment.

[0028] FIG. 8 is a diagram showing the region of the Hachijo neural tube detected through an artificial intelligence model according to one embodiment and the region of the Hachijo neural tube determined based on entropy.

[0029] FIG. 9 is a diagram illustrating a method for generating a three-dimensional heatmap based on a two-dimensional heatmap according to one embodiment.

[0030] FIG. 10 is a diagram illustrating a graphic processing method for pixels corresponding to a Hachijo neural tube according to one embodiment.

[0031] FIG. 11 is a diagram illustrating a user operation method for checking the probability value of a pixel in an area that is displayed in a gradient color or blurred according to one embodiment being a Hachijo neural tube.

[0032] Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only and may be modified and implemented in various forms. Accordingly, actual implementations are not limited to the specific embodiments disclosed, and the scope of this specification includes modifications, equivalents, or substitutions included in the technical concept described by the embodiments.

[0033] In this document, each of the following phrases may include any one of the items listed together in the corresponding phrase, or any combination of A, B, and C, or any combination of all of them. Terms such as "A or B," "at least one of A and B," "at least one of A, B, and C," "at least one of A, B, or C," and "a combination of one or more of A, B, and C" may be used to describe various components, but these terms should be interpreted solely for the purpose of distinguishing one component from another. For example, the first component may be named the second component, and similarly, the second component may also be named the first component.

[0034] When it is stated that a component is "connected" to another component, it should be understood that it may be directly connected to or coupled with that other component, or that there may be other components in between.

[0035] The singular expression includes the plural expression unless the context clearly indicates otherwise. In this specification, terms such as "comprising" or "having" are intended to specify the existence of the described features, numbers, steps, actions, components, parts, or combinations thereof, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0036] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this specification.

[0037]

[0038] Hereinafter, embodiments will be described in detail with reference to the attached drawings. In the description with reference to the attached drawings, identical components are given the same reference numeral regardless of the drawing number, and redundant descriptions thereof will be omitted.

[0039]

[0040] FIG. 1 is a drawing for illustrating a device for visualizing the Hachijo nerve tube according to one embodiment.

[0041] Referring to FIG. 1, the device (100) may include one or more processors (110) and a memory (120) that loads or stores a computer program (130) executed by the processors (110). The processors (110) and the memory (120) may be connected to each other via a communication link (e.g., a bus) (140). Optionally, the device (100) may further include a transceiver (150), and the transceiver (150) may be used for data exchange, such as the transmission and / or reception of data between the device (100) and another device. The components included in the device (100) of FIG. 1 are merely examples, and a person skilled in the art to which this disclosure pertains will understand that other general-purpose components may be included in addition to the components shown in FIG. 1.

[0042] ​​The processor (110) can control the overall operation of each component of the device (100). The processor (110) may be implemented as a circuitry (e.g., a processing circuit) such as a system on chip (SoC) or an integrated circuit (IC). The processor (110) may include one or more processors. For example, the processor (110) may include a combination of one or more processors such as a central processing unit (CPU), a micro processor unit (MPU), a micro controller unit (MCU), a graphic processing unit (GPU), a neural processing unit (NPU), a digital signal processor (DSP), an application processor (AP), a communication processor (CP), or any form of processor well known in the art of the present disclosure. Additionally, the processor (110) can perform operations on at least one application or computer program (130) for executing a method / operation according to various embodiments of the present disclosure.

[0043] The memory (120) can store one or more combinations of various data, instructions, and information used by a component (e.g., processor (110)) included in the device (100). The memory (120) may include volatile memory and / or nonvolatile memory.

[0044] A computer program (130) may include one or more actions in which methods / actions according to various embodiments of the present disclosure are implemented, and may be stored in memory (120) in the form of software. Here, the action may correspond to instructions implemented in the computer program (130). For example, a computer program (130) may include instructions to perform the following operations: setting a region of interest (ROI) based on the inferior alveolar nerve canal detected in three-dimensional oral data through an artificial intelligence model; generating a histogram for each of the detailed regions based on the HU (hounsfield unit) values ​​for the pixels of the detailed regions set for the region of interest in a two-dimensional cross-section including the region of interest; calculating the entropy of each of the detailed regions based on the probability distribution of the HU values ​​calculated for each bin interval of the histogram; determining whether the pixels included in each of the detailed regions are the inferior alveolar nerve canal based on the calculated entropy; and generating a three-dimensional heatmap of the inferior alveolar nerve canal based on whether the pixels included in each of the detailed regions are the inferior alveolar nerve canal.

[0045] When a computer program (130) is loaded into memory (120), the processor (110) can perform methods / operations according to various embodiments of the present disclosure by executing a plurality of operations to implement the computer program (130).

[0046] A communication link (140) may include a path for transmitting at least one of various data, commands, and information between components included in the device (100). The communication link (140) may be, for example, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, but the types of such buses are merely examples and are not limited to the above examples. For example, in FIG. 1, the bus is represented as a single line for convenience of explanation, but in reality, multiple buses or various types of buses may be included.

[0047] The execution screen of a computer program (130) can be displayed through a display (160). In the case of FIG. 1, the display (160) is depicted as a separate device connected to the device (100), but in the case of a device (100) such as a terminal that a user can carry, such as a smartphone or tablet, the display (160) can be a component of the device (100). The screen displayed on the display (160) may be the result of the execution of the program or before information is input into the program.

[0048]

[0049] FIG. 2 is a diagram illustrating a method for visualizing the lower alveolar nerve tube of a device according to one embodiment. In one embodiment, at least one of the operations of FIG. 2 may be performed simultaneously or in parallel with other operations, and the order between the operations may be changed. Additionally, at least one of the operations may be omitted, and other operations may be performed additionally. The operations illustrated in FIG. 2 may be performed by a processor (e.g., processor (110) of FIG. 1) of a device (e.g., device (100) of FIG. 1).

[0050] ​In operation (210), the processor can set a region of interest based on the inferior alveolar nerve canal detected in the three-dimensional oral data through an artificial intelligence model. To do this, the processor can receive three-dimensional oral data (e.g., CT image data) of the patient's oral cavity from an external device (e.g., computed tomography, CT) through a transceiver (e.g., transceiver (150) of FIG. 1).

[0051] Referring to FIG. 3, the three-dimensional oral data may include a two-dimensional cross-section in the direction of a coronal view looking at the patient's head from the front (hereinafter referred to as a coronal cross-section) (310), a two-dimensional cross-section in the direction of a sagittal view looking at the patient's head from the side (hereinafter referred to as a sagittal cross-section) (320), a two-dimensional cross-section in the direction of an axial view looking down at the patient's head from the top (hereinafter referred to as an axial cross-section) (330), and a screen (340) that renders the patient's head in three dimensions. However, the configuration and arrangement of such three-dimensional oral data are merely examples and are not limited to the above examples.

[0052] The processor can input the received 3D oral data into an artificial intelligence model to detect major structures including the inferior alveolar nerve canal. More specifically, the artificial intelligence model of the present disclosure can receive 3D oral data as input, analyze it to distinguish major structures (e.g., mandibular bone, inferior alveolar nerve canal), and output a masked result based on this.

[0053] For example, as shown in FIG. 4, the artificial intelligence model can distinguish the mandibular bone (410) and the inferior alveolar nerve canal (420) in the input three-dimensional oral data, and can display the distinguished mandibular bone (410) and inferior alveolar nerve canal (420) by masking them with different colors. However, such a masking method is merely one example and is not limited to the above example. For example, the artificial intelligence model may mask the mandibular bone and the inferior alveolar nerve canal together, or mask them in a form other than color.

[0054] According to one embodiment, 3D oral data masked through an artificial intelligence model can be used to train the artificial intelligence model. For example, the artificial intelligence model may be one of a U-net, a convolutional neural network (CNN), a recurrent neural network (RNN), a deep belief network (DBN), and a restricted Boltzmann machine (RBM), but the types of such artificial intelligence models are merely examples and are not limited to the above examples.

[0055] According to one embodiment, the processor can set an area of ​​interest that is extended by a preset thickness from the area of ​​the inferior alveolar nerve canal detected in three-dimensional oral data through an artificial intelligence model. For example, the processor can set an area (520) that is extended by about 2 mm from the upper border and lower border based on the area (510) of the inferior alveolar nerve canal detected through the artificial intelligence model as shown in FIG. 5 as an area of ​​interest. If the area (510) of the inferior alveolar nerve canal detected through the artificial intelligence model is a cylinder with a diameter of about 3 mm, the processor can set a cylinder with a diameter of about 7 mm containing said cylinder as an area of ​​interest.

[0056] In operation (220), the processor may generate a histogram for each of the sub-regions based on the HU (hounsfield unit) values ​​for pixels of the sub-regions set within the region of interest in at least one two-dimensional cross-section including the region of interest. For example, the processor may divide the region of interest included in the coronal cross-section, sagittal cross-section, and axial cross-section into a plurality of sub-regions, and then generate a histogram of HU values ​​representing the radiation absorption density extracted for each sub-region.

[0057] However, the HU value for pixels representing the same structure may vary depending on the type of imaging equipment used to extract the HU value. Therefore, the processor can perform image normalization to utilize the extracted HU value. Through this, the processor can reduce noise and provide consistent analysis data by adjusting the extracted HU value to the same resolution.

[0058] According to one embodiment, the processor can normalize the extracted HU value based on the settings of the imaging equipment used to extract the HU value as shown in FIG. 6 (e.g., window width and window level). For example, the processor can adjust the HU value based on the settings of the imaging equipment to provide reliable data in an analysis process based on an artificial intelligence model.

[0059] The processor can analyze the distribution of the image-normalized HU values ​​in this way and generate a histogram showing the range of HU values ​​and the corresponding frequency. Referring to FIG. 7, within the region of interest, there may be compact bone (750) corresponding to the inferior alveolar nerve canal (710) having a relatively low HU value, the upper margin (720) and lower margin (730) of the inferior alveolar nerve canal (710), cancellous bone (740), and compact bone corresponding to the mandibular bone.

[0060] According to one embodiment, when creating an artificial intelligence model for detecting Hachijo neural tubes, the operator may expect that two lines are connected in the empty space (760) within the region of interest and mask the area.

[0061] In contrast, since the processor must visualize the empty space (760) as is, when detecting pixels in the area and calculating entropy values ​​to visualize it, the empty space (760) can be visualized by displaying it in a gradient color or by blurring it so that it is not recognized as a Hachijo neural tube. Through this, the processor can require user attention to the area and can improve the reliability of the Hachijo neural tube detection result.

[0062] In addition, if user interaction (e.g., mouse over) is identified in an area displayed in a gradient color or blurred, the processor may display and provide to the user a probability value that the pixel corresponding to the user interaction is a Hachijo neural tube.

[0063] For example, on average, air has a HU value of about -1000, fat has a HU value between about -100 and -50, soft tissue has a HU value between about 20 and 60, bone has a HU value between about 700 and 3000, and the inferior alveolar nerve tube can have a HU value between about -100 and 100.

[0064] Therefore, the processor can divide the range of HU values ​​of the Hachijo neural tube, approximately -100 to 100, into pre-set bin intervals. For example, by dividing the range of HU values ​​of the Hachijo neural tube, approximately -100 to 100, into bin intervals of 20 HU, the processor can set a total of 10 bin intervals, such as Bin 1: [-100, -80], Bin 2: [-80, -60], Bin 3: [-60, -40], …, Bin 10: [80, 100]. The processor can generate a histogram for the detailed region by calculating the number of pixels belonging to each bin interval based on the set bin intervals.

[0065] In operation (230), the processor can calculate the entropy of each of the sub-regions based on the probability distribution of HU values ​​calculated for each bin interval of the histogram for the sub-regions. More specifically, the processor can calculate the probability distribution of each bin interval as shown in Equation 1 below by dividing the number of pixels belonging to each bin interval in the histogram for the sub-regions by the total number of pixels within the sub-regions.

[0066]

[0067] for example, The number of pixels belonging to is 30, and The number of pixels belonging to is 50, and The number of pixels belonging to is 25, and,..., If the number of pixels belonging to is 60 and the total number of pixels within the detailed area is 500, 0.06, 0.1, 0.05,..., Each can be calculated as 0.12.

[0068] The processor, based on the probability distribution calculated in this way, the entropy of each detailed region It can be calculated using the following mathematical formula 2.

[0069]

[0070] Here, N represents the total number of bin intervals. Entropy It is an indicator that measures the diversity or uncertainty of data and can show how diversely pixel values ​​in a detailed area are distributed.

[0071] In operation (240), the processor can determine whether the pixels contained in each of the sub-regions are Hachijo neural tubes based on the calculated entropy. The entropy value is an indicator of the diversity and complexity of pixels within the sub-region and can be used to distinguish the tissue structure within the sub-region.

[0072] More specifically, a higher entropy value indicates that pixel values ​​in a detailed area can have a wide range, which means there is significant variation between pixels; thus, it may indicate that the detailed area possesses a complex tissue structure (e.g., bone, complex soft tissue). Conversely, a lower entropy value indicates that pixel values ​​in a detailed area can have a uniform range; this means there is minimal variation between pixels, which may indicate that the detailed area possesses a simple tissue structure (e.g., the interior of the inferior alveolar nerve canal).

[0073] For example, if the entropy value of a sub-region is 0.0 or higher and less than 0.7, the processor may determine that the pixels contained in the sub-region are inside the inferior alveolar neural canal. Additionally, if the entropy value of a sub-region is 0.7 or higher and less than 1.5, the processor may determine that the pixels contained in the sub-region are at least one of the superior or inferior margin of the inferior alveolar neural canal. Or, if the entropy value of a sub-region is 1.5 or higher, the processor may determine that the pixels contained in the sub-region are bone or complex soft tissue.

[0074] However, the entropy distinction value for determining the structure of a pixel included in such a detailed area as one of the different organizational structures is merely one example and is not limited to the above example.

[0075] In operation (250), the processor compares the match rate between the region of the Hachijo neural tube determined based on the calculated entropy and the region of the Hachijo neural tube detected by the artificial intelligence model with a preset value, and if the match rate is less than the preset value, the bin spacing of the histogram can be adjusted in operation (260). Here, the match rate can be calculated as shown in Equation 3 below by comparing the number of overlapping pixels between the two regions with the total number of pixels of the region detected by the artificial intelligence model.

[0076]

[0077] For example, if any of the match rates calculated for each of the coronal section, sagittal section, and axial section is less than a value specified by the user (e.g., 95%), the processor can adjust the bin interval for generating a histogram for the detailed region.

[0078] FIG. 8 is a diagram showing a region (810) of a lower neural tube detected by an artificial intelligence model according to one embodiment and a region (820) of a lower neural tube determined based on entropy. Referring to FIG. 8, if the agreement rate between the region (810) of a lower neural tube detected by an artificial intelligence model and the region (820) of a lower neural tube determined based on entropy is less than 95%, the processor may reduce or expand the bin interval for generating a histogram for a detailed region.

[0079] The setting of the bin intervals of such a histogram can have a significant impact on the accuracy of HU value analysis. For example, if the bin intervals of the histogram are too wide, detailed changes in HU values ​​may not be captured, leading to inaccurate results in entropy calculation; conversely, if the bin intervals are too narrow, the entropy values ​​may fluctuate unstably due to the influence of noise.

[0080] According to one embodiment, if the matching rate between the region of the Hachijo neural tube (810) detected through an artificial intelligence model and the region of the Hachijo neural tube (820) determined based on entropy is less than a preset value, the processor may reduce or expand the bin interval of the initially set histogram and then re-detect the Hachijo neural tube based on the entropy based on the HU value. At this time, the processor may reduce or expand the bin interval of the histogram according to a preset ratio (e.g., 50%).

[0081] Alternatively, the processor may determine the direction of adjustment for the bin spacing of the histogram based on the difference between the match rate between the region of the Hachijo neural tube (810) detected by the artificial intelligence model and the region of the Hachijo neural tube (820) determined based on entropy and a preset value. For example, the processor may reduce the bin spacing as the difference between the match rate between the region of the Hachijo neural tube (810) detected by the artificial intelligence model and the region of the Hachijo neural tube (820) determined based on entropy and the preset value is greater, and expand the bin spacing as the difference between the match rate between the region of the Hachijo neural tube (810) detected by the artificial intelligence model and the region of the Hachijo neural tube (820) determined based on entropy and the preset value is smaller.

[0082] The processor can re-detect the Hachijo neural tube region through the bin intervals adjusted in this way, and then re-compare the detection result with the detection result of the artificial intelligence model. The processor can repeatedly perform this process until the match rate between the region of the Hachijo neural tube (810) detected by the artificial intelligence model and the region of the Hachijo neural tube (820) determined based on entropy becomes greater than or equal to a preset value.

[0083] In operation (270), the processor can generate a three-dimensional heatmap of the Hachijo neural tube based on whether the pixels included in each of the detailed regions are Hachijo neural tubes. More specifically, the processor can generate a two-dimensional heatmap based on whether the pixels included in each of the detailed regions are Hachijo neural tubes in each two-dimensional cross-section, and generate a three-dimensional heatmap by aligning and integrating the generated two-dimensional heatmaps.

[0084] For example, as shown in FIG. 9 (a), (b), and (c), if the entropy value for a detailed area of ​​each 2D cross-section is in a first range (e.g., 0.0 or more and less than 0.7), the processor may display the pixels contained in that detailed area in a color set by the user (e.g., red) corresponding to the inside of the inferior alveolar neural tube. Additionally, if the entropy value for a detailed area of ​​each 2D cross-section is in a second range (e.g., 0.7 or more and less than 1.5), the processor may display the pixels contained in that detailed area in a color set by the user (e.g., green) corresponding to the upper or lower edge of the inferior alveolar neural tube. In this way, the processor can generate a 3D heatmap by aligning consecutive 2D heatmaps as shown in FIG. 9 (d) based on the 2D heatmap generated for each 2D cross-section.

[0085] At this time, the processor can induce user attention to the corresponding area by visualizing the pixels in the area by displaying them in gradient colors according to a probability distribution as shown in Fig. 10 or by blurring them, in order to provide clearer information about the pixels corresponding to the upper or lower edge of the Hachijo neural tube.

[0086] Alternatively, if a user operation (e.g., mouse over) (1110) is identified in an area that is displayed in a gradient color according to a probability distribution as in FIG. 11 or blurred, the processor may display and provide to the user a probability value (1120) that the pixel corresponding to the user operation is a Hachijo neural tube.

[0087]

[0088] The hardware device described above may be configured to operate as one or more software modules to perform the operation of the embodiment, and vice versa.

[0089] ​Although the embodiments have been described above with reference to the limited drawings, those skilled in the art can apply various technical modifications and variations based thereon. For example, suitable results may be achieved even if the described techniques are performed in a different order than described, and / or if the components of the described system, structure, device, circuit, etc. are combined or assembled in a form different from described, or replaced or substituted by other components or equivalents.

[0090] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims set forth below.

Claims

1. In a method of operation for a device for visualizing the Hachijo neural tube, The operation of setting a region of interest (ROI) based on the inferior alveolar nerve canal detected in 3D oral data through an artificial intelligence model; The operation of generating a histogram for each of the detailed regions based on HU (Hounsfield Unit) values ​​for pixels of the detailed regions set for the region of interest in a two-dimensional cross-section including the region of interest; The operation of calculating the entropy of each of the detailed regions based on the probability distribution of the HU values ​​calculated for each bin interval of the histogram; An operation to determine whether a pixel included in each of the detailed regions is a Hachijo neural tube based on the above-calculated entropy; and The operation of generating a 3D heatmap for the Hachijo neural tube based on whether the pixels included in each of the detailed regions are Hachijo neural tubes. A method of operation including 2. In Paragraph 1, The operation of determining whether the above-mentioned Hachijo nerve tube exists is, If the entropy calculated for the above detailed region is within a preset first range, the operation of determining a pixel belonging to the above detailed region as being inside the above Hachijo neural tube; and If the entropy calculated for the above detailed region is within a preset second range, the operation of determining a pixel belonging to the above detailed region as at least one of the upper or lower margin of the above Hachijo neural tube A method of operation including 3. In Paragraph 2, The above first range is, having an entropy value lower than the second range mentioned above, Method of operation.

4. In Paragraph 1, An operation to adjust the bin intervals of the histogram when the concordance rate between the region of the Hachijo neural tube determined based on the above-calculated entropy and the region of the Hachijo neural tube detected through the artificial intelligence model is less than a preset value; and Operation of recalculating the probability distribution and entropy of the HU values ​​based on the bin intervals of the adjusted histogram to re-determine whether the pixels included in each of the detailed regions are Hachijo neural tubes. A method of operation that further includes 5. In Paragraph 1, The operation of generating the above 3D heatmap is, The operation of generating a two-dimensional heatmap for the two-dimensional cross-section based on whether the pixels included in each of the above detailed regions are Hachijo neural tubes; and The operation of generating a 3D heatmap by aligning the above-mentioned 2D heatmap. A method of operation including 6. In Paragraph 1, The operation of setting the above region of interest is, The operation of setting the region of interest as an area extended by a preset thickness compared to the area of ​​the Hachijo neural tube detected by the above artificial intelligence model. A method of operation including 7. In Paragraph 1, An operation to normalize the HU value based on the window width and window level of the imaging equipment used to extract the HU value. A method of operation that further includes 8. In Paragraph 1, The operation of generating the above 3D heatmap is, The operation of visualizing the areas corresponding to the upper and lower margins of the above-mentioned Hachijo nerve tube by displaying them in gradient colors or blurring them. A method of operation including 9. In Paragraph 8, The above visualization action is, Operation of displaying a probability value that the pixel corresponding to the user operation is a Hachijo neural tube when the user operation is identified in the area displayed by the above gradient color or the above blurred area. A method of operation including 10. A computer-readable recording medium storing a computer program that executes the method of paragraph 1.

11. In a device for visualizing the Hachijo neural tube, At least one processor including a processing circuit; and Memory comprising one or more storage media that store instructions Includes, When the above commands are executed individually or collectively by the at least one processor, the device, Based on the inferior alveolar nerve canal detected in 3D oral data using an artificial intelligence model, a region of interest (ROI) is established, and In a two-dimensional cross-section including the region of interest, a histogram for each of the detailed regions is generated based on the HU (Hounsfield Unit) values ​​for the pixels of the detailed regions set for the region of interest, and Calculate the entropy of each of the detailed regions based on the probability distribution of the HU values ​​calculated for each bin interval of the above histogram, and Based on the above calculated entropy, determine whether the pixels included in each of the above detailed regions are Hachijo neural tubes, and Generating a 3D heatmap for the Hachijo neural tube based on whether the pixels included in each of the above detailed regions are Hachijo neural tubes, device.

Images