Method and system for reconstructing a three-dimensional model of the external auditory canal based on computed tomography

Abstract

The present invention provides a method and system for reconstructing a three-dimensional model of the external auditory canal based on computed tomography (CT). The method comprises the following steps: 1) performing a CT scan on a subject's head to obtain a plurality of CT images; 2) determining the external auditory canal display region in the plurality of CT images; 3) performing edge extraction on each external auditory canal display region to obtain a chain code of the external auditory canal contour; 4) performing two-dimensional curve fitting on the extracted chain code using an elliptical Fourier transform to obtain a two-dimensional continuous curve model of the external auditory canal in each layer of the CT image; and 5) splicing the two-dimensional continuous curve models in all CT images to obtain a three-dimensional model of the external auditory canal. The present invention does not require the injection of foreign matter into a real ear canal, does not cause adverse reactions to the subject, and can accurately, efficiently, and quickly obtain a three-dimensional model of the external auditory canal.

Description

Technical Field

[0001] The present invention relates to the field of three-dimensional reconstruction technology and the field of three-dimensional structural reconstruction of the external auditory canal, and in particular to a method and system for reconstructing a three-dimensional model of the external auditory canal based on computed tomography, which can be used for hearing aid fitting, personalized appearance and structure design of in-ear headphones, and acoustic simulation. Background Art

[0002] In-ear headphones, a common communication device, are increasingly popular with consumers. However, due to the standardized production process, the experience of wearing in-ear headphones varies from person to person. Long-term wear can cause discomfort, primarily due to the mismatch between the headphone's shape and the ear canal. During the hearing aid fitting process, the hearing aid's shape also needs to be matched to the three-dimensional structure of the external auditory canal to minimize discomfort.

[0003] The complex structure of the external auditory canal, characterized by its tendency to grow into the skull, makes it difficult to directly measure using various laser scanning devices. Furthermore, the external auditory canal is connected to the inner ear through the eardrum, making it highly sensitive, making it difficult to obtain a 3D model of a real person's external auditory canal. Safely, accurately, and efficiently obtaining a 3D model of a real person's external auditory canal presents multiple technical challenges.

[0004] Stinson of the National Medical Center of Canada and Lawton et al. of the University of Southampton in the UK first measured and analyzed a three-dimensional model of the human ear canal in 1989 (MR Stinson, BW Lawton., Specification of the geometry of the human ear canal for the prediction of sound-pressure level distribution [J]. Journal of the Acoustical Society of America, 1989, 85(6): 2492-2503.). They injected silicone rubber into the ear canal and, after the rubber was formed, measured the three-dimensional coordinates of 1,000 sampling points on the surface to establish the ear canal model. This method is time-consuming and may have certain side effects on the subjects, making it difficult to promote and apply in large quantities. Therefore, during their research, they only collected ear canal model data from 15 subjects. In 2018, Stefan's research group at the University of Copenhagen proposed using magnetic resonance imaging (MRI) to create three-dimensional models of human ear canals (S. Darkner, S. Sommer, A. Schuhmacher, et al., An Average of the Human Ear Canal: Recovering Acoustical Properties via Shape Analysis. https: / / arxiv.org / abs / 1811.03848v1, 2018.). Due to the working principle of MRI, MRI scans require the filling of the human ear canal with a contrast agent. Stefan's group chose to use rapeseed oil as a contrast agent, which was difficult to implement. MRI scans also have low spatial resolution, requiring powerful image processing algorithms. They also take a long time to scan, are prone to artifacts, and are expensive, all of which hinder the widespread application of MRI.

[0005] Therefore, existing methods for reconstructing ear canal models all require injecting contrast agents or silicone rubber into the subject's ear canal, which is difficult to implement and can easily cause discomfort to the subject. In severe cases, side effects such as allergies may occur. In addition, powerful image algorithms are required for subsequent reconstruction, which results in a large amount of computational workload. At the same time, existing methods for reconstructing ear canal models also have the disadvantages of long scanning time, low scanning spatial resolution, and high price, which are not conducive to large-scale promotion and application. Summary of the Invention

[0006] The purpose of the present invention is to overcome the shortcomings of existing auditory canal model reconstruction methods, such as easy discomfort to the subjects, large computational workload, long scanning time, low scanning spatial resolution and high price, and thus provide a method and system for reconstructing a three-dimensional model of the external auditory canal based on computed tomography.

[0007] To solve the above technical problems, the technical solution of the present invention provides a method for reconstructing a three-dimensional model of the external auditory canal based on computed tomography, comprising the following steps:

[0008] Step 1) performing a computed tomography scan on the subject's head to obtain a plurality of computed tomography images;

[0009] Step 2) determining the external auditory canal display area in the plurality of computed tomography images;

[0010] Step 3) performing edge extraction on each external auditory canal display area to obtain a chain code of the external auditory canal contour;

[0011] Step 4) performing two-dimensional curve fitting on the extracted chain code using elliptical Fourier transform to obtain a two-dimensional continuous curve model of the external auditory canal in each layer of the computed tomography image;

[0012] Step 5) splicing the two-dimensional continuous curve models in all the computed tomography images to obtain a three-dimensional model of the external auditory canal.

[0013] As an improvement to the above method, step 1) specifically includes performing a CT scan of the subject's head along the z-axis, scanning one CT image slice at intervals Δz to obtain a plurality of CT image slices. The number of slices obtained is related to the z-axis length of the scanned head. For example, if the z-axis length of the head is 20 cm and the scanning interval is 0.1 cm, the number of slices obtained is 200. If the scanning interval is 0.05 cm, the number of slices obtained is 400.

[0014] As an improvement to the above method, step 3) specifically includes: converting the plurality of computed tomography images into grayscale images, performing edge extraction on each external auditory canal display area using the Canny edge extraction method, and establishing a vector segment mapping of adjacent pixels starting from any pixel point along the edge of the external auditory canal contour to obtain a chain code V of the external auditory canal contour:

[0015] V=α1α2α3…α K (1)

[0016] Among them, α1 represents the first vector segment, α2 represents the second vector segment, α3 represents the third vector segment, and α KRepresents the Kth vector line segment, where K represents the total number of vector line segments of this continuous curve.

[0017] As an improvement to the above method, the method of performing two-dimensional curve fitting on the extracted chain code using elliptical Fourier transform to obtain a two-dimensional continuous curve model of the external auditory canal in a layer of computed tomography image specifically includes:

[0018] Step 4-1) Based on the chain code V of the external auditory canal contour, the speed of moving along the vector segment is the unit speed, and the vector segment α passing through the i-th vector segment is calculated. i The time Δt i :

[0019]

[0020] Among them, α i represents the i-th vector segment, α i ∈{0,1,2,K,7},

[0021] Step 4-2) Calculate the displacement Δx of the i-th vector segment in the x-axis direction i and the displacement Δy in the y-axis direction i ;in,

[0022] Δx i =sgn(6-α i )sgn(2-α i ) (3)

[0023] Δy i =sgn(4-α i )sgn(α i ) (4)

[0024] Step 4-3) Get the time t taken to pass through the first p vector segments p and the total time T spent passing through all vector segments, where

[0025]

[0026] Wherein, K represents the total number of vector line segments of the two-dimensional continuous curve;

[0027] Step 4-4) Get the coordinate x in the x-axis direction after passing through the first p vector segments p and the coordinate y in the y-axis direction p ;in,

[0028]

[0029] Step 4-5) Calculate the DC component A0 in the x-axis direction and the DC component C0 in the y-axis direction; wherein,

[0030]

[0031]

[0032] Among them, x p-1 is the coordinate in the x-axis direction after passing through the first p-1 vector segments; y p-1 is the coordinate in the y-axis direction after passing through the first p-1 vector segments;

[0033] Step 4-6) Calculate the nth order first ellipticity coefficient a in the x-axis direction n and the second ellipticity coefficient b n , and the nth order third ellipticity coefficient c in the y-axis direction n and the third ellipticity coefficient d n ,in,

[0034]

[0035] Among them, t p-1 is the time it takes to pass through the first p-1 vector segments;

[0036] Step 4-7) Obtain the coordinates {x(t), y(t)} of the two-dimensional continuous curve model of the external auditory canal in the current layer of the computed tomography image, where

[0037]

[0038] Where N is the total order of two-dimensional curve fitting.

[0039] As an improvement to the above method, step 5) specifically includes: splicing the two-dimensional continuous curve models of all computed tomography images along the fault direction to obtain a three-dimensional model of the external auditory canal, wherein the fault direction is the Z-axis direction, and the value of the j-th layer of computed tomography image in the z-axis direction is Z j .

[0040] To achieve another object of the present invention, the present invention provides a system for reconstructing a three-dimensional model of the external auditory canal based on computed tomography, the system comprising: a computed tomography scanner, a two-dimensional curve fitting module and a three-dimensional model reconstruction module, wherein:

[0041] The computerized tomography scanner is used to perform a computerized tomography scan on the subject's head to obtain a plurality of computerized tomography images;

[0042] The two-dimensional curve fitting module is used to determine the external auditory canal display area in the plurality of computed tomography images; to extract the edge of each external auditory canal display area to obtain a chain code of the external auditory canal contour; and to perform two-dimensional curve fitting on the extracted chain code using elliptical Fourier transform to obtain a two-dimensional continuous curve model of the external auditory canal in each layer of the computed tomography image;

[0043] The three-dimensional model reconstruction module is used to splice the two-dimensional continuous curve models in all the computer tomography images to obtain a three-dimensional model of the external auditory canal.

[0044] The advantages of the present invention are that the method for reconstructing a three-dimensional model of the external auditory canal based on computed tomography provided by the present invention firstly obtains a plurality of CT images based on the principle that x-rays will be absorbed and attenuated by tissue after passing through human tissue by CT, and extracts edges of the images to obtain a chain code of a closed curve of the external auditory canal contour; secondly, the closed curve is fitted using an elliptical Fourier descriptor to obtain a mathematical expression of a two-dimensional curve of the external auditory canal projection of the current slice; finally, the two-dimensional curves of the external auditory canal projection extracted from all CT images are spliced ​​to obtain a three-dimensional model of the external auditory canal. The present invention combines the advantages of fast scanning speed and high spatial resolution of CT scanners, does not require the injection of foreign matter into the ear canal of a real person, and will not produce adverse reactions to the subject, and can accurately, efficiently and quickly obtain a three-dimensional model of the external auditory canal. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A flowchart of a method for reconstructing a three-dimensional model of the external auditory canal based on computed tomography provided by the present invention;

[0046] Figure 2 Schematic diagram of computed tomography image;

[0047] Figure 3 This is a schematic diagram of the display area of ​​the external auditory canal in a computed tomography image;

[0048] Figure 4 is the binary image extracted from the edge of the external auditory canal;

[0049] Figure 5 It is a schematic diagram of vector segment chain coding;

[0050] Figure 6 It is a two-dimensional continuous curve model reconstructed by the elliptic Fourier descriptor;

[0051] Figure 7 This is a schematic diagram of the reconstructed three-dimensional model of the external auditory canal. DETAILED DESCRIPTION

[0052] The technical solution provided by the present invention is further illustrated below with reference to embodiments.

[0053] Computed tomography (CT) uses the principle that x-rays are absorbed and attenuated by tissue after passing through it. Tissues of varying densities absorb x-rays differently. By emitting x-rays at different angles of incidence and processing them with a computer, the density distribution of the object within a specific section can be determined. Because CT scans do not require the insertion of foreign objects into the subject, they offer rapid scanning speeds and high spatial resolution.

[0054] The purpose of the present invention is to provide a fast, efficient and safe external auditory canal three-dimensional structure reconstruction method and system using computed tomography technology. The present invention is described in detail below with reference to the accompanying drawings.

[0055] Example 1

[0056] The method for reconstructing a three-dimensional model of the external auditory canal based on computed tomography provided by the present invention comprises the following steps:

[0057] Step 1) Computed tomography experiment of subjects:

[0058] The first step in implementing this patent is to conduct a computerized tomography experiment. Figure 2 A CT scan image of a subject is given, with 350 scan layers. The figure shows the scan image of the 121st layer.

[0059] Step 2) Determine the external auditory canal area:

[0060] Due to the difference in density between the external auditory canal cavity and the surrounding human tissue, the two-dimensional display area of ​​the external auditory canal in the current scanning layer can be determined. Figure 3 Shown in the grey box.

[0061] Step 3) Extract the edges of the image within the auditory canal box:

[0062] Since the air density is relatively uniform, the image color coding in the external auditory canal display area tends to be consistent; however, in the external auditory canal wall, the density of human tissue will suddenly change relative to the cavity, so the image color coding will suddenly change at the external auditory canal contour. The edge extraction algorithm can be used to extract the external auditory canal contour. Since the edge extraction algorithm only supports grayscale images, the original scanned image needs to be converted into a grayscale image in advance for processing. Figure 4 The binary image obtained by using the Canny edge detection algorithm is given.

[0063] Step 4) Starting from any pixel point along the edge of the external auditory canal, a vector segment mapping of adjacent pixel points is established as follows: Figure 6 As shown, the chain coding V of the external auditory canal contour can be obtained:

[0064] V=α1α2α3Kα K (1)

[0065] where α i is the i-th vector segment symbol, α i ∈{0,1,2,K,7}, K represents the total number of vector segments of this continuous curve.

[0066] Assuming that the speed of movement along the vector segment is unit speed, the time Δt taken to pass the i-th vector segment is i for

[0067]

[0068] The i-th vector segment is displaced Δx in the x-axis direction i :

[0069] Δx i =sgn(6-α i )sgn(2-α i ) (3)

[0070] The i-th vector segment is displaced Δy in the y-axis direction i :

[0071] Δy i =sgn(4-α i )sgn(α i ) (4)

[0072] The time t it takes to pass through the first p vector segments p for:

[0073]

[0074] The total time T spent passing through all vector segments is:

[0075]

[0076] The coordinate x in the x-axis direction passes through the first p vector segments p for:

[0077]

[0078] The coordinate y in the y-axis direction of the vector line segment passing through the first p p for:

[0079]

[0080] Using formula (7), the DC component A0 in the x-axis direction can be calculated as

[0081]

[0082] Using equations (2) to (8), we can calculate the nth order ellipticity coefficient a in the x-axis direction: n and b n for:

[0083]

[0084] Using formula (8), the DC component C0 in the y-axis direction can be calculated as

[0085]

[0086] Using equations (2) to (8), we can calculate the nth order ellipticity coefficient c in the y-axis direction: n and d n for:

[0087]

[0088] The mathematical model of the two-dimensional continuous curve coordinates {x(t), y(t)} of the external auditory canal contour is:

[0089]

[0090] Where N is the order of the fitting curve. The higher the order, the smaller the fitting error.

[0091] From the above formula, we can know that after obtaining the chain coding of the external auditory canal contour, we can calculate the coefficients of each order elliptical Fourier descriptor and the period time through formula (4-13) to construct a two-dimensional continuous curve model of the external auditory canal contour. Figure 6 A two-dimensional continuous curve (indicated in gray) of the external auditory canal contour reconstructed using a 20th-order elliptical Fourier transform is given.

[0092] Step 5) The external auditory canal contour curves obtained by the tomography are spliced ​​along the tomographic direction to obtain a three-dimensional structural model of the external auditory canal model as follows:

[0093] [x(t),y(t),z j ] (17)

[0094] where Z j is the value of the jth CT image in the z-axis direction. During the CT scanning process, there will be a distance interval Δz between each layer of CT image, so z j =j*Δz. Figure 7 The three-dimensional structure of the external auditory canal of a subject reconstructed using CT scan images is given.

[0095] Example 2

[0096] The present invention provides a system for reconstructing a three-dimensional model of the external auditory canal based on computed tomography, which is used to implement the above-mentioned method for reconstructing a three-dimensional model of the external auditory canal based on computed tomography. The system includes: a computed tomography scanner, a two-dimensional curve fitting module and a three-dimensional model reconstruction module, wherein:

[0097] The computerized tomography scanner is used to perform a computerized tomography scan on the subject's head to obtain a plurality of computerized tomography images;

[0098] The two-dimensional curve fitting module is used to determine the external auditory canal display area in the plurality of computed tomography images; to extract the edge of each external auditory canal display area to obtain a chain code of the external auditory canal contour; and to perform two-dimensional curve fitting on the extracted chain code using elliptical Fourier transform to obtain a two-dimensional continuous curve model of the external auditory canal in each layer of the computed tomography image;

[0099] The three-dimensional model reconstruction module is used to splice the two-dimensional continuous curve models in all the computer tomography images to obtain a three-dimensional model of the external auditory canal.

[0100] As can be seen from the above detailed description of the present invention, the method and system provided by the present invention for reconstructing a three-dimensional model of the external auditory canal based on computed tomography first extracts the edges of a slice image in the computed tomography image to obtain a chain code of a closed curve of the external auditory canal contour; secondly, the closed curve is fitted using an elliptical Fourier descriptor to obtain a two-dimensional curve mathematical expression of the external auditory canal projection of the current slice; finally, the two-dimensional curves of the external auditory canal projection extracted from all computed tomography images are spliced ​​together to obtain a three-dimensional model of the external auditory canal. This method does not require the injection of foreign matter into a real human ear canal, does not cause adverse reactions to the subject, and can accurately, efficiently, and quickly obtain a three-dimensional model of the external auditory canal.

[0101] Finally, it should be noted that the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention and are intended to be encompassed by the claims of the present invention.

Claims

1. A method for reconstructing a three-dimensional model of the external auditory canal based on computed tomography, comprising the following steps: Step 1) performing a computed tomography scan on the subject's head to obtain a plurality of computed tomography images; Step 2) determining the external auditory canal display area in the plurality of computed tomography images; Step 3) performing edge extraction on each external auditory canal display area to obtain a chain code of the external auditory canal contour; Step 4) performing two-dimensional curve fitting on the extracted chain code using elliptical Fourier transform to obtain a two-dimensional continuous curve model of the external auditory canal in each layer of the computed tomography image; Step 5) splicing the two-dimensional continuous curve models in all the computed tomography images to obtain a three-dimensional model of the external auditory canal; The method of performing two-dimensional curve fitting on the extracted chain code using elliptical Fourier transform to obtain a two-dimensional continuous curve model of the external auditory canal in a layer of computed tomography image specifically includes: Step 4-1) Based on the chain code V of the external auditory canal contour, the speed of moving along the vector segment is the unit speed, and the vector segment α passing through the i-th vector segment is calculated. i The time Δt i : Among them, α i represents the i-th vector segment, Step 4-2) Calculate the displacement Δx of the i-th vector segment in the x-axis direction i and the displacement Δy in the y-axis direction i ;in, Δx i =sgn(6-a i )sgn(2-a i ) (3) Δy i =sgn(4-a i )sgn(a i ) (4) Step 4-3) Get the time t taken to pass through the first p vector segments p and the total time T spent passing through all vector segments, where Wherein, K represents the total number of vector line segments of the two-dimensional continuous curve; Step 4-4) Get the coordinate x in the x-axis direction after passing through the first p vector segments p and the coordinate y in the y-axis direction p ;in, Step 4-5) Calculate the DC component A0 in the x-axis direction and the DC component C0 in the y-axis direction; wherein, Among them, x p-1 is the coordinate in the x-axis direction after passing through the first p-1 vector segments; y p-1 is the coordinate in the y-axis direction after passing through the first p-1 vector segments; Step 4-6) Calculate the nth order first ellipticity coefficient a in the x-axis direction n and the second ellipticity coefficient b n , and the nth order third ellipticity coefficient c in the y-axis direction n and the third ellipticity coefficient d n ,in, Among them, t p-1 is the time it takes to pass through the first p-1 vector segments; Step 4-7) Obtain the coordinates {x(t), y(t)} of the two-dimensional continuous curve model of the external auditory canal in the current layer of the computed tomography image, where Where N is the total order of two-dimensional curve fitting.

2. The method for reconstructing a three-dimensional model of the external auditory canal based on computed tomography according to claim 1, characterized in that: The step 1) specifically includes: performing a computed tomography scan on the subject's head along the Z-axis direction, scanning one layer of computed tomography image at every interval Δz to obtain a plurality of layers of computed tomography images.

3. The method for reconstructing a three-dimensional model of the external auditory canal based on computed tomography according to claim 1, wherein: The step 3) specifically includes: converting the plurality of computed tomography images into grayscale images, performing edge extraction on each external auditory canal display area using the Canny edge extraction method, and establishing a vector segment mapping of adjacent pixels starting from any pixel point along the edge of the external auditory canal contour to obtain a chain code V of the external auditory canal contour: V＝α1α2α3…α K (1) Among them, α1 represents the first vector segment, α2 represents the second vector segment, α3 represents the third vector segment, and α K Represents the Kth vector line segment, where K represents the total number of vector line segments of this continuous curve.

4. The method for reconstructing a three-dimensional model of the external auditory canal based on computed tomography according to claim 1, wherein: The step 5) specifically includes: splicing the two-dimensional continuous curve models of all CT images along the fault direction to obtain a three-dimensional model of the external auditory canal, wherein the fault direction is the Z-axis direction, and the value of the j-th CT image in the Z-axis direction is Z j .

5. A system for reconstructing a three-dimensional model of the external auditory canal based on computed tomography, used to implement the method for reconstructing a three-dimensional model of the external auditory canal based on computed tomography according to any one of claims 1 to 4, characterized in that: The system includes: a computer tomography scanner, a two-dimensional curve fitting module and a three-dimensional model reconstruction module, wherein: The computerized tomography scanner is used to perform a computerized tomography scan on the subject's head to obtain a plurality of computerized tomography images; The two-dimensional curve fitting module is used to determine the external auditory canal display area in the plurality of computed tomography images; to extract the edge of each external auditory canal display area to obtain a chain code of the external auditory canal contour; and to perform two-dimensional curve fitting on the extracted chain code using elliptical Fourier transform to obtain a two-dimensional continuous curve model of the external auditory canal in each layer of the computed tomography image; The three-dimensional model reconstruction module is used to splice the two-dimensional continuous curve models in all the computer tomography images to obtain a three-dimensional model of the external auditory canal.

Images