Method for quantitatively evaluating depth of tooth crack based on optical coherence tomography image

Abstract

The present application relates to the technical field of oral medical detection, in particular to a tooth crack depth quantitative evaluation method based on optical coherence tomography image, steps comprising: acquiring dual-wavelength interference spectrum data, performing spatial phase alignment processing to generate a depth-wavelength dual-channel intensity matrix; performing wavelength difference fitting along the depth direction of the light path to generate a local attenuation coefficient matrix; matching a spectral absorption characteristic model to calculate a medium component ratio distribution atlas; substituting into a light medium refractive index mapping model to calculate a dynamic local refractive index matrix; performing non-uniform segmented integration to reconstruct physical space depth distribution data; matching with a crack tip phase threshold to determine a physical termination point and output a depth quantitative evaluation value. The present application realizes high-precision reconstruction of optical path to physical space depth through a dynamic local refractive index matrix, determines the crack termination point in combination with the phase coherence gradient, and significantly improves the accuracy and reliability of tooth crack depth evaluation.

Description

Technical Field

[0001] This invention relates to the field of oral medical testing technology, specifically to a method for quantitative assessment of the depth of tooth microcracks based on optical coherence tomography (OCT) images. Background Technology

[0002] Dental imaging technology based on optical coherence tomography (OCT) utilizes the low-coherence interference principle of near-infrared light to acquire high-resolution, non-invasive tomographic images of dental tissues. In recent years, it has gradually moved from laboratory research to the auxiliary diagnosis of clinical endodontics. With the continuous development of frequency-scanning OCT and multi-wavelength interferometry, OCT is increasingly widely used in the microscopic observation of microstructural changes such as microcracks on the crown surface, early enamel demineralization, and microleakage at the margins of dental restorations, providing a new technical means for the early non-invasive detection of dental hard tissue diseases.

[0003] In the field of diagnosing dental cracks, existing OCT depth assessment techniques mainly involve acquiring axial or two-dimensional tomographic optical signals of the tooth, using algorithms to identify signal abrupt changes or strong reflection boundaries at the crack location in the image, and converting the optical path length into physical depth based on a preset refractive index of a single tooth tissue. This method can obtain relatively accurate geometric depth measurements in ideal, dry, and uncontaminated tooth specimens, providing a feasible technical approach for the preliminary quantitative assessment of cracks.

[0004] However, in real-life oral diagnostic scenarios, the interior of microcracks in teeth is easily filled with saliva or tissue fluid under capillary forces, and the interior of the cracks often contains exogenous pigments or early micro-demineralization products caused by bacterial metabolism. These mixed media not only exhibit wavelength-dependent spectral absorption and anisotropic scattering in the near-infrared band, but also cause complex local refractive index discontinuities along the depth direction of the crack. Existing technologies use physical models with a uniform refractive index and ignore media absorption attenuation, which leads to severe optical path distortion and false cliff-like attenuation of deep signals when the light beam penetrates the composite media layer. This causes the system to misjudge the media absorption abrupt change as the physical crack endpoint, or the calculated physical depth of the crack may deviate significantly from the actual depth due to refractive index deviation, failing to meet the clinical need for accurate assessment of the risk and prognosis of microcracks near the pulp. Summary of the Invention

[0005] In view of the above-mentioned shortcomings of the existing technology, the present invention provides a method for quantitative evaluation of tooth microcrack depth based on optical coherence tomography images, which can effectively solve the problems mentioned in the existing technology.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] This invention provides a method for quantitatively assessing the depth of tooth microcracks based on optical coherence tomography images, the specific steps of which include:

[0008] S100. Obtain dual-wavelength interference spectral data obtained by scanning the tooth crack region of the target tooth, wherein the dual-wavelength interference spectral data includes a deep reflection intensity sequence of a first predetermined wavelength and a second predetermined wavelength.

[0009] S200. Perform spatial phase alignment processing on the dual-wavelength interference spectral data to generate a depth-wavelength dual-channel intensity matrix. Based on the depth-wavelength dual-channel intensity matrix, perform wavelength difference fitting along the depth-direction light path to generate a local attenuation coefficient matrix.

[0010] S300. Based on the local attenuation coefficient matrix, a preset spectral absorption characteristic model is matched to calculate the medium composition ratio distribution map of the mixed medium inside the hidden crack.

[0011] S400. Substitute the medium composition ratio distribution map into the preset optical medium refractive index mapping model to calculate the dynamic local refractive index matrix along the microcrack extension path.

[0012] S500. Based on the dynamic local refractive index matrix, perform non-uniform piecewise integration on the optical sequence in the depth-wavelength dual-channel intensity matrix to reconstruct the physical space depth distribution data.

[0013] S600: Match the physical spatial depth distribution data with the preset crack tip phase threshold to determine the physical termination point of the tooth crack, and output the physical depth quantitative evaluation value of the tooth crack.

[0014] Furthermore, the step of performing spatial phase alignment processing on the dual-wavelength interferometric spectral data to generate a depth-wavelength dual-channel intensity matrix includes:

[0015] Extract the first deep reflection intensity sequence corresponding to the first predetermined wavelength and the second deep reflection intensity sequence corresponding to the second predetermined wavelength from the dual-wavelength interference spectral data;

[0016] Using the surface reflection peak of the target tooth as a spatial reference position, the first deep reflection intensity sequence and the second deep reflection intensity sequence are cross-correlated and registered.

[0017] Gaussian filtering and background noise reduction are performed on the registered sequence to construct the depth-wavelength dual-channel intensity matrix determined by both spatial depth index and wavelength index.

[0018] Furthermore, based on the depth-wavelength dual-channel intensity matrix, wavelength difference fitting is performed along the depth-direction light path to generate a local attenuation coefficient matrix, which is calculated using the following formula:

[0019] ;

[0020] in, For depth Elements in the local attenuation coefficient matrix at that location; For the first predetermined wavelength With the second predetermined wavelength The wavelength difference; The preset depth-sliding sampling step size; and These are the intensity values ​​of the first and second predetermined wavelengths at depth z in the depth-wavelength dual-channel intensity matrix, respectively.

[0021] Furthermore, the step of calculating the media composition ratio distribution map of the mixed medium inside the hidden crack by matching the local attenuation coefficient matrix with a preset spectral absorption characteristic model includes:

[0022] Extract the values ​​of each depth node from the local attenuation coefficient matrix;

[0023] Retrieve the preset reference attenuation coefficient for the liquid saliva phase and the preset calibration attenuation coefficient for the solid sediment phase from the preset spectral absorption feature library;

[0024] Based on the preset benchmark attenuation coefficient and the preset calibration attenuation coefficient, the values ​​of each depth node are linearly decoupled and normalized to obtain the distribution map of the medium composition ratio representing the volume fraction of solid sediments in the mixed medium at each depth node.

[0025] Furthermore, the dynamic local refractive index matrix along the microcrack propagation path is calculated by substituting the medium composition ratio distribution map into a preset optical medium refractive index mapping model, using the following formula:

[0026] ;

[0027] in, is the value in the dynamic local refractive index matrix at depth z; The pre-defined refractive index of the saliva medium; To calibrate the refractive index of the pre-defined solid deposit, This represents the volume fraction ratio at depth z in the medium composition ratio distribution map.

[0028] Furthermore, based on the dynamic local refractive index matrix, the optical sequence in the depth-wavelength dual-channel intensity matrix is ​​non-uniformly piecewise integrated to reconstruct the physical spatial depth distribution data, which is then calculated using the following formula:

[0029] ;

[0030] in, This refers to the physical spatial depth distribution data value corresponding to the optical path length L. This represents the dynamic local refractive index value at a small optical path length l. It represents a tiny optical path increment along the direction of light propagation.

[0031] Furthermore, the step of matching the physical spatial depth distribution data with a preset crack tip phase threshold to determine the physical termination point of the tooth crack includes:

[0032] Extract the complex signal phase data from the depth-wavelength dual-channel intensity matrix;

[0033] Calculate the phase coherence gradient sequence of the data distributed along the depth of the physical space;

[0034] Identify the depth node in the phase coherence gradient sequence that first falls below a preset phase coherence threshold;

[0035] The depth nodes below the preset phase coherence threshold are determined as the physical termination points of the tooth microcrack.

[0036] Furthermore, the method also includes:

[0037] After calculating the local attenuation coefficient matrix, identify the numerical step abrupt change points that occur along the depth direction in the local attenuation coefficient matrix;

[0038] The capillary edge of the air-liquid phase interface inside the hidden crack is located based on the numerical step abrupt change point.

[0039] Based on the capillary edge position, the dynamic local refractive index matrix is ​​subjected to segmented refractive index boundary correction on both sides of the capillary edge position.

[0040] Furthermore, the method also includes:

[0041] Obtain the surface normal vector of the target tooth in the crack inlet region;

[0042] Calculate the local incident angle between the scanning beam and the surface normal vector;

[0043] Based on the local incident angle, the reconstructed physical space depth distribution data is mapped in three dimensions using a preset law of light refraction to correct the oblique incident geometric angle error.

[0044] Furthermore, after outputting the quantitative assessment value of the physical depth of the tooth crack, the method further includes:

[0045] Obtain the three-dimensional spatial position data of the top wall of the pulp chamber of the target tooth;

[0046] Calculate the shortest spatial distance between the physical termination point and the three-dimensional spatial position data of the top wall of the pulp cavity to obtain the remaining dentin wall thickness value;

[0047] If the remaining dentin wall thickness is less than a preset safe wall thickness threshold, a pulp risk warning signal will be triggered.

[0048] The technical solution provided by this invention has the following advantages compared with the known prior art:

[0049] This invention extracts the local attenuation coefficient through dual-wavelength differential fitting and further decouples it to obtain the distribution spectrum of the medium composition ratio inside the hidden crack. Based on this, a dynamic local refractive index matrix is ​​constructed along the crack extension path. Combined with non-uniform piecewise integration, a high-precision reconstruction from optical path to physical spatial depth is achieved. At the same time, with the piecewise refractive index boundary correction at the capillary edge position and the oblique incidence geometry correction, the accuracy of quantitative assessment of the physical depth of tooth hidden cracks is significantly improved, overcoming the systematic depth measurement error caused by the traditional method's assumption of a single constant refractive index.

[0050] This invention extracts complex signal phase data and calculates the phase coherence gradient sequence. The position where the phase coherence gradient first falls below a preset threshold is used as the criterion for determining the physical termination point of the crack. This effectively avoids the problem of missed or false judgments caused by weak light signals and severe intensity attenuation in the crack tip region, and improves the reliability and stability of the detection of the termination point of the tooth microcrack depth. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0052] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0054] The present invention will be further described below with reference to embodiments.

[0055] Example:

[0056] Reference Figure 1 The method for quantitatively evaluating the depth of tooth microcracks based on optical coherence tomography images provided in this invention includes the following steps:

[0057] Step 1: Acquire Dual-Wavelength Interference Spectroscopy Data. A dual-wavelength optical coherence tomography (OCT) device is used to scan the microcracked area of ​​the target tooth point-by-point or line-by-line to acquire dual-wavelength interference spectral data. The dual-wavelength interference spectral data includes the depth reflection intensity sequence of a first predetermined wavelength and a second predetermined wavelength. In this embodiment, the selection of the first and second predetermined wavelengths should satisfy the requirement of having differentiated spectral absorption characteristics in the mixed medium within the hard tissue of the tooth and the microcrack. For example, two near-infrared bands with sufficient wavelength spacing can be selected to ensure the sensitivity of subsequent wavelength differential fitting to changes in the medium composition. This step is a routine data acquisition process for optical coherence tomography; the specific scanning parameters can be set according to the model of the equipment used and the clinical examination requirements.

[0058] Step 2: Perform spatial phase alignment processing on the dual-wavelength interference spectral data to generate a depth-wavelength dual-channel intensity matrix.

[0059] First, the first deep reflection intensity sequence corresponding to the first predetermined wavelength and the second deep reflection intensity sequence corresponding to the second predetermined wavelength are extracted from the dual-wavelength interferometric spectral data. Then, using the surface reflection peak of the target tooth as the spatial reference position, the first and second deep reflection intensity sequences are cross-correlated and registered to determine the optimal axial offset between the two sequences. The second deep reflection intensity sequence (or the first deep reflection intensity sequence) is then translated according to the offset to align the two sequences in the depth direction. After registration, Gaussian filtering and background noise reduction are performed on the registered sequences to suppress high-frequency noise and DC background interference. Finally, a depth-wavelength dual-channel intensity matrix, denoted as M(z,λ), is constructed, determined by both the spatial depth index and the wavelength index. The spatial depth index z corresponds to the sampling depth position along the deep optical path, and the wavelength index λ corresponds to either the first or second predetermined wavelength.

[0060] It should be noted that due to the difference in dispersion characteristics of the dual-wavelength probe light in dental tissue, there is an axial misalignment between the first and second depth reflection intensity sequences in the original data. Using the surface reflection peak of the target tooth as a spatial reference for cross-correlation registration allows the use of the enamel surface, a highly reflective interface, as a common spatial reference. This eliminates the depth misalignment caused by wavelength differences and system dispersion, ensuring that subsequent wavelength difference fitting is performed at the same physical depth, thus avoiding the introduction of spurious attenuation coefficients due to spatial misalignment.

[0061] Step 3: Based on the depth-wavelength dual-channel intensity matrix, perform wavelength difference fitting along the depth-direction light path to generate a local attenuation coefficient matrix.

[0062] The following formula is used for calculation:

[0063] ;

[0064] in, For depth Elements in the local attenuation coefficient matrix at that location; For the first predetermined wavelength With the second predetermined wavelength The wavelength difference; The preset depth-sliding sampling step size; and These are the intensity values ​​at depth z corresponding to the first and second predetermined wavelengths in the depth-wavelength dual-channel intensity matrix, respectively.

[0065] The local attenuation coefficient matrix is ​​obtained by sliding the sample point by point along the depth optical path with a preset depth sliding sampling step size.

[0066] By calculating the cross-ratio of two signals with different wavelengths at the same depth location, fluctuations in light source intensity, changes in fiber coupling efficiency, differences in tooth surface reflectivity, and system common-mode noise can be effectively eliminated, retaining only the local attenuation information determined by the intrinsic absorption and scattering characteristics of the medium. The selection of the depth-sliding sampling step size needs to strike a balance between depth resolution and statistical stability. If the step size is too small, the calculation results will be significantly affected by speckle noise; if the step size is too large, the depth positioning accuracy will be lost. In this embodiment, the interval can be set to several to more than ten sampling points based on the signal-to-noise ratio and depth sampling density.

[0067] Step 4: Based on the local attenuation coefficient matrix, match the preset spectral absorption characteristic model to calculate the distribution map of the medium composition ratio of the mixed medium inside the hidden crack.

[0068] The values ​​of each depth node in the local attenuation coefficient matrix are extracted. Preset baseline attenuation coefficients for the liquid saliva phase and preset calibration attenuation coefficients for the solid sediment phase are retrieved from a preset spectral absorption feature library. The preset spectral absorption feature library was obtained through preliminary experiments using dual-wavelength optical coherence tomography (OCT) on liquid saliva samples and solid sediment samples (such as food residue, plaque mineralization, or dental calculus samples) of known composition. The intrinsic attenuation coefficients of each individual medium phase at the first and second predetermined wavelengths are recorded. Based on the preset baseline and preset calibration attenuation coefficients, the values ​​of each depth node are linearly decoupled. The measured local attenuation coefficients are decomposed into attenuation contribution components of the liquid saliva phase and the solid sediment phase. A normalized calculation is then performed to obtain a medium composition ratio distribution map representing the volume fraction of solid sediment in the mixed medium at each depth node, denoted as R(z).

[0069] It should be noted that the hidden fractures are not filled with a single homogeneous medium, but rather a mixed medium containing both liquid saliva phase and solid sediment phase, with significant differences in the composition ratio of the medium at different depths. By matching and decoupling the measured local attenuation coefficient with a pre-defined spectral absorption characteristic model, the spatial distribution characteristics of the medium composition inside the hidden fractures can be quantitatively characterized. The medium composition ratio distribution map is the basic input data for subsequent accurate calculation of the dynamic local refractive index matrix.

[0070] Step 5: Substitute the medium composition ratio distribution map into the preset optical medium refractive index mapping model to calculate the dynamic local refractive index matrix along the microcrack extension path.

[0071] The following formula is used for calculation:

[0072] ;

[0073] in, is the value in the dynamic local refractive index matrix at depth z; The pre-defined refractive index of the saliva medium; To calibrate the refractive index of the pre-defined solid deposit, This represents the volume fraction ratio at depth z in the medium composition ratio distribution map.

[0074] The preset refractive index of the saliva medium and the preset calibrated refractive index of the solid sediment can be obtained by measuring the corresponding medium samples under standard temperature conditions using an Abbe refractometer, and then pre-entered as constants.

[0075] Traditional optical coherence tomography (OCT) typically uses a single tissue-average refractive index for optical path to physical depth conversion during depth reconstruction. However, in the unique scenario of dental microcracks, the upper part of the crack may be dominated by low-refractive-index saliva, while the lower part exhibits a higher equivalent refractive index due to the accumulation of solid deposits. Using a constant refractive index assumption would introduce significant depth reconstruction errors. This step, by combining a medium composition ratio distribution map with an optical medium refractive index mapping model, constructs a dynamic local refractive index matrix that varies along the microcrack's propagation path. This matrix accurately reflects the velocity variation of the light signal propagating within the microcrack, laying the optical physics foundation for subsequent high-precision physical depth reconstruction.

[0076] Step 6: Based on the dynamic local refractive index matrix, perform non-uniform piecewise integration on the optical sequence in the depth-wavelength dual-channel intensity matrix to reconstruct the physical spatial depth distribution data.

[0077] The following formula is used for calculation:

[0078] ;

[0079] in, This refers to the physical spatial depth distribution data value corresponding to the optical path length L. This represents the dynamic local refractive index value at a small optical path length l. It represents a tiny optical path increment along the direction of light propagation.

[0080] In the actual numerical implementation, the total optical path is discretized into several small optical path segments. The dynamic local refractive index value at each small optical path segment is determined by linear interpolation of the dynamic local refractive index matrix obtained in step five. The physical depth increment corresponding to each small optical path segment is accumulated segment by segment to obtain the real physical spatial depth at the corresponding optical path L.

[0081] Optical coherence tomography (OCT) systems directly measure and output the optical path length, not the actual physical depth. Because the dynamic local refractive index varies at different depths along the path of a microcrack, the light propagation speed changes accordingly. Therefore, a simple linear method of dividing the optical path length by the average refractive index cannot be used for depth conversion. A non-uniform piecewise integration method accumulates the actual physical depth corresponding to each small optical path length segment along the light propagation direction, achieving a high-precision nonlinear mapping from optical path length space to physical space, significantly improving the accuracy of quantitative assessment of dental microcrack depth.

[0082] Step 7: Match the physical spatial depth distribution data with the preset crack tip phase threshold to determine the physical termination point of the tooth crack and output the quantitative evaluation value of the physical depth of the tooth crack.

[0083] First, the complex signal phase data is extracted from the depth-wavelength dual-channel intensity matrix. Specifically, the interference intensity signal in the depth-wavelength dual-channel intensity matrix is ​​subjected to a Hilbert transform along the depth direction to construct an analytic signal. The real part of the analytic signal is the original interference intensity signal, and the imaginary part is the orthogonal component of the original interference intensity signal after the Hilbert transform. The argument of the analytic signal is the complex signal phase data.

[0084] Then, the phase coherence gradient sequence of the data distributed along the physical spatial depth is calculated. Specifically, a sliding calculation window is set along the depth direction corresponding to the physical spatial depth distribution data. Within the sliding calculation window, the absolute difference of the complex signal phase data between adjacent depth nodes is calculated, or the statistical variance of the complex signal phase data within the sliding calculation window is calculated, which serves as the phase coherence metric at the center depth position of the sliding calculation window. The calculation window is slid along the depth direction with a preset step size to obtain a phase coherence gradient sequence covering the entire depth range. In the region where the fissure exists, due to the inhomogeneity of the medium inside the fissure and the multiple scattering of the light signal, the coherence of the complex signal phase data is low, and the phase coherence gradient sequence shows a high value; while at the physical termination point of the fissure tip, the fissure terminates, the continuity of the tooth hard tissue is restored, the coherence of the light signal is significantly enhanced, and the phase coherence gradient sequence shows a significant decrease.

[0085] Next, the depth node in the phase coherence gradient sequence that first falls below a preset phase coherence threshold is identified. The preset phase coherence threshold is set based on the statistical baseline of phase coherence in normal tooth hard tissue regions; for example, it can be set as a proportional coefficient of the mean phase coherence of normal tooth tissue. The depth node that first falls below the preset phase coherence threshold is determined as the physical termination point of the tooth crack.

[0086] Finally, based on the physical spatial depth distribution data value corresponding to the physical termination point, the quantitative assessment value of the physical depth of the tooth crack is output.

[0087] The intensity signal of reflected light in the fracture tip region is often extremely weak due to depth attenuation and scattering. Relying solely on intensity thresholds can easily lead to misjudging noise as the termination point or missing the point due to signal overload. Complex signal phase data still exhibits good stability for weak signals. By monitoring the gradient change in phase coherence, the physical termination point of the fracture can be located more reliably, avoiding the limitations of traditional intensity threshold-based methods.

[0088] Step 8: Identify the numerical step abruption points that appear along the depth direction in the local attenuation coefficient matrix, locate the capillary edge position, and implement segmented refractive index boundary correction.

[0089] After calculating the local attenuation coefficient matrix, first-order difference operations or edge detection operations are performed on the element values ​​of the local attenuation coefficient matrix along the depth direction to identify numerical step abrupt change points. These numerical step abrupt change points correspond to the capillary edge position of the air-liquid phase interface inside the hidden crack, i.e., the meniscus position where air and liquid saliva phase contact at the opening of the hidden crack. On the upper side (air side) of the capillary edge position, the optical medium refractive index mapping model should use the air refractive index; on the lower side (inside the crack) of the capillary edge position, the dynamic local refractive index matrix calculated in step five is used. Based on the capillary edge position, piecewise refractive index boundary correction is applied to the dynamic local refractive index matrix on both sides of the capillary edge position to ensure that the non-uniform piecewise integral in step six uses the correct refractive index value for depth transformation at the phase interface.

[0090] There is a distinct air-liquid phase interface in the upper opening region of the hidden crack. The refractive index changes drastically at the interface. If the segmented refractive index boundary is not corrected, the non-uniform segmented integration in step six will produce a significant depth accumulation error at the phase interface, resulting in a systematic deviation in the physical spatial depth distribution data in the crack entrance region.

[0091] Step 9: Obtain the surface normal vector of the target tooth in the crack inlet area and correct for oblique incidence geometry error.

[0092] Obtain the surface normal vector of the target tooth in the crack inlet region. Specifically, the three-dimensional surface scanning function of the optical coherence tomography (OCT) device can be used to obtain the three-dimensional point cloud data of the tooth surface in the crack inlet region, or the three-dimensional morphology of the tooth surface can be reconstructed using the obtained tomographic image sequence and surface reconstruction algorithm. The local surface in the crack inlet region is fitted with a plane or curved surface, and the normal direction of the local surface, i.e., the surface normal vector, is calculated.

[0093] Calculate the local incident angle between the scanning beam and the surface normal vector. Specifically, the incident direction of the scanning beam is determined by the optical path geometry parameters of the dual-wavelength optical coherence tomography device, and the local incident angle is obtained by vector dot product operation in conjunction with the aforementioned surface normal vector.

[0094] Based on the local incident angle, the reconstructed physical space depth distribution data is mapped to a three-dimensional spatial vector using a preset law of light refraction to correct the oblique incidence geometric angle error. Specifically, the refraction direction of the scanning beam after entering the tooth tissue from the air is calculated according to the law of light refraction (Snell's law), and the refraction angle is obtained. Then, the physical space depth component along the direction of the refracted beam obtained in step six is ​​converted into a depth component along the normal direction of the tooth tissue surface (i.e., the tooth tissue thickness direction) through three-dimensional geometric projection, thereby correcting the oblique incidence geometric angle error.

[0095] In actual clinical scanning, due to the complexity of tooth anatomy and the limited space within the patient's mouth, it is often difficult to ensure that the scanning beam is precisely perpendicular to the surface of the crack. If the oblique incidence effect is ignored, the reconstructed physical depth will be measured along the beam direction rather than the tissue normal direction, leading to a systematic bias in the depth assessment value, which may underestimate or overestimate the true crack penetration depth. By correcting through three-dimensional spatial vector mapping, the final output quantitative assessment value of physical depth is ensured to reflect the true penetration depth along the tooth tissue thickness direction, thus improving the reliability of clinical assessment.

[0096] Step 10: Calculate the remaining dentin wall thickness and trigger a pulp risk warning.

[0097] After outputting the quantitative assessment value of the physical depth of the tooth crack, the three-dimensional spatial location data of the pulp chamber roof of the target tooth is obtained. The three-dimensional spatial location data of the pulp chamber roof can be obtained through preoperative cone-beam computed tomography (CBCT) data, or by using optical coherence tomography (OCT) in deep imaging mode to locate and image the pulp chamber roof while scanning the tooth crack area.

[0098] The three-dimensional spatial coordinates of the physical termination point determined in step seven are spatially registered with the three-dimensional spatial position data of the pulp cavity roof. The shortest spatial distance between the physical termination point and the three-dimensional spatial position data of the pulp cavity roof is calculated. The shortest spatial distance is the remaining dentin wall thickness value.

[0099] The remaining dentin wall thickness is compared with a preset safe wall thickness threshold. The preset safe wall thickness threshold is set according to clinical endodontic treatment guidelines, for example, it can be set to 0.5 mm. If the remaining dentin wall thickness is less than the preset safe wall thickness threshold, it indicates that the tooth crack has approached or threatened the pulp tissue, triggering the output of a pulp risk warning signal, prompting clinicians to take appropriate pulp protection or treatment intervention measures in a timely manner.

[0100] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for quantitatively assessing the depth of tooth microcracks based on optical coherence tomography (OCT) images, characterized in that, The specific steps include: S100. Obtain dual-wavelength interference spectral data obtained by scanning the tooth crack region of the target tooth, wherein the dual-wavelength interference spectral data includes a deep reflection intensity sequence of a first predetermined wavelength and a second predetermined wavelength. S200. Perform spatial phase alignment processing on the dual-wavelength interference spectral data to generate a depth-wavelength dual-channel intensity matrix. Based on the depth-wavelength dual-channel intensity matrix, perform wavelength difference fitting along the depth-direction light path to generate a local attenuation coefficient matrix. S300. Based on the local attenuation coefficient matrix, a preset spectral absorption characteristic model is matched to calculate the medium composition ratio distribution map of the mixed medium inside the hidden crack. S400. Substitute the medium composition ratio distribution spectrum into the preset optical medium refractive index mapping model to calculate the dynamic local refractive index matrix along the microcrack extension path. S500. Based on the dynamic local refractive index matrix, perform non-uniform piecewise integration on the optical sequence in the depth-wavelength dual-channel intensity matrix to reconstruct the physical space depth distribution data. S600: Match the physical spatial depth distribution data with the preset crack tip phase threshold to determine the physical termination point of the tooth crack, and output the physical depth quantitative evaluation value of the tooth crack.

2. The method for quantitative assessment of tooth microcrack depth based on optical coherence tomography images according to claim 1, characterized in that, The step of performing spatial phase alignment processing on the dual-wavelength interferometric spectral data to generate a depth-wavelength dual-channel intensity matrix includes: Extract the first deep reflection intensity sequence corresponding to the first predetermined wavelength and the second deep reflection intensity sequence corresponding to the second predetermined wavelength from the dual-wavelength interference spectral data; Using the surface reflection peak of the target tooth as a spatial reference position, the first deep reflection intensity sequence and the second deep reflection intensity sequence are cross-correlated and registered. Gaussian filtering and background noise reduction are performed on the registered sequence to construct the depth-wavelength dual-channel intensity matrix determined by both spatial depth index and wavelength index.

3. The method for quantitatively evaluating the depth of tooth microcracks based on optical coherence tomography images according to claim 1, characterized in that, The local attenuation coefficient matrix is ​​generated by performing wavelength difference fitting along the depth-wavelength dual-channel intensity matrix, and then calculated using the following formula: ； in, For depth Elements in the local attenuation coefficient matrix at that location; For the first predetermined wavelength With the second predetermined wavelength The wavelength difference; The preset depth-sliding sampling step size; and These are the intensity values ​​of the first and second predetermined wavelengths at depth z in the depth-wavelength dual-channel intensity matrix, respectively.

4. The method for quantitative assessment of tooth microcrack depth based on optical coherence tomography images according to claim 1, characterized in that, The step of calculating the composition ratio distribution map of the mixed medium inside the hidden crack by matching the local attenuation coefficient matrix with a preset spectral absorption characteristic model includes: Extract the values ​​of each depth node from the local attenuation coefficient matrix; Retrieve the preset reference attenuation coefficient for the liquid saliva phase and the preset calibration attenuation coefficient for the solid sediment phase from the preset spectral absorption feature library; Based on the preset benchmark attenuation coefficient and the preset calibration attenuation coefficient, the values ​​of each depth node are linearly decoupled and normalized to obtain the distribution map of the medium composition ratio representing the volume fraction of solid sediments in the mixed medium at each depth node.

5. The method for quantitatively evaluating the depth of tooth microcracks based on optical coherence tomography images according to claim 1, characterized in that, The dynamic local refractive index matrix along the microcrack propagation path is calculated by substituting the medium composition ratio distribution spectrum into a preset optical medium refractive index mapping model, using the following formula: ； in, is the value in the dynamic local refractive index matrix at depth z; The pre-defined refractive index of the saliva medium; To calibrate the refractive index of the pre-defined solid sediment, This represents the volume fraction ratio at depth z in the medium composition ratio distribution map.

6. The method for quantitative assessment of tooth microcrack depth based on optical coherence tomography images according to claim 1, characterized in that, The physical spatial depth distribution data is reconstructed by performing non-uniform piecewise integration on the optical sequence in the depth-wavelength dual-channel intensity matrix based on the dynamic local refractive index matrix, and calculated using the following formula: ； in, This corresponds to the physical spatial depth distribution data value at the optical path length L. This represents the dynamic local refractive index value at a small optical path length l. It represents a tiny optical path increment along the direction of light propagation.

7. The method for quantitative assessment of tooth microcrack depth based on optical coherence tomography images according to claim 1, characterized in that, The step of matching the physical spatial depth distribution data with a preset crack tip phase threshold to determine the physical termination point of the tooth crack includes: Extract the complex signal phase data from the depth-wavelength dual-channel intensity matrix; Calculate the phase coherence gradient sequence of the data distributed along the depth of the physical space; Identify the depth node in the phase coherence gradient sequence that first falls below a preset phase coherence threshold; The depth nodes below the preset phase coherence threshold are determined as the physical termination points of the tooth microcrack.

8. The method for quantitative assessment of tooth microcrack depth based on optical coherence tomography images according to claim 1, characterized in that, The method further includes: After calculating the local attenuation coefficient matrix, identify the numerical step abrupt change points that occur along the depth direction in the local attenuation coefficient matrix; The capillary edge of the air-liquid phase interface inside the hidden crack is located based on the numerical step abrupt change point. Based on the capillary edge position, the dynamic local refractive index matrix is ​​subjected to segmented refractive index boundary correction on both sides of the capillary edge position.

9. The method for quantitative assessment of tooth microcrack depth based on optical coherence tomography images according to claim 1, characterized in that, The method further includes: Obtain the surface normal vector of the target tooth in the crack inlet region; Calculate the local incident angle between the scanning beam and the surface normal vector; Based on the local incident angle, the reconstructed physical space depth distribution data is mapped in three dimensions using a preset law of light refraction to correct the oblique incident geometric angle error.

10. The method for quantitative evaluation of tooth microcrack depth based on optical coherence tomography images according to claim 1, characterized in that, After outputting the quantitative assessment value of the physical depth of the tooth crack, the method further includes: Obtain the three-dimensional spatial position data of the top wall of the pulp chamber of the target tooth; Calculate the shortest spatial distance between the physical termination point and the three-dimensional spatial position data of the top wall of the pulp cavity to obtain the remaining dentin wall thickness value; If the remaining dentin wall thickness is less than a preset safe wall thickness threshold, a pulp risk warning signal will be triggered.

Images