Method and system for multi-root dental radiology bone loss calculation

Abstract

The present application provides a multi-root tooth radiology bone loss calculation method, comprising: obtaining a three-dimensional tooth region of a target tooth; identifying a plurality of tooth roots and corresponding root axes of each tooth root; attributing the entire mesial surface or distal surface to the root axis with the smallest average vertical distance on the mesial surface or the distal surface, the average vertical distance being the average value of the vertical distances from a plurality of sampling points on the to-be-attributed tooth surface to the corresponding root axis; cutting the buccal surface or the lingual surface along the root bifurcation plane or the midplane of the root axes of the two tooth roots when the buccal surface or the lingual surface is close to or across the root bifurcation, to obtain a tooth root associated sub-region; determining an enamel-cementum junction reference point, an alveolar crest reference point, and an apical reference point of the tooth root in each sub-region; and calculating a root-associated radiology bone loss index along the root axis of the tooth root. The present application also provides a multi-root tooth radiology bone loss calculation system, an electronic device, and a readable storage medium.

Description

Technical Field

[0001] This invention relates to a method and system for calculating radiographic bone loss in multi-rooted teeth. Background Technology

[0002] Molars typically have two or three roots, and the spatial orientation of each root is not identical. In cases of severe distal root resorption, the mesial root may still retain relatively good bone support. Clinically, comparable quantitative results of bone loss are required, presented by root and surface.

[0003] If current three-dimensional measurements still use a common long axis running through the crown and root shaft to measure all tooth surfaces, the bone height and root length in a particular root direction will be distorted by the tilt of the common axis. The correspondence between tooth surfaces and roots will also be eliminated. Severe resorption of distal roots will be diluted by the common axis or recorded on tooth surfaces that do not belong to that root.

[0004] Periapical examination of multi-rooted teeth is prone to missing roots and multiple roots. Without tooth position priors and root number constraints, the set of apical points cannot be stably mapped to specific tooth roots. Without backtracking rules, fused roots and third molars may be incorrectly cut, resulting in a high proportion of false positives.

[0005] Therefore, current radiographic bone loss calculations still need improvement in terms of using a common long axis for uniform measurement in multi-rooted teeth and the lack of attribution between tooth surfaces and roots. Summary of the Invention

[0006] To address one of the aforementioned technical problems, this invention provides a method and system for calculating radiographic bone loss in multi-rooted teeth.

[0007] According to one aspect of the present invention, a method for calculating radiographic bone loss in multi-rooted teeth includes: acquiring a three-dimensional tooth region of a target tooth; identifying multiple roots and their corresponding root axes; assigning the mesial or distal surface as a whole to the root axis with the smallest average vertical distance, wherein the average vertical distance is the average of the vertical distances from a plurality of sampling points on the tooth surface to be assigned to the corresponding root axis; dividing the buccal or lingual surface into two root-associated sub-regions along the bifurcation plane or the midline of the root axes of the two roots when the difference between the average distances of the buccal or lingual surface and the two roots is less than a distance threshold, or when the buccal or lingual surface crosses the root bifurcation; determining a cementoenamel junction reference point, an alveolar ridge reference point, and an apical reference point of the root within each root-associated sub-region; and calculating a root-associated radiographic bone loss index for the sub-region along the root axis of the root.

[0008] According to at least one embodiment of the present invention, the expected number of roots is given based on the tooth position number, and the measured number of roots is based on the tooth roots identified in the image. When the measured number of roots is inconsistent with the expected number of roots, a backtracking process is initiated.

[0009] According to at least one embodiment of the present invention, an independent root number detection branch is set in the root identification model to obtain root number detection results, and the number of root regions is constrained according to the root number detection results.

[0010] According to at least one embodiment of the present invention, a cross-section of the tooth region is scanned from the coronal direction along the common tooth long axis. When the tooth region stably splits into multiple connected domains with an area exceeding an area threshold in a cross-section with a continuous axial length of not less than a set value, the axial position of the cross-section is determined as the root bifurcation plane. Each connected domain in the root bifurcation plane is taken as a different tooth root, and the common tooth long axis is used to find the root bifurcation. The root-associated radiometric bone loss index is measured along the root axis corresponding to each tooth root.

[0011] According to at least one embodiment of the present invention, for each tooth root, the root axis is fitted with the sequence of center points of the cross-section of the tooth root region, or the direction from the root bifurcation point to the root apex reference point is used as the root axis.

[0012] According to at least one embodiment of the present invention, a set of sampling points are taken along the local cementoenamel boundary curve or the midline of the tooth surface to be divided, the vertical distance from each sampling point to each root axis is calculated, and the average distance is calculated for the tooth surface.

[0013] According to at least one embodiment of the present invention, a measurement slice is generated that passes through a common tooth long axis and forms an angle with a mesiodistal reference axis, wherein the tooth long axis divides the slice into two half-slices. Multiple apical reference points are projected onto a plane perpendicular to the tooth long axis, and the angle between each projection point and the mesiodistal reference axis is calculated. For each half-slice, the tooth root corresponding to the apical reference point with the smallest difference in angle with the half-slice is selected for root association calculation of the half-slice.

[0014] According to at least one embodiment of the present invention, the bone height from the cementoenamel boundary reference point to the alveolar ridge reference point in the sub-region, parallel to the corresponding root axis, is determined as the root-related bone height; the absolute value of the projection of the difference between the root apex reference point and the cementoenamel boundary reference point in the sub-region in the direction of the corresponding root axis is determined as the root-related root length; and the root-related radiographic bone loss index is determined based on the ratio of the two.

[0015] According to at least one embodiment of the present invention, when a fused root or a stable connected domain cannot be split, the calculation is reduced to a single tooth long axis and unsplit roots are marked. When the root axis fitting of a certain tooth fails or the root-associated root length is less than the root length threshold, the tooth is discarded without affecting other tooth roots. In the case of a third molar or morphological abnormality, the calculation is reverted to a single tooth long axis.

[0016] According to at least one embodiment of the present invention, the index of each root-associated sub-region, the heaviest sub-region of each root, and the heaviest point of the whole tooth are output, and the heaviest point of the whole tooth is taken as the maximum index of each root-associated sub-region, without projecting each root reference point back to the common tooth long axis before calculating the proportion of the whole tooth.

[0017] According to at least one embodiment of the present invention, the expected number of roots for incisors, canines and most mandibular premolars is 1, the expected number of roots for maxillary first premolars is 2, the expected number of roots for maxillary first molars and second molars is 3, and the expected number of roots for mandibular first molars and second molars is 2.

[0018] According to at least one embodiment of the present invention, regional growth is performed within the tooth region using the apical reference point as a marker to obtain each root region, as an alternative to the splitting of connected domains.

[0019] According to at least one embodiment of the present invention, the cementoenamel junction reference point and the alveolar ridge reference point are measured along the corresponding root axis, and the measurement reference for each tooth surface is the root axis of the tooth root to which that tooth surface belongs.

[0020] According to another aspect of the present invention, a multi-rooted tooth radiographic bone loss calculation system includes: a tooth region acquisition module for acquiring a three-dimensional tooth region of a target tooth; a tooth root identification module for identifying multiple tooth roots and the root axis corresponding to each tooth root; a sub-region division module for assigning the mesial or distal surface as a whole to the root axis with the smallest average vertical distance on the mesial and distal surfaces, wherein the average vertical distance is the average of the vertical distances from a number of sampling points on the tooth surface to be assigned to the corresponding root axis; and when the difference between the average distances of the buccal or lingual surface and the two tooth roots is less than a distance threshold, or when the buccal or lingual surface crosses the root bifurcation, dividing the buccal or lingual surface into two root-associated sub-regions along the bifurcation plane or the midline of the root axes of the two tooth roots; a reference point determination module for determining a cementoenamel junction reference point, an alveolar ridge reference point, and an apical reference point of the tooth root within each root-associated sub-region; and a root association index calculation module for calculating the root-associated radiographic bone loss index of the sub-region along the root axis of the tooth root.

[0021] According to one aspect of the present invention, an electronic device is provided, comprising: a memory storing execution instructions, and a processor executing the execution instructions stored in the memory, such that the processor performs the method as described in any of the preceding claims.

[0022] According to one aspect of the present invention, a readable storage medium is provided, wherein executable instructions are stored therein, which, when executed by a processor, are used to implement the method as described in any of the preceding claims.

[0023] Compared with the prior art, the present invention has the following beneficial effects.

[0024] The mesial and distal surfaces are assigned to the root axis with the smallest average vertical distance. The buccal and lingual surfaces are cut along the bifurcation plane or the midline of the root axes when they are close to or cross the root bifurcation. The tooth surface can be mapped to a specific root. Second, the root-associated radiographic bone loss index is calculated along the root axis of that root. Severe resorption on a root is not diluted by the tilt of the common tooth long axis. Tooth position numbering gives the expected number of roots, based on imaging. For fused roots, unstable splitting, or abnormal third molar morphology, the tooth is reduced to a single tooth long axis and marked as unsplit; these tooth positions can still be calculated. The most critical point of the whole tooth is the maximum index in each root-associated subregion; the reference points of each root are not projected back to the common tooth long axis before calculating a whole tooth proportion. For single-rooted teeth, fused roots, or failed splitting, the tooth is reduced to a single tooth long axis and marked as unsplit; these tooth positions still output tooth surface-level indices. Attached Figure Description

[0025] Figure 1 A method for calculating radiographic bone loss in multi-rooted teeth according to an embodiment of the present invention is shown.

[0026] Figure 2 A schematic diagram of three roots of the maxillary molars and two roots of the mandibular molars is shown according to one embodiment of the present invention.

[0027] Figure 3 A schematic diagram of the connected domain splitting and root bifurcation planes appearing when scanning along the long axis of a common tooth according to an embodiment of the present invention is shown.

[0028] Figure 4 A schematic diagram showing the relationship between each root axis and the common tooth long axis is shown in one embodiment of the present invention.

[0029] Figure 5 A schematic diagram is shown according to one embodiment of the present invention, showing the approximate belonging of the adjacent face and the separation of the cheek and tongue faces along the cutting plane.

[0030] Figure 6 A schematic diagram of root-associated bone height and root-associated root length is shown in one embodiment of the present invention.

[0031] Figure 7 A schematic diagram of a fused root retraction to a single tooth long axis is shown according to one embodiment of the present invention.

[0032] Figure 8 A schematic diagram of a multi-rooted tooth radiographic bone loss calculation system according to an embodiment of the present invention is shown.

[0033] Figure 9 A schematic diagram of an electronic device according to an embodiment of the present invention is shown. Detailed Implementation

[0034] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be noted that, for ease of description, only the parts relevant to the present invention are shown in the accompanying drawings.

[0035] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined with each other. The technical solution of this invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0036] Unless otherwise stated, the exemplary embodiments shown are to be understood as providing exemplary features of various details that provide ways in which the technical concept of the invention can be implemented in practice. Therefore, unless otherwise stated, the features of the various embodiments may be additionally combined, separated, interchanged, and rearranged without departing from the technical concept of the invention.

[0037] The terminology used herein is for the purpose of describing particular embodiments and is not intended to be limiting. As used herein, unless the context clearly indicates otherwise, the singular forms “a” and “the” are intended to include the plural forms as well. Furthermore, when the terms “comprising” and “including” and variations thereof are used in this specification, they indicate the presence of the stated features, steps, operations, and components, but do not exclude the presence or addition of other features, steps, operations, and components.

[0038] According to one embodiment of the present invention, a method for calculating radiographic bone loss in multi-rooted teeth is provided.

[0039] Figure 1 A method for calculating radiographic bone loss in multi-rooted teeth according to an embodiment of the present invention is shown.

[0040] like Figure 1 As shown, the method may include steps 102 to 112. This method is for multi-rooted teeth such as molars. First, multiple roots and their respective root axes are identified. Then, the mesial and distal surfaces are assigned to the nearest root axis. The buccal and lingual surfaces are cut when crossing roots. Finally, the root-related radiographic bone loss index is calculated under the root axis. When a single-rooted tooth, fused root, or branched root fails, the calculation is reduced to the long axis of a single tooth.

[0041] In step 102, the three-dimensional tooth region of the target tooth is obtained.

[0042] The tooth region is segmented from single-tooth instances on cone-beam CT, represented as a single-tooth binary mask. The system reads the DICOM sequence, sorts the image sequence according to the image location information, and reads spatial information such as pixel spacing, slice thickness, and image orientation to reconstruct the 3D volume data. There are no special restrictions on the acquisition device model, tube voltage and current parameters, or reconstruction algorithm.

[0043] Reference points for the cementoenamel junction and alveolar ridge can be derived from extreme values ​​of section thermograms, manual calibration, or other test results. The apical reference point can be derived from extreme values ​​of three-dimensional thermograms or from the center of the apical extreme zone within the root region.

[0044] In step 104, multiple tooth roots and the root axis corresponding to each tooth root are identified.

[0045] Figure 2 A schematic diagram of three roots in the maxillary molars and two roots in the mandibular molars according to an embodiment of the present invention is shown. The expected number of roots is given according to tooth position number: incisors, canines, and most mandibular premolars have an expected number of 1 root; the maxillary first premolar has an expected number of 2 roots, corresponding to the buccal and palatal roots; the maxillary first and second molars have an expected number of 3 roots, corresponding to the mesiobuccal, distobuccal, and palatal roots; and the mandibular first and second molars have an expected number of 2 roots, corresponding to the mesial and distal roots. The third molar exhibits significant morphological variation and is processed based on measured connected regions and thermal peak numbers, with regression allowed.

[0046] Prior knowledge is used to constrain root division and name roots, but it does not replace the image. If the measured number of roots is inconsistent with the prior knowledge, the image shall prevail, and a regression process shall be initiated. In one embodiment, the second mesiobuccal root of the maxillary molar is not used as a separate root axis for bone resorption partitioning. When the mesial root of the mandibular molar is further divided, the main implementation method still calculates based on a single mesial axis.

[0047] The multi-root tooth detection model can add an independent root number detection branch to the original structure. This branch outputs the root number category or root number regression value. The root number detection result is used to constrain the number of subsequent connected domains or the number of thermal peaks, but does not directly rewrite the root axis direction.

[0048] Figure 3 This diagram illustrates the appearance of connected domain splitting and root bifurcation planes when scanning along the long axis of a common tooth according to an embodiment of the present invention. First, the long axis of the common tooth is fitted to locate root bifurcation. The initial direction of the long axis can be derived from the principal components of the tooth region or from the fitting of the midpoint of the crown-root segment. A cross-section is scanned along this axis from the coronal direction to the root. The scanning step size can be one voxel layer thickness. Two-dimensional connected domain analysis is performed on the tooth region on each layer, recording the number of connected domains and the area of ​​each connected domain.

[0049] If, in several consecutive layers, the tooth region splits from a single connected domain into multiple connected domains whose areas all exceed the threshold, then the axial position of the first stable split is recorded as the root bifurcation plane. The continuous axial length can be no less than 1.5 mm. The area threshold is used to remove small fragments near the root apex, and the continuous length is used to remove single-layer noise. Each connected domain in the root bifurcation root direction grows into a different root region. The root trunk and crown in the coronal direction of the root bifurcation are not included in a single root and are only used for calculations related to the common axis and cementoenamel boundary.

[0050] The connected domains at the root bifurcation point need to be grown in three dimensions to obtain the complete root region. This can be done by connecting the connected domains in each layer from the bifurcation plane towards the root, according to their adjacency. If a connected domain in a certain layer merges or disappears, the growth of that root is stopped, and the resulting growth is taken as the root region.

[0051] For each tooth root, a sequence of center points of the cross-section from the root bifurcation to the root apex is taken, and a straight line is fitted to obtain the root axis, with the direction pointing from the root bifurcation to the root apex. In an alternative method, the vector pointing from the root bifurcation point to the root apex reference point is directly used as the root axis. The root apex reference point can be the three-dimensional thermodynamic extremum in the root region or the center of the root-square extreme region of the root mask.

[0052] When fitting the root axis, the center of the cross-section perpendicular to the common axis within the root region is taken. Alternatively, a cross-section perpendicular to the current estimated direction of the root can be taken. The process is iterated once or twice, with least squares used for linear fitting. If the fitting residual is too large, it indicates significant root curvature. In this case, a broken line or spline can be used as the root axis, and the local tangential direction is taken as the projection direction during measurement.

[0053] Figure 4 A schematic diagram illustrating the relationship between each root axis and the common tooth long axis is shown in one embodiment of the present invention. In the diagram, α is the angle of deviation of the root axis relative to the common tooth long axis. The measurement reference is each root axis, and the common axis is used to find bifurcation. In the alternative root division method, the effective extreme value of the three-dimensional periapical thermogram is used as a marker to perform regional growth or watershed within the tooth region to obtain each root region. Root division is not limited to the use of thermograms.

[0054] A root count detection branch can be added after the encoder of the tooth root recognition model. This branch outputs the root count category, such as 1, 2, 3, and greater than 3. During training, tooth position priors and artificial root counts are used as supervision. During inference, if the detected root count matches the number of connected component splits, the split result is adopted. If they do not match, connected component splits are assumed first, and the detected root count is written into the quality control marker. If a root is missed, the area threshold can be lowered and the scan can be repeated.

[0055] The three maxillary roots are named mesobuccal root, distal buccal root, and palatal root in the prior knowledge. After the connected domain is split into three regions, they can be named according to the relationship between the position of the center of each region on the horizontal plane and the tangent of the dental arch and the buccal-lingual direction. The root closest to the palatal side is the palatal root. Of the remaining two roots, the one closer to the middle is the mesobuccal root, and the one closer to the distal side is the distal buccal root. Of the two mandibular roots, the root bifurcates to the mesobuccal side, which is the mesobuccal root, and to the distal side, which is the distal root. If the mesobuccal root splits again at the root direction, in one implementation, it still maintains a single axis.

[0056] In step 106, on the mesial and distal surfaces, the mesial surface or the distal surface as a whole is assigned to the root axis with the smallest average vertical distance. The mesial surface of the mandibular first molar is assigned to the mesial root, and the distal surface is assigned to the distal root. The distal surface of the maxillary first molar is assigned to the distal buccal root.

[0057] In step 108, on the buccal and lingual surfaces, when the difference in the average distance between the buccal or lingual surface and the two tooth roots is less than a distance threshold, or when the buccal or lingual surface crosses the root bifurcation, the buccal or lingual surface is divided into two root-associated sub-regions along the bifurcation plane or the midline of the root axes of the two tooth roots. In one embodiment, the distance threshold is 1 mm. After incision, the sub-region on the mesial side of the root bifurcation is assigned to the mesial root or mesobuccal root, and the sub-region on the distal side is assigned to the distal root or distal buccal root.

[0058] Figure 5 This diagram illustrates the assignment of proximal surfaces and the separation of buccal and lingual surfaces along the incision plane according to an embodiment of the present invention. One implementation of the distance definition involves taking a set of sampling points along the local cementoenamel boundary curve or along the midline of the sector within the circumferential interval or tooth surface sector to be divided. The vertical distance from each sampling point to each root axis is calculated, the interval is averaged, and the interval is assigned to the root with the smallest average distance.

[0059] The sampling points for the average vertical distance can be taken along the local cementoenamel boundary curve of the tooth surface with equal arc lengths, for example, every 0.5 mm. The vertical distance from point p to the root axis is calculated as a point to a straight line, and the root axis is recorded as passing through point p. , direction is ,but: ,in This represents the perpendicular distance from point p to the root axis.

[0060] The average of all sampling points on the tooth surface is calculated to obtain the distance from the surface to each root. For the mesial and distal surfaces, the smallest distance is assigned to the entire surface. For the buccal and lingual surfaces, the difference between the smallest and second smallest distances is considered. If the difference is less than 1 mm, the surface is cut. The cutting plane is the cross section of the midline of the root axes of the two roots at the root bifurcation height, or the root bifurcation plane can be used directly.

[0061] The mesial and distal surfaces are usually each mainly adjacent to one root. The mesial or distal surface as a whole is assigned to the root axis with the smallest average vertical distance. The mesial surface of the mandibular first molar is assigned to the mesial root, and the distal surface is assigned to the distal root. The distal surface of the maxillary first molar is assigned to the distal buccal root. The relationship of the palatal adjacent surfaces is determined according to the distance from the palatal root.

[0062] The buccal and lingual surfaces often span two tooth roots. When the difference in the average distance between the same clinical tooth surface and the two roots is less than a threshold, such as 1 mm, or when the tooth surface anatomically spans the root bifurcation, the tooth surface is divided into two root-related sub-regions along the midline of the root axes or along the bifurcation plane, and calculated separately. The buccal and lingual surfaces of mandibular molars are treated by incision. In maxillary molars with three roots, the buccal surface can be incised according to the mesial buccal root and the distal buccal root, while the palatal surface is treated separately as the palatal root.

[0063] In one alternative method, the distance can be assigned using the angle between the central ray of the tooth surface and each root axis, with the smallest angle being used. Another assignment method consistent with the slice measurement path is as follows: The slice passes through the long axis of the tooth and forms a certain angle with the mesial-distal reference axis, for example, one slice is taken every 30°. The long axis of the tooth divides the slice into two halves. Each half has a cementoenamel junction and an alveolar ridge. Multiple apical points are projected onto a plane perpendicular to the long axis of the tooth, and the angle between each projection point and the mesial-distal reference axis is calculated. Each half also has an angle relative to the mesial-distal reference axis. The apical point with the smallest angle difference is selected for root association calculation of that half. In half-slice angle matching, the apical points are projected onto a plane perpendicular to the long axis of the tooth, and the angle between each projection point and the mesial-distal reference axis is calculated. Each half also has an angle relative to the mesial-distal reference axis. For each half-section, select the root tip with the smallest angle difference. ,in ,in, The root number ultimately selected for this half-section, with the apical reference point being... , The origin of the local coordinate system is x, and x / y are the first and second axes of the local coordinate system. The root to which the root tip belongs is the half-section. Every 30° section will cover 12 half-sections.

[0064] The two classification methods can be used independently or cross-validated. In one implementation, half-slice angle matching is used as the main path; in another implementation, the entire surface rule is used as the main path. Sub-regions that fail the verification can be marked and then manually confirmed.

[0065] In step 110, the cementoenamel junction reference point, alveolar ridge reference point, and apical reference point of the tooth root are determined in each root-associated sub-region.

[0066] In step 112, the root-associated radiometric bone loss index of the subregion is calculated along the root axis of the tooth root.

[0067] Figure 6 A schematic diagram of root-associated bone height and root-associated root length is shown in one embodiment of the present invention. For the associated sub-region s of the tooth root r, the root is first divided, then the tooth surface and root sub-regions are divided according to the nearest root axis, and the corresponding root-associated radial bone loss percentage is calculated. The cementoenamel boundary reference point for this sub-region is denoted as... The alveolar ridge reference point is The root tip reference point is The unit vector of the axis is The root-associated bone height, root-associated root length, and root association parameters are as follows: Among them, the height of the root-related bone Root association root length .

[0068] In one implementation, without deducting physiological distance, the ratio is directly obtained from the root-associated bone height and root-associated root length, and the measurement axis is this root axis. Even if the cementoenamel junction and alveolar ridge are taken from slices passing through a common axis, they are still projected onto that root axis. When determining reference points within a sub-region, half-slices or tooth surface sampling points already assigned to that root are used; apical reference points are taken from the root's own apex. The output includes the index of each root-associated sub-region, the heaviest sub-region of each root, and the heaviest point of the entire tooth. The heaviest point of the entire tooth is the maximum index among all root-associated sub-regions. .

[0069] The report provides the root and tooth surface or sub-region to which the site belongs, and may also provide the heaviest sub-region for each root. When summarizing the indicators, the median of each sub-region for that root can be given simultaneously, with the maximum value still taken for the heaviest site of the entire tooth. When the distal root axis is deflected by an angle α relative to the common axis, if the projection is changed to the common axis z, then: Where α is the deflection angle of the distal root axis relative to the common axis. For the height of the root-related bone, This is the bone height measured along the common tooth's long axis z. The root length is also multiplied by the cosine α; after changing the measurement axis to this root axis, α is no longer included in the projection.

[0070] Figure 7A schematic diagram of the regression of a fused root to a single tooth long axis is shown in one embodiment of the present invention. The regression is performed in the following order: when a fused root or one that cannot split into a stable connected region, the entire tooth is regressed to a single tooth long axis, output according to tooth surface indices, and unsplit roots are marked. When it is desired that only two of the three roots will stably split, it is calculated as two roots, and the third root is marked as invalid. The heaviest point of the entire tooth is taken from the calculated sub-region. If the root axis fitting of a certain root fails or the root length is less than a threshold, such as 5 mm, the root is discarded without affecting other roots. When the number of apical thermal peaks is inconsistent with the number of split roots, the number of split roots in the mask is used. If there is no thermal peak in the root, the center of the root-side extreme region of the mask is used as the root apex of the root. When the third molar or the morphology is abnormal, regression to a single axis is given priority.

[0071] The regression judgment can be performed sequentially. First, it is determined whether the splitting is stable. If not, the entire tooth is regressed and unsplit roots are marked. If yes, the number of roots is checked to see if it matches the expected number. If not, the actual number of roots is used for calculation, and missing roots are marked as invalid. Next, it is checked whether the root length of each root reaches the threshold. If not, the root is discarded. Finally, it is checked whether the third molar or quality control mark is morphologically abnormal. If so, forced regression is allowed. The tooth surface level indicators output by the regression can record unsplit roots.

[0072] The following is an example of two mandibular first molars.

[0073] The target tooth is the left mandibular first molar with a desired number of roots of 2. The long axis of the common tooth is fitted by the crown and the middle of the root trunk. Scanning along this axis, at a distance of about 4.5 mm apical to the cementoenamel junction, the tooth region within a continuous range of 2 mm is stably split into two connected domains with areas exceeding the threshold. This position is denoted as the root bifurcation plane. The two connected domains apical to the root grow into the mesial root and the distal root.

[0074] The mesial root axis is approximately 12° mesial relative to the common tooth length axis, and the distal root axis is approximately 18° distal relative to the common tooth length axis. The average vertical distance from the mesial surface to the mesial root axis is approximately 1.1 mm, and to the distal root axis is approximately 3.4 mm, with the entire surface belonging to the mesial root. The distance from the distal surface to the distal root axis is approximately 0.9 mm, and to the mesial root axis is approximately 3.6 mm, with the entire surface belonging to the distal root. The difference in average distance between the buccal surface and the two roots is approximately 0.4 mm, which is less than the 1 mm threshold. The buccal surface is cut along the root bifurcation plane to form the buccal mesial region and the buccal distal region. The lingual surface is also cut in the same way.

[0075] In the distal root distal surface region, the root-associated bone height is approximately 5.8 mm, the root-associated root length is approximately 11.5 mm, and the root association index is approximately 50%. In the mesial root mesial surface region, the root-associated bone height is approximately 1.7 mm, the root-associated root length is approximately 13 mm, and the root association index is approximately 13%. The buccal distal surface region accounts for approximately 46%, and the buccal mesial surface region accounts for approximately 18%. If the distal surface is calculated using the common tooth long axis, the distal surface bone height is approximately 4.1 mm, the root length is approximately 14 mm, and the proportion is approximately 29%. The most significant point of the entire tooth output by this method is approximately 50% of the distal root distal surface.

[0076] The following is an example of a three-rooted maxillary first molar.

[0077] The target tooth is the first molar on the right maxilla, with a desired number of roots of 3. The root number detection branch outputs 3. Scanning along the common axis, it stably splits into three connected domains at approximately 5 mm apical to the cementoenamel junction, corresponding to the mesobuccal root, distal buccal root, and palatal root, respectively. The mesobuccal root axis is deviated approximately 15° mesobuccally relative to the common axis, the distal buccal root axis is deviated approximately 20° distal buccally, and the palatal root axis is deviated approximately 25° palatally. The mesobuccal surface is close to the mesobuccal and palatal roots. It is cut along the midline of the root axes of the two roots. The entire distal surface is assigned to the distal buccal root, and the buccal surface is cut along the midline of the mesobuccal and distal buccal roots. The entire palatal surface is assigned to the palatal root.

[0078] The distal facial region of the distal buccal root has a bone height of approximately 6.2 mm and a root length of approximately 12 mm, accounting for approximately 52% of the total. The mes buccal lateral region of the mes buccal root accounts for approximately 21%, and the palatal lateral region of the palatal root accounts for approximately 16%. The most critical point of the entire tooth is the distal facial region of the distal buccal root.

[0079] In the half-slice angle matching embodiment, the included angle interval is 30°. A slice with an included angle of about 60% is divided into two half-slices by the tooth long axis. The difference between the included angle of the distal buccal half-slice and the included angle of the projection of the distal buccal root apex on the horizontal plane is the smallest. The distal buccal root apex is selected to calculate the index of this half-slice. The difference between the included angle of the mesial lingual half-slice and the projection of the palatal root is the smallest. The palatal root apex is selected.

[0080] In the fused root embodiment, the buccal root and palatal root of the maxillary second molar did not show stable splitting, and the continuous scan was always a single connected domain. The whole tooth marker was not divided into roots and degenerated into a single tooth long axis, outputting four tooth surface indices.

[0081] The two roots of the maxillary first premolar are treated as buccal roots and palatal roots. The buccal and palatal surfaces usually do not need to be cut across roots. The proximal surfaces can be assigned to the whole surface according to the nearest root axis. Single-rooted anterior teeth are regressed and output tooth surface level indicators. In the whole mouth process, molars are assigned by root, and anterior teeth are automatically assigned to a single axis. The report fields are consistent, but molars have additional root names and sub-region names.

[0082] In one implementation, the desired number of roots is based on a tooth position table, with a continuous length of 1.5 mm for connected regions, a threshold of 1 mm for the distance difference between buccal and lingual incisions, a lower limit of 5 mm for root length, and a half-slice interval of 30°.

[0083] The report outputs the root-associated bone height, root-associated root length, and root-associated radiographic bone loss index for each root-associated subregion. It also provides the name of the root and the name of the tooth surface or subregion to which the subregion belongs. The most severely affected subregion for each root is listed separately. The most severely affected point for the entire tooth is the maximum value of each subregion, with the root and tooth surface indicated. Cases without root segmentation are marked as such in the report. Four tooth surface indices are output. Severity indicators can be generated based on set thresholds. These indicators are for radiographic reference only and are not considered clinical diagnostic conclusions.

[0084] The expected number of roots for the maxillary first premolar is 2, corresponding to the buccal and palatal roots. After a stable split is observed during scanning along the common axis, the buccal and palatal root axes are fitted separately. The proximal surfaces are assigned to the nearest root axis, with the buccal surface usually assigned to the buccal root and the palatal surface usually assigned to the palatal root. If the difference in distance between the buccal surface and the two roots is less than a threshold, the tooth can still be cut along the midline, and the root association index is calculated along their respective root axes.

[0085] In the third molar embodiment, the morphological variation is large. If the connected domain scan cannot stably split, the entire tooth reverts to a single tooth long axis and is marked as unsplit. If it stably splits into two roots, it is calculated as two roots as measured, and not treated as three roots.

[0086] Quality control markers may include: discrepancies between the measured number of roots and the expected number of roots, a root length less than 5 mm, excessive root axis fitting residuals, and conflicts between half-slice angle matching and whole-surface attribution. Conflicting sub-regions can be manually confirmed.

[0087] When a reference point is determined within a sub-region, a half-section or tooth surface sampling point that has already been assigned to the root is used. The cementoenamel junction reference point and alveolar ridge reference point can be derived from the extreme values ​​of the section thermogram or from manual calibration. The apical reference point is taken from the root apex of the root itself. When there is no thermal peak in the root, the center of the root-side extreme region of the root mask is used as the root apex of the root.

[0088] During system deployment, the root recognition module can independently update the root number detection branch without changing the index calculation formula. The sub-region division module can switch between vertical distance and angle matching. The backtracking module provides a marked single-axis result when root division fails. The system can be deployed on a local imaging workstation or on a server that interfaces with the hospital information system.

[0089] In one implementation, the scanning step size is one voxel layer thickness, the continuous axial length of connected regions is 1.5 mm, the cutting distance difference threshold is 1 mm, the lower limit of root length is 5 mm, the half-slice angle interval is 30°, and the sampling point interval is 0.5 mm. The initial direction of the common tooth long axis can be derived from the principal components of the tooth region or from fitting the center point of the crown-root midsection. The scanning step size can be one voxel layer thickness, and two-dimensional connected region analysis is performed on the tooth region on each layer. When the number of connected regions changes from 1 to N, and the area of ​​all N connected regions exceeds the area threshold, the number of consecutive layers is counted. After the axial length corresponding to the consecutive layers reaches 1.5 mm, the layer that first meets the conditions is recorded as the root bifurcation plane.

[0090] Among the alternative root segmentation methods, the effective extreme values ​​of the three-dimensional periapical thermogram are used as markers. Regional growth or watersheds are performed within the tooth region to obtain each root region. Regional growth can extend from each periapical marker to the coronal direction until they meet near the root bifurcation. Watersheds can use the root bifurcation plane as the watershed ridge. The root regions obtained by both methods can be used to fit the root axis.

[0091] The incision plane is a cross-section of the midline of the root axes of the two roots at the bifurcation height, or the bifurcation plane can be used directly. After incision, the subregion belonging to the mesial side of the bifurcation is assigned to the mesial root or mesiobuccal root, and the distal side is assigned to the distal root or distalbuccal root. In the case of three maxillary roots, the buccal surface is incised according to the mesiobuccal root and distalbuccal root, and the palatal surface is assigned to the palatal root separately. If the mesial surface is adjacent to both the mesiobuccal root and the palatal root, it can also be incised along the midline of the root axes of the two roots.

[0092] In the embodiment of the maxillary first premolar, the desired number of roots is 2. The buccal root axis is deviated buccally by about 20° relative to the common axis, and the palatal root axis is deviated palatally by about 22°. The average vertical distance from the mesial surface to the buccal root axis is small. The entire surface is attributed to the buccal root, and the entire palatal surface is attributed to the palatal root. The buccal surface does not need to be cut across the roots. The root association index is calculated along their respective root axes.

[0093] When the root axis fitting of a certain root fails or the root-related root length is less than 5mm, the root is discarded without affecting other roots. When the number of apical thermal peaks is inconsistent with the number of split roots, the number of split roots in the mask is used as the standard. The tooth surface level index records that have not been split are rolled back to distinguish them from cases that have been calculated according to the root axis.

[0094] After steps 102 to 112, the multi-rooted teeth obtain root-associated radiometric bone loss indices measured by root axis.

[0095] The expected number of roots is determined by tooth position: 1 for incisors, canines, and most mandibular premolars; 2 for maxillary first premolars; 3 for maxillary first and second molars; 2 for mandibular first and second molars; and the third molar is processed according to the measured connected domain. The expected number of roots is used to constrain root division and name roots, and is not used to replace the image.

[0096] The output report may include the names of each root, the names of each root-associated sub-region, root-associated bone height, root-associated root length, root-associated radiographic bone loss index, the heaviest sub-region of each root, the heaviest point of the entire tooth, and quality control markers such as undivided roots, insufficient root length, and classification conflicts. Auxiliary prompts are for radiographic reference only and are not used as clinical diagnostic conclusions. The heaviest point of the entire tooth is the maximum index within each root-associated sub-region; the reference points of each root are not projected back to the common tooth long axis before calculating the proportion of the entire tooth.

[0097] The measurement axis is the root axis of the tooth root. Even if the cementoenamel junction and alveolar ridge are taken from slices passing through the common axis, they are still projected onto the root axis. The root-related bone height is the distance from the cementoenamel junction reference point to the alveolar ridge reference point in the sub-region, parallel to the corresponding root axis. The root-related root length is the absolute value of the projection of the difference between the root apex reference point and the cementoenamel junction reference point in the sub-region in the direction of the corresponding root axis.

[0098] According to another aspect of the present invention, a system for calculating radiographic bone loss in multi-rooted teeth is provided.

[0099] Figure 8 A schematic diagram of a multi-rooted tooth radiographic bone loss calculation system according to an embodiment of the present invention is shown. The multi-rooted tooth radiographic bone loss calculation system may include a tooth region acquisition module 1002, a tooth root identification module 1004, a sub-region division module 1006, a reference point determination module 1008, a root association index calculation module 1010, and a backtracking processing module 1012.

[0100] The tooth region acquisition module is used to acquire the three-dimensional tooth region of the target tooth. The tooth root recognition module is used to identify multiple tooth roots and the root axis corresponding to each tooth root. The sub-region division module is used to divide the circumferential tooth surface or tooth surface sub-region to the corresponding tooth root according to the spatial distance relationship with each root axis, thus obtaining the tooth root associated sub-region. The reference point determination module is used to determine the cementoenamel junction reference point, alveolar ridge reference point, and root apex reference point of the tooth root in each tooth root associated sub-region. The root association index calculation module is used to calculate the root association radiographic bone loss index of the sub-region along the root axis of the tooth root. The regression processing module is used to degenerate to a single tooth long axis calculation when fused roots, root separation failure, or morphological abnormalities occur.

[0101] In one embodiment, the root identification module may further include a root number detection submodule, a connected domain scanning submodule, and a root axis fitting submodule. The sub-region division module may further include a proximal surface assignment submodule, a buccal-lingual incision submodule, and a half-slice angle matching submodule. The processing logic of each submodule corresponds to the above-described method implementation, and will not be repeated here.

[0102] The system can be deployed on a dental imaging workstation or server. The input is a single tooth region and optional reference points. The output is the indicators of each root-related sub-region, the most critical point of each root, and the most critical point of the whole tooth. After the server completes the analysis, it returns the root-related indicators to the client for the physician to view.

[0103] Figure 9 A schematic diagram of an electronic device according to an embodiment of the present invention is shown. The electronic device includes a memory 1300 and a processor 1200. The memory 1300 stores execution instructions, and the processor 1200 executes the execution instructions stored in the memory 1300 to implement a method for calculating radiographic bone loss in multi-rooted teeth.

[0104] The device may include corresponding modules that perform one or more steps in the above process. Modules may be hardware modules specifically configured to perform the corresponding steps, or implemented by a processor configured to perform the corresponding steps, or stored in a computer-readable medium for implementation by a processor, or implemented by some combination thereof.

[0105] This hardware architecture can be implemented using a bus architecture. A bus architecture can include any number of interconnect buses and bridges, depending on the specific application and overall design constraints. The bus connects various circuits, including one or more processors, memory, and hardware modules. It can also connect other circuits such as peripherals, voltage regulators, and power management circuits.

[0106] A bus can be an industry-standard architecture bus, an external device interconnection bus, or an extended industry-standard architecture bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, only one connection line is used in the attached diagram, but this does not imply that there is only one bus or one type of bus.

[0107] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process. The scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed in a different order than those shown or discussed.

[0108] The processor executes the various methods and processes described above. The method implementations of this invention can be implemented as software programs tangibly contained in a machine-readable medium, such as memory. When the software program is loaded into memory and executed by the processor, one or more steps of the methods described above can be performed. Alternatively, the processor can be configured in other suitable ways to execute the methods described above.

[0109] A readable storage medium can be any means that can contain, store, communicate, propagate, or transmit programs for use by an instruction execution system, apparatus, or device. More specific examples include: electrical connections having one or more wires, portable computer disk drives, random access memory, read-only memory, erasable and programmable read-only memory, fiber optic devices, and portable read-only memory.

[0110] Various parts of this invention can be implemented in hardware, software, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software stored in memory and executed by a suitable instruction execution system. If implemented in hardware, they can be implemented using discrete logic circuits, application-specific integrated circuits, programmable gate arrays, field-programmable gate arrays, etc., which are well known in the art.

[0111] Those skilled in the art will understand that all or part of the steps of the methods described above can be implemented by a program instructing related hardware. The program can be stored in a readable storage medium. When executed, the program includes one or a combination of the steps of the method implementation.

[0112] In the various embodiments of this invention, the functional units can be integrated into one processing module, or each unit can exist physically separately, or two or more units can be integrated into one module. The integrated module can be implemented in hardware or as a software functional module.

[0113] The present invention also provides an electronic device, comprising: a memory storing execution instructions; and a processor or other hardware module, wherein the processor or other hardware module executes the execution instructions stored in the memory, causing the processor or other hardware module to perform the above-described method.

[0114] The present invention also provides a readable storage medium storing executable instructions, which, when executed by a processor, are used to implement the above-described method.

[0115] In one implementation, during the full-mouth procedure, molars are assigned to their respective root regions, while anterior teeth automatically revert to a single long axis. For molars, the output includes indicators for each root-associated sub-region, the most critical point for each root, and the most critical point for the entire tooth. When a root is fused, a third molar, or a root separation failure occurs, it is marked as unseparated, and four tooth surface indicators are output. Output fields include root name, sub-region name, root-associated bone height, root-associated root length, and root-associated radiographic bone loss indicator.

[0116] This invention first identifies multiple root axes, then divides the tooth surface or sub-regions according to spatial distance relationships to the corresponding tooth roots, and calculates root-associated radiographic bone loss indices along the root axes. In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," or "some examples" indicate that a specific feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Specific features, structures, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified. Those skilled in the art should understand that the above embodiments are merely for clearly illustrating the invention and are not intended to limit the scope of the invention. For those skilled in the art, other changes or modifications can be made based on the above disclosure, and these changes or modifications are still within the scope of this invention.

Claims

1. A method for calculating radiographic bone loss in multi-rooted teeth, characterized in that, include: Obtain the three-dimensional tooth region of the target tooth; Identify multiple tooth roots and the root axis corresponding to each tooth root; On the mesial and distal surfaces, the mesial surface or the entire distal surface is assigned to the root axis with the smallest average vertical distance, where the average vertical distance is the average of the vertical distances from several sampling points on the tooth surface to be assigned to the corresponding root axis. On the buccal and lingual surfaces, when the difference in the average distance between the buccal or lingual surface and the two tooth roots is less than a distance threshold, or when the buccal or lingual surface crosses the root bifurcation, the buccal or lingual surface is divided into two root-associated subregions along the bifurcation plane or the midline plane of the root axes of the two tooth roots. Within each root-associated subregion, determine the cementoenamel junction reference point, alveolar ridge reference point, and apical reference point of the root; and Calculate the root-associated radiometric bone loss index for this subregion along the root axis of the tooth root.

2. The method as described in claim 1, characterized in that, The identification of multiple tooth roots and the root axis corresponding to each tooth root includes: Given the expected number of roots based on the tooth position number; and Based on the tooth roots identified in the image, a backtracking process is initiated when the actual number of roots differs from the expected number.

3. The method as described in claim 2, characterized in that, Also includes: A separate root number detection branch is set up in the tooth root recognition model to obtain the root number detection results; as well as The number of root regions is constrained based on the root number detection results.

4. The method as described in claim 1, characterized in that, The identification of multiple tooth roots and the root axis corresponding to each tooth root includes: Scan the cross-section of the tooth region from the coronal direction to the apical direction along the long axis of the common tooth; When the tooth region stably splits into multiple connected domains with an area exceeding the area threshold in a cross section with a continuous axial length of not less than a set value, the axial position of the cross section is determined as the root bifurcation plane. Each connected region on the root side of the root bifurcation plane is considered a different tooth root; and The common tooth long axis is used to locate root bifurcation, and the root-associated radiometric bone loss index is measured along the root axis corresponding to each tooth root.

5. The method as described in claim 4, characterized in that, For each tooth root, fit the root axis using the sequence of center points of the cross-section of that root region; or The direction from the root bifurcation point to the root apex reference point is taken as the root axis.

6. The method as described in claim 1, characterized in that, The average vertical distance is determined as follows: Take a set of sampling points along the local cementoenamel boundary curve of the tooth surface to be divided or along the midline of the tooth surface; Calculate the vertical distance from each sampling point to each root axis and average the distance for that tooth surface.

7. The method as described in claim 1, characterized in that, The root-associated subregion is also determined as follows: A measurement slice is generated that passes through a common tooth long axis and forms an angle with a mesial-distal reference axis, wherein the tooth long axis divides the slice into two halves. Multiple apical reference points are projected onto a plane perpendicular to the tooth's long axis, and the angle between each projection point and the mesial-distal reference axis is calculated; and For each half-slice, select the tooth root corresponding to the apical reference point with the smallest difference in angle with the half-slice, and use it for the root association calculation of that half-slice.

8. The method as described in claim 1, characterized in that, The calculation of the root-associated radiographic bone loss index for this sub-region along the root axis of the tooth root includes: The bone height from the cementoenamel junction reference point to the alveolar ridge reference point within this sub-region, parallel to the corresponding root axis, is defined as the root-related bone height. The absolute value of the projection of the difference between the root apex reference point and the cementoenamel junction reference point of the sub-region onto the corresponding root axis direction is determined as the root-associated root length; and The root-associated radiographic bone loss index is determined based on the ratio of the root-associated bone height to the root-associated root length.

9. The method as described in claim 1, characterized in that, This also includes rollback in the following ways: When there is a fused root or when a stable connected domain cannot be split, the calculation is reduced to a single tooth major axis, and unsplit roots are marked. If the root axis fitting for a particular tooth root fails or the root-associated root length is less than the root length threshold, that tooth root is discarded without affecting other tooth roots; and In cases of third molar or morphological abnormalities, regress to a single tooth long axis.

10. The method as described in claim 8, characterized in that, Output the root-associated radiographic bone loss index for each root-associated subregion, the heaviest subregion of each root, and the heaviest point of the entire tooth; and The most critical point of the whole tooth is taken as the maximum index in each root-related sub-region, instead of projecting each root reference point back to the long axis of the common tooth before calculating the proportion of the whole tooth.

11. The method as described in claim 2, characterized in that, The expected number of roots for incisors, canines, and most mandibular premolars is 1; The expected number of roots for the maxillary first premolar is 2; The expected number of roots for the maxillary first and second molars is 3; and The expected number of roots for the mandibular first and second molars is 2.

12. The method as described in claim 4, characterized in that, Using the apical reference point as a marker, regional growth is performed within the tooth region to obtain each root region, serving as an alternative method for connected domain splitting.

13. The method as described in claim 1, characterized in that, The cementoenamel junction reference point and alveolar ridge reference point are measured along the corresponding root axis, and the measurement reference for each tooth surface is the root axis of the tooth root to which that tooth surface belongs.

14. A system for calculating radiographic bone loss in multi-rooted teeth, characterized in that, include: The tooth region acquisition module is used to acquire the three-dimensional tooth region of the target tooth. The tooth root recognition module is used to identify multiple tooth roots and the root axis corresponding to each tooth root; The sub-region division module is used to assign the mesial surface or the entire distal surface to the root axis with the smallest average vertical distance on the mesial and distal surfaces. The average vertical distance is the average of the vertical distances from several sampling points on the tooth surface to be assigned to the corresponding root axis. When the difference between the average distance of the buccal or lingual surface and the two tooth roots is less than a distance threshold, or when the buccal or lingual surface crosses the root bifurcation, the buccal or lingual surface is divided into two root-associated sub-regions along the root bifurcation plane or the midline plane of the root axes of the two tooth roots. The reference point determination module is used to determine the cementoenamel junction reference point, alveolar ridge reference point, and apical reference point of each tooth root within each root-associated sub-region; and The root-related index calculation module is used to calculate the root-related radiographic bone loss index of the sub-region along the root axis of the tooth root.

15. An electronic device, characterized in that, include: The memory stores execution instructions; as well as A processor that executes execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1 to 13.

16. A readable storage medium, characterized in that, The readable storage medium stores execution instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 13.

Images