US 10,500,017 B2 · Method and apparatus for tooth body automatic preparation by digital controlled laser light and tooth retainer

Abstract

A digital control laser automatic tooth preparation method and device and a tooth positioner are provided. The device includes an intra-oral three-dimensional scanner, a dental laser, an oral working end, an oral and maxillofacial cone beam CT scanner, a computer, a tooth positioner, a negative-pressure suction device and a real-time monitoring device. The computer is connected respectively with the intra-oral three-dimensional scanner, the dental laser, the oral working end, the oral and maxillofacial cone beam CT scanner, the negative-pressure suction device, and the real-time monitoring device. The dental laser is connected with the oral working end of the digital control laser tooth preparation control system through a light guiding arm (1). The oral working end of the digital control laser tooth preparation control system is connected with the tooth positioner and the real-time monitoring device. The negative-pressure suction device is connected with the tooth positioner.

A method for digital control laser automatic tooth preparation, comprising the following steps:

1) acquiring three-dimensional surface scanning data of a crown portion of a target tooth with an intra-oral three-dimensional scanner, acquiring three-dimensional volume data of the crown portion of the target tooth with an oral and maxillofacial cone beam CT scanner, and storing the two aforementioned sets of data respectively;
2) registering the two sets of data with a tooth preparation CAD software and unifying the data in a same coordinate system, and on a computer screen extracting edges of a tooth preparation, defining design parameters used for the tooth preparation, and separating a dental enamel model portion, a dentin model portion and a dental pulp cavity model portion, so as to complete virtual modeling of the tooth preparation and obtain data of a virtual tooth preparation model, and storing the data of the virtual tooth preparation model in a STL format;
and
3) generating parameters of a cutting process with a laser tooth preparation CAM software according to the data of the virtual tooth preparation model, and outputting the parameters of the cutting process into a digital control laser tooth preparation control system, wherein the parameters of the cutting process comprise a focal spot diameter, a spot motion path and a scanning speed of the laser during tooth preparation;
wherein the method further comprises the following steps with the tooth preparation CAD software:
1) reading the three-dimensional data reconstructed from a CT image, and performing lighting and rendering;
2) reading the three-dimensional surface scanning data acquired by scanning the target tooth and the three-dimensional volume data acquired by the CT scanner and storing the same;
3) acquiring a three-dimensional surface scanning data model of the target tooth by the intra-oral three-dimensional scanner, dividing the three-dimensional surface scanning data model into a fusion region and a non-fusion region, wherein in the non-fusion region point set is fixed, and clipping and extracting a fusion curved surface from the three-dimensional surface scanning data model;
4) picking up three or more corresponding feature points from a three-dimensional volume data model acquired by the CT scanner and the three-dimensional surface scanning data model acquired by the intra-oral three-dimensional scanner, achieving alignment of the two models through alignment of the feature points, calibrating the feature points, denoted by si and si′ respectively, on the three-dimensional volume data model and the three-dimensional surface scanning data model, calculating a geometric transform matrix R, T between the feature points through si′=siR+T and applying the matrix to the three-dimensional volume data model and the three-dimensional surface scanning data model to perform an initial transformation;
5) performing a registration by adopting an Iterative Closest Point (ICP) algorithm, comprising rotating and translating a coordinate system of the three-dimensional surface scanning data model and a coordinate system of the three-dimensional volume data model, and minimizing the distance between homologous points by minimizing a least square approximate objective function between the volume data model and the surface scanning data model;
6) introducing a Laplace operator coefficient matrix L, so as to ensure that the transformation between Cartesian coordinates and differential coordinates is performed in a matrix multiplication manner, as shown in equation (2) below:
Δ=LV,L=I−D−1A (2),
wherein Δ={δi} is a differential coordinate of a mesh surface, i is the index number of a mesh vertex, V is a Cartesian coordinate of the mesh vertex, D is a diagonal matrix, and A is an adjacency matrix of the mesh;
7) realizing extraction of salient-featured points from the three-dimensional surface scanning data model, wherein a constraint condition is to search, in the three-dimensional volume data model, closest points corresponding to the feature points extracted from the three-dimensional surface scanning data model;
8) during deformation from deformable meshes of the three-dimensional volume data model to fixed meshes of the three-dimensional surface scanning data model, employing a multi-iteration deformation strategy, comprising dividing a single deformation into a multi-step deformation, wherein a deformation amount, denoted by b, for each iteration is calculated according to equation (3):
b
=
1
n
-
k
,
k
=
0
,
1
⁢
⁢
…
⁢
⁢
n
-
1
,
(
3
)
wherein n is the number of iterations, and k is the current number of iteration;
9) quantifying and assessing a degree of closeness between a deformable point in the three-dimensional volume data model and a target point in three-dimensional surface scanning data model based on a weight matrix, wherein the weight matrix is defined by equation (4) below:
wi=wd(k,d)×wa(α) (4),
wherein, wd(k,d) is a distance weighting function, k is the current number of iteration times, i is an index of a feature point, d is a distance between a deforming point gi and a target deformation point vi, Wa(α) is an angle weighting function, and α is an included angle between a normal direction of the deforming point gi and a ray direction li from a gravity point of the three-dimensional surface scanning data model to the deforming point;
and
10) introducing the weight matrix W into a linear equation and an energy function, so as to obtain equations (5) and (6),
(
L
wI
m
×
m
)
⁢
V
′
=
L
′
⁢
V
′
=
(
δ
wC
)
,
(
5
)
E
⁡
(
V
′
)
=

L
′
⁢
V
′
-
δ

2
+
∑
j
∈
C
⁢
w
⁢

v
j
′
-
c
j

2
,
(
6
)
wherein L is a Laplacian matrix before the target mesh is deformed, I is an m-order unit matrix, L′ is a Laplacian matrix after the target mesh is deformed, V′ are the vertices after the target mesh is deformed, δ is a Laplacian coordinate of the target mesh, C is the deformation amount of the vertex, v′j is a coordinate after the deformation of a constraint vertex j, and cj is a deformation amount of the constraint vertex j;
and
wherein the equation (6) is minimized for solution, and variable values x′, y′, z′ obtained by minimizing the equation (6) are used as the expected deformation position of the deformable mesh, wherein precomputation based on Cholesky decomposition is performed during the solution;
11) extracting a long axis of a tooth, wherein a directional vector in a space is S=(m, n, p) and a linear equation passing through any point (x0,y0,z0) is as shown in equation (7):
x
-
x
0
m
=
y
-
y
0
n
=
z
-
z
0
p
,
(
7
)
wherein x0, y0, z0 are three-dimensional coordinate values passing through the any point, and (m, n, p) is a coordinate representation of the directional vector, and an error equation as shown in equation (8) is obtained according to a best square approximation theory:
ƒ(εi1,εi2,εi3)=Σi=0q(εi12+εi22+εi32) (8),
wherein εi1, εi2, εi3 are respectively error components of the ith point of a point cloud in directions of the X axis, the Y axis and the Z axis of a fitting straight line, and q is the number of the points;
and wherein this non-linear ternary quadratic equation is solved by optimizing the directional vector (m, n, p), and the optimized directional vector (m, n, p) can be used with known midpoints to solve the axis of the tooth;
12) projecting the points onto the three-dimensional volume data model along a curvature direction point by point, and connecting respective projected points to obtain a projection curve;
13) biasing the three-dimensional volume data model by biasing respective points in a normal vector direction of each point, wherein multiple vectors on the mesh vertex are classified depending on types and are endowed with different weight values, and a calculation equation of a bias direction is shown in equation (9):
V
→
Offset
=
∑
j
=
1
n
⁢
W
j
·
N
→
i
,
j
,
(
9
)
wherein Voffset is the bias direction of a point, 1 . . . n are the numbers of triangular plates surrounding the mesh vertex in a circle, Wj is a weight value varying with the type of the triangular plate, Nij is a normal vector corresponding to respective triangular plates surrounding the mesh vertex in a circle;
14) setting up a model constraint equation set (10) by analyzing dimensional chain relationships which constrain the model:
{
L
2
=
∑
i
=
1
n
⁢
l
i
H
=
∑
i
=
1
n
⁢
h
i
tan
⁢
⁢
α
=
l
m
h
m
,
(
10
)
wherein, L represents a total buccolingual length;
H represents a total longitudinal height of a certain section;
l1 and l3 respectively represent a shoulder width and a distance from a surface edge to a long axis of the tooth;
h1 and h3 respectively represent a surface preparation thickness and a distance from the projection curve to a certain lower designated location;
α is an axial wall inclination of the preparation;
and l2 and h2 are respectively an edge length and a height corresponding to α;
for a complete preparation, n=3 and m=2, wherein l1, h1 and α are three parameterized parameter values;
for a specific tooth, H, L and h3 are fixed values;
and when the parameter values are given, the model constraint equation set (10) is the one containing three unknowns, that is, l2, l3 and h2, thus a set of values of l2, l3 and h2 can be determined uniquely from the three known equations, such that the sizes of respective portions of the model are determined;
15) deriving a tooth preparation model through a parametric operation method, and realizing parametric and dynamic modifications of the tooth preparation model through a history-based method;
and
16) detecting an error size of the tooth preparation model by calculating angular and size constraints of respective sections of the tooth preparation model, and displaying an error distribution condition in a colored cloud picture.

207 more elements

Method and apparatus for tooth body automatic preparation by digital controlled laser light and tooth retainer

Life of this patent

Method and apparatus for tooth body automatic preparation by digital controlled laser light and tooth retainer

Life of this patent

Abstract

Claims3