阅读说明：本技术 一种术前ctmr扫描图像与术中病人坐标的配准算法 (Registration algorithm for preoperative CTMR scanning image and intraoperative patient coordinates ) 是由 伯斯坦·巴勃罗大卫 于 2021-08-04 设计创作，主要内容包括：本发明公开了一种术前CTMR扫描图像与术中病人坐标的配准算法,包括如下步骤：S1、从术前的CTMR提取脊柱线、左右PSIS连接线点云,获得CTMR的术前点云；S2、术中,在病人的脊柱上滑动EMTS传感器从而获得脊柱线点云,在左右PSIS的连接线上滑动EMTS传感器从而获得左右PSIS连接线点云,结果进行存储；S3、预处理,调整EMTS的点云,使得十字的两条交叉线来自于交集,匹配扫描的长度；S4、配准；初始化,变换CTMR的点云以至于它的交集与EMTS相一致；然后使用ICP算法,进行从CTMR到EMTS的点云配准；最后利用刚性配准的变换矩阵对CTMR图像进行变换。本发明实现了两种点云信息的配准和校准。(The invention discloses a registration algorithm of a CTMR scanning image before operation and the coordinates of a patient in operation, which comprises the following steps: s1, extracting spinal line and left and right PSIS connecting line point clouds from the preoperative CTMR to obtain preoperative point clouds of the CTMR; s2, in the operation, sliding an EMTS sensor on the spine of a patient to obtain spine line point clouds, sliding the EMTS sensor on connecting lines of left and right PSIS to obtain point clouds of connecting lines of the left and right PSIS, and storing the results; s3, preprocessing, adjusting the point cloud of the EMTS, enabling two cross lines of the cross to come from an intersection and matching the scanning length; s4, registering; initializing, and transforming the point cloud of the CTMR so that the intersection of the point cloud is consistent with the EMTS; then point cloud registration from CTMR to EMTS is carried out by using an ICP algorithm; and finally, transforming the CTMR image by using the rigidly registered transformation matrix. The invention realizes the registration and calibration of two kinds of point cloud information.)

1. An algorithm for registration of preoperative CTMR scan images with intraoperative patient coordinates, characterized by: the method comprises the following steps:

s1, preoperative point cloud obtaining, namely extracting spinal line and left and right PSIS connecting line point clouds from preoperative CTMR to obtain preoperative point cloud of the CTMR;

s2, acquiring an EMTS point cloud in an operation, sliding an EMTS sensor on a spine of a patient in the operation to obtain a spine line point cloud, and storing a result of the spine line point cloud; sliding an EMTS sensor on connecting lines of the left PSIS and the right PSIS to obtain point clouds of the left PSIS connecting lines and the right PSIS connecting line points, and storing results of the point clouds of the left PSIS connecting lines and the right PSIS connecting lines;

s3, preprocessing, adjusting the point cloud of the EMTS, enabling two cross lines of the cross to come from an intersection and matching the scanning length;

s4, registering; initializing, and transforming the point cloud of the CTMR so that the intersection of the point cloud is consistent with the EMTS; then point cloud registration from CTMR to EMTS is carried out by using an ICP algorithm; and finally, transforming the CTMR image by using the rigidly registered transformation matrix.

2. The pre-operative CTMR scan image and intra-operative patient coordinates registration algorithm of claim 1, wherein: the preoperative point cloud acquisition method comprises the following specific steps:

s11, removing the scanning bed;

s12, positioning the MSP of the midsagittal plane of the patient to be parallel to the front and back axes of the image;

s13, positioning the left and right PSIS coordinates and assuming that the left and right PSIS coordinates are positioned on the same longitudinal slice;

s14, displaying the skin on the CT through a window;

s15, extracting point cloud connected with PSIS coordinates;

s151, in the longitudinal slice k containing the PSIS coordinates, starting from the column corresponding to the left PSIS coordinates and ending at the column j corresponding to the right PSIS coordinates;

s152, finding the last pixel of the column in each slice and saving the 3D coordinates of the last pixel;

s16, extracting point cloud of spinal lines;

s161, searching in each frame of axial slice k;

s162, searching the rearmost pixels on the column according to the MSP, and saving the 3D coordinates of the rearmost pixels.

3. The pre-operative CTMR scan image and intra-operative patient coordinates registration algorithm of claim 2, wherein: in step S12, if the MSP is not parallel to the front-to-back axis of the image, then the alignment is rotated.

4. The pre-operative CTMR scan image and intra-operative patient coordinates registration algorithm of claim 3, wherein: the rotational alignment refers to a counter-rotational alignment.

5. The pre-operative CTMR scan image and intra-operative patient coordinates registration algorithm of claim 2, wherein: in steps S152 and S162, the coordinate values of the 3D coordinates are represented by i, j, k.

Technical Field

The invention belongs to the technical field of image processing, and particularly relates to a registration algorithm of a pre-operative CTMR scanning image and an intra-operative patient coordinate.

Background

In recent years, medical imaging techniques have been rapidly developed, and Computed Tomography (CT), magnetic resonance imaging (MR), and the like are widely used in clinical diagnosis. Both CT and MR are popular and very important means for clinical imaging examinations. Among them, the basic principle of CT is: the technology of scanning the bedding surface with a certain thickness of a human body by using an X-ray beam and detector combination device is used for reconstructing the acquired data containing the attenuation difference of the voxels to the X and the spatial distribution information to obtain an image reflecting the spatial structure and the material density, and is convenient and fast, clear in image and high in density resolution; multi-layer scanning is adopted, a strong image post-processing function is achieved, and a 3D image can be constructed; the CT enhancement has better development on the blood supply condition and the perfusion state of the lesion part and has extremely high value for differential diagnosis. The basic principle of MR is: the human body is placed in a special magnetic field, and the hydrogen atomic nucleus in the human body is excited by radio frequency pulse to cause the hydrogen atomic nucleus to resonate and absorb energy. After stopping the radio frequency pulse, the hydrogen nucleus emits radio signals according to specific frequency, releases the absorbed energy, is recorded by a receiver outside the body, and is processed by an electronic computer to obtain an image. No radiation; the soft tissue resolution is excellent, muscles, ligaments and nerves can be clearly displayed, and blood vessel imaging of skeleton interference can be obtained.

For better operation, the coordinates of the patient need to be determined during operation, the preoperative point cloud information needs to be extracted by an expert through the sliding of the probe on the spine and the lumbar skin of the patient, but how to determine the coordinates of the patient more accurately needs to be registered by the extracted point cloud information of the CTMR scanning image.

Disclosure of Invention

The invention aims to solve the problems and provides a registration algorithm of a CTMR scanning image before operation and the coordinates of a patient during operation. According to point cloud information extracted by CT, point cloud information before operation is extracted by an expert through the sliding of a probe on the spine and the waist skin of a patient, and registration is carried out on the two kinds of point cloud information based on an iterative closest point algorithm (ICP); and using the rigid transformation matrix obtained by registration for calibrating the coordinate space and the CT space of the electromagnetic sensor.

In order to achieve the purpose, the technical scheme of the invention is as follows:

an algorithm for registration of preoperative CTMR scan images with intraoperative patient coordinates comprising the steps of:

s1, preoperative point cloud obtaining, namely extracting spinal line and left and right PSIS connecting line point clouds from preoperative CTMR to obtain preoperative point cloud of the CTMR;

s2, acquiring an EMTS point cloud in an operation, sliding an EMTS sensor on a spine of a patient in the operation to obtain a spine line point cloud, and storing a result of the spine line point cloud; sliding an EMTS sensor on connecting lines of the left PSIS and the right PSIS to obtain point clouds of the left PSIS connecting lines and the right PSIS connecting line points, and storing results of the point clouds of the left PSIS connecting lines and the right PSIS connecting lines;

s3, preprocessing, adjusting the point cloud of the EMTS, enabling two cross lines of the cross to come from an intersection and matching the scanning length;

s4, registering; initializing, and transforming the point cloud of the CTMR so that the intersection of the point cloud is consistent with the EMTS; then point cloud registration from CTMR to EMTS is carried out by using an ICP algorithm; and finally, transforming the CTMR image by using the rigidly registered transformation matrix.

As an improvement to the above technical solution, the preoperative point cloud acquisition specifically includes:

s11, removing the scanning bed;

s12, positioning a midsagittal plane (MSP) of the patient to be parallel to the front-back axis of the image;

s13, positioning the left and right PSIS coordinates and assuming that the left and right PSIS coordinates are positioned on the same longitudinal slice;

s14, displaying the skin on the CT through a window;

s15, extracting point cloud connected with PSIS coordinates;

s151, in the longitudinal slice k containing the PSIS coordinates, starting from the column corresponding to the left PSIS coordinates and ending at the column j corresponding to the right PSIS coordinates;

s152, finding the last pixel of the column in each slice and saving the 3D coordinates of the last pixel;

s16, extracting point cloud of spinal lines;

s161, searching in each frame of axial slice k;

s162, searching the rearmost pixels on the column according to the MSP, and saving the 3D coordinates of the rearmost pixels

As an improvement to the above technical solution, in step S12, if the MSP cannot be parallel to the front-back axis (anterior-posterior axis) of the image, the MSP is rotationally aligned.

As an improvement to the above technical solution, the rotational alignment refers to a reverse rotational alignment.

As a modification of the above-described technical solution, in step S152 and step S162, the coordinate values of the 3D coordinates are represented by i, j, k.

Compared with the prior art, the invention has the advantages and positive effects that:

the registration algorithm of the preoperative CTMR scanning image and the intraoperative patient coordinate is based on a 'lock and key' principle, an Electromagnetic tracking system (EMTS) is used for obtaining intraoperative point cloud, preoperative point cloud is extracted from the preoperative CTMR of a patient, the preoperative point cloud and the preoperative point cloud are mutually used as a lock and a key, namely the lock and the key represent cross-shaped point cloud characteristics extracted from spine and waist skin of the patient, and registration is carried out on the point cloud and corresponding point cloud in CT based on iterative closest point algorithm (ICP). And using the rigid transformation matrix obtained by registration for calibrating the coordinate space and the CT space of the electromagnetic sensor.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.

FIG. 1 is a schematic diagram of a state of extracting a spinal line point cloud from EMTS;

FIG. 2 is a schematic diagram of a PSIS link point cloud extraction from EMTS;

FIG. 3 is a schematic diagram of three states of an unregistered, preprocessed and registered point cloud;

FIG. 4 is an unregistered three-dimensional image;

fig. 5 is a registered three-dimensional image.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived from the embodiments of the present invention by a person skilled in the art without any creative effort, should be included in the protection scope of the present invention.

As shown in fig. 1 to 5, the present embodiment discloses a registration algorithm of a pre-operative CTMR scan image and intra-operative patient coordinates, which includes the following steps:

s1, preoperative point cloud obtaining, namely extracting spinal line and left and right PSIS connecting line point clouds from preoperative CTMR to obtain preoperative point cloud of the CTMR;

s2, acquiring an EMTS point cloud in an operation, sliding an EMTS sensor on a spine of a patient in the operation to obtain a spine line point cloud, and storing a result of the spine line point cloud; sliding an EMTS sensor on connecting lines of the left PSIS and the right PSIS to obtain point clouds of the left PSIS connecting lines and the right PSIS connecting line points, and storing results of the point clouds of the left PSIS connecting lines and the right PSIS connecting lines;

s3, preprocessing, adjusting the point cloud of the EMTS, enabling two cross lines of the cross to come from an intersection and matching the scanning length;

s4, registering; initializing, and transforming the point cloud of the CTMR so that the intersection of the point cloud is consistent with the EMTS; then point cloud registration from CTMR to EMTS is carried out by using an ICP algorithm; and finally, transforming the CTMR image by using the rigidly registered transformation matrix.

The preoperative point cloud acquisition method comprises the following specific steps:

s11, removing the scanning bed;

s12, positioning a midsagittal plane (MSP) of the patient to be parallel to the front-back axis of the image;

s13, positioning the left and right PSIS coordinates and assuming that the left and right PSIS coordinates are positioned on the same longitudinal slice;

s14, displaying the skin on the CT through a window;

s15, extracting point cloud connected with PSIS coordinates;

s151, in the longitudinal slice k containing the PSIS coordinates, starting from the column corresponding to the left PSIS coordinates and ending at the column j corresponding to the right PSIS coordinates;

s152, finding the last pixel of the column in each slice and saving the 3D coordinates of the last pixel;

s16, extracting point cloud of spinal lines;

s161, searching in each frame of axial slice k;

s162, searching the rearmost pixels on the column according to the MSP, and saving the 3D coordinates of the rearmost pixels

In step S12, if the MSP is not parallel to the anteroposterior axis of the image (anti-spatial axis), then the MSP is rotationally aligned. The rotational alignment refers to a counter-rotational alignment. In steps S152 and S162, the coordinate values of the 3D coordinates are represented by i, j, k.

The registration algorithm of the preoperative CTMR scanning image and the intraoperative patient coordinate is based on a 'lock and key' principle, an Electromagnetic tracking system (EMTS) is used for obtaining intraoperative point cloud, preoperative point cloud is extracted from the preoperative CTMR of a patient, the preoperative point cloud and the preoperative point cloud are mutually used as a lock and a key, namely the lock and the key represent cross-shaped point cloud characteristics extracted from spine and waist skin of the patient, and registration is carried out on the point cloud and corresponding point cloud in CT based on iterative closest point algorithm (ICP). And using the rigid transformation matrix obtained by registration for calibrating the coordinate space and the CT space of the electromagnetic sensor.