DESC:Description
The following Specification describes the invention.
FIELD OF THE INVENTION:
[001] Embodiments described herein generally pertain to the field of spinal surgery, an automated method for facilitating the accurate placement of pedicle screws or implants in spinal surgery.
BACKGROUND OF THE INVENTION:
[002] Pedicle-screw placement is one of the most common surgery performed to stabilise vertebrae, arrest relative displacement, and enable them to fuse. The pedicle is in close proximity to the spinal canal, which stores an integral portion of the central nervous system. Adjacent to the spine is numerous vital organs. Therefore, the accuracy of pedicle access, in position and direction, is of utmost importance as the margin for error is very small. The consequence of misplacement can lead to disastrous complications.

[003] Patent Application Publication No US 20040240715 A1, published on Dec. 2, 2004, relates to methods and computer systems for determining the placement of pedicle screws in spinal surgery. In accordance with one aspect of the invention, the locus of the centre of minimum transverse pedicle widths for each axial slice is stored. These data represent a locus of points used to determine the optimum path by performing a linear least-square fit. The disadvantage of this technique is that solution is dependent on a reference coordinate system, which does not account for the relative orientation of a pedicle. Further the pedicle cross-sectional area is simplified to be one-dimensional as transverse pedicle width. This simplification of two-dimensional space can be the source of error and can lead to inaccurate assessment. An assessment reveals that the two critical shortcomings in the method may result in the pedicle-screw path being less optimal or false pedicle-screw axis. The technique is less likely to provide optimum placement in length and direction. In contrast, the new method of the present invention uses the locus of the centre of the pedicle coronal cross-section to determine the optimum pedicle-screw axis. A line passing through or nearby locus of centres of pedicle bottleneck region assures maximum margin on the transverse directions to the line. A line that pivots at the area-centroid of the smallest sectional area and pass through area-centroid of adjacent layers on either side in the least perpendicular distance, as described more particularly hereinafter. It assures maximum margin on the transverse directions. The method of the present invention sets an autonomous a processing steps for determining the maximum screw diameter for intraosseous (within bone) screw placement.
[004] Patent Application Publication No US 20070232960A1, published on Oct. 4, 2007, relates to a method of determining the pedicle base circumference and the pedicle isthmus to facilitate screw placement through a pedicle of a vertebral body during spinal surgery. It discloses a method wherein identifying the pedicle base circumference as the areas of the outer cortical surface. The adjacent second lines perpendicular to the tangent at that location are at the greatest angle with respect to one another. The invention explains a method of identifying the pedicle isthmus as the areas of the outer cortical surface where the second lines that are opposed to each other are closest to being parallel to one another. This technique has practical limitations while operating in a digital space where multiple data sets(pixels) exist with the highest and equal angle. The method of the present invention provides an approach leading to a non-ambiguous solution. The autonomous a processing steps defines the pedicle isthmus and centre of the pedicle isthmus.
[005] Patent Application Publication No US 7,835,497 B2, published on Oct. 16, 2010, relates to a method for automatic evaluation of tomographic image data records of a patient and generation of slices in the correct orientation in the region of interest.
SUMMARY OF INVENTION
[006] The invention described herein is a method for accurate analysis of geometric properties of pedicle of a patient’s spine using medical imaging data obtained from one or more medical imaging devices like stationary computed tomography (CT), mobile CT system, positron emission tomography- computed tomography (PET-CT). In at least one embodiment, the present invention discloses a method for automatic evaluation of tomographic image data of a patient spine to accurately determine the sizing and orientation of pedicle screws, as well as comprehensive visualization of pedicles using Machine-independent Multi-Planar Reconstruction (MiMPR), in image space. The method comprises of following steps:
1. acquisition of serially stacked images of the patient scan in accordance with DICOM standards of a tomography system,
2. generation of dimensionally true point cloud representation of the vertebrae,
3. automatic determination of medically significant areas and the centre of the critical section of pedicles,
4. automatic determination of the spatial position and orientation of the vertebrae,
5. automatic determination of appropriate sizing of pedicle screw and orientation of pedicle screw axis with respect to the pose of vertebrae,
6. display of Machine-independent Multi-Planar Reconstruction (MiMPRs) of pedicles of relevant vertebrae.
[007] The advantageous embodiment of the aforementioned method is patient-specific and an intensity-independent approach. In other words, this enables and provides same visualization independent of 16 bits, 12bits, or 8-bit depth display units.
[008] In accordance with the method of present invention, a point cloud of vertebra boundary is obtained using computer implement technique comparing multiple filters and edge scan method. Subsequently, the pedicle region is segmented and partitioned into left and right pedicle.
[009] Once segmented point cloud of each pedicle region is obtained, the method further segments the point cloud into multiple coronal section. Each of the section segregates point cluster which is used to determine the geometry properties like area, area-centre, second moment of area. The collection of area centroids of progressive sections will enable to generate the pedicle-screw axis converging to the optimum pedicle screw axis. The estimation reduces the computation time required for searching the optimum axis, thus improving efficiency and accuracy in subsequent steps. In this context, it should be noted that initially chosen coronal sections are parallel to the tomographic machine XZ plane (the seed axis, that is the 0th iteration considers axis parallel to the machine Y-axis).
[010] In further embodiments of the method, additional steps are implemented to determine the accurate pedicle-screw axis. These steps comprise of the following.
1. Pedicle sections perpendicular to the new pedicle axis is considered. In each of the section, the geometric properties of the pedicle boundary is determined.
2. The section corresponding to smallest sectional area is identified. The position of the area-centroid is stored as the critical section.
3. Further, the area-centroid of the neighbouring sections adjacent to the critical section on either side were found and recorded.
4. Establishment of the pedicle screw axis is performed based on the area centroid of the critical section and area-centroids of the neighbouring sections on either side of critical section.
The analysis of pedicle coronal area and utilization of neighbouring section data aid in achieving accurate pedicle axis which passes medially to narrowest region of the pedicle.
[011] In relation to the automatic determination of appropriate sizing of pedicle screw and orientation of pedicle screw axis, after we optimised the location pedicle isthmus and its orientation, a further step is taken to optimise the direction of pedicle screw and diameter. The motive to implement henceforth technique is that the screw should be at equal distance from either side of the boundary. A technique employed is to expand the radius of the circle in the critical section until it intercepts either side of the cortical boundary (sectional boundary). On interception on any side, the centre of the circle is marginally shifted diametrically opposite to the wall. The above procedure is repeated till the centre lies equidistant from the boundaries with in a small predetermined margin to obtain the maximum screw diameter. This process ensures correct direction and aligns the pedicle screw within the pedicle isthmus boundaries.
[012] In relation to automatic creation of MiMPR of pedicle of the vertebra, inventors propose novel method that involves autonomous generation the planes that contains pedicle screw axis called as true-sagittal and true-axial section of pedicle. The invention also discloses a method to generate the plane that is normal to the pedicle screw axis called as true-coronal section of the pedicle. This approach aims to enhance the accuracy and perception of the pedicle structure, thereby improving the overall quality and reliability of the surgery.
[013] In relation to autonomous determination of the spatial position and orientation of the vertebrae, inventors propose a novel method for the autonomous identification of four body- landmarks. Based on these landmarks, a machine-independent vertebral coordinate system (VCS) is established. The unique, patient specific and accurate definition of VCS sets the MiMPRs independent of DICOM. The VCS is known with respect to Machine Coordinate System (MCS). The MCS is a coordinate system which is attached to imaging device. It localises every pixel in DICOM images. The pedicle-screw axis is also known with respect to MCS. Therefore, orientation and location of pedicle-screw axis can be evaluated with respect to VCS. The advantage with this approach is that it determines orientation of pedicle-screw axis with respect to true position and orientation of corresponding vertebrae. The present invention is based on analytical and logical assessment. The invention provides a deterministic and reproducible alternative approach to the existing intuitive and subjective approach. The approach enables an assessment based, accurate, image guided screw placement surgery. The domain professionals will further utilize the above non-ambiguous and autonomous features of the invention.
BRIEF DESCRIPTION OF THE DRAWINGS
Embodiments of the invention will be described in more detail with the aid of the figures in the following text, with the figures illustrating only those features which are necessary
for understanding of embodiments of the invention. The following reference symbols are used in this case: 0: imaging device like stationary CT scanner, mobile CT scanner etc. 1: machine coordinate system; 2: X-ray tube; 3: detector; 4: gantry housing; 5: patient; 6: movable patient couch; 7: computer workstation; 8: scan area; 9: stack of images of L2 vertebra; 10,11:Geometric centre of spinal canal; 12: Geometric centre of vertebral body; 13:geometric centre of pedicle isthmus; 14: ROI enclosed by rectangle; 15: coronal Area variation of left pedicle of vertebra; 16: coronal Area variation of right pedicle of vertebra; 15A: The location corresponding to minimum area of left pedicle; 16A: The location corresponding to minimum area of right pedicle; 17: line joining area-centre of SC and centre of circle; 18: a coronal area of left pedicle represented in form of point cluster; 19: a coronal area of left pedicle represented in form of point cluster C0 : isocentre of machine coordinate system ROI: Region of interest; M{SP}_P: Mid sagittal plane of pedicle; M{CP}_P: Mid coronal plane of pedicle; S: machine Sagittal plane; L1 to L5: lumbar vertebrae;P_L: Pedicle, left; P_R: Pedicle, right; T1 to T12: thoracic vertebrae; VB: vertebral bodies;
{W1}: working vertebra coordinate system in pixel space;{W2}: working spinal canal coordinate system in pixel space; u: axes along column of image in pixel space; v: axes along row of image in pixel space; VBB: Vertebral body outer boundary; SCB: Spinal canal outer boundary; MCS: Machine Coordinate System; {\hat{X}}_C, {\hat{Y}}_C, {\hat{Z}}_C: principal axes of imaging device; ; {\hat{X}}_L, {\hat{Y}}_L, {\hat{Z}}_L: principal axes of Left pedicle; {\hat{X}}_R, {\hat{Y}}_R, {\hat{Z}}_R: principal axes of Right pedicle; {\hat{X}}_V, {\hat{Y}}_V, {\hat{Z}}_V: principal axes of Vertebra
In detail, in the figures:
FIG. 1: shows a schematic illustration of an imaging device with coordinate system at isocentre,
FIG. 2: shows a side view and front view of a complete spinal column with a scan area;
FIG. 3: shows a schematic illustration of a vertebra where in case 1. it is ideally placed and in case 2. in general, it is rotated axially with respect to machine coordinate system (0);
FIG. 4: shows an illustration of autonomous rectangle enclosure of the region of the vertebra on axial slice of the scan area from FIG. 2;
FIG. 5: shows an illustration of cortical bone detected within rectangle obtained from FIG.3;
FIG. 6: shows an illustration of (A) fixing of working coordinate system{W1}at approximate centre of vertebra and (B) fixing of working coordinate system {W2} at approximate centre of spinal canal from FIG. 4;
FIG. 7: shows an illustration of autonomous detection of outer edge of a vertebra (VBB) and spinal canal (SCB);
FIG. 8: shows (A) partition of boundary obtained in FIG.6 by joining centre of Spinal canal and vertebral body and (B) axially stacked 3D point cloud of left and right pedicle of patient;
FIG. 9: shows (A) an illustration for obtaining multiple parallel coronal section, over the data shown in FIG.8(B), to obtain optimum Pedicle screw axis;
FIG. 9: shows (B) an illustration a sectional area;
FIG. 10: Shows variation of computed area of the 2D coronal section of left and right pedicle the pedicle boundaries, obtained from FIG. 8(B);
FIG. 11: shows an illustration of true Coronal section of pedicle and machine coronal section of pedicle;
FIG. 12: shows a illustration of iterative solution to obtain optimum pedicel screw;
FIG. 13: shows a visualization of pedicle isthmus and a solution to obtain optimum pedicle screw diameter;
FIG. 14: shows a formation of vertebral coordinate system and visualisation of pedicle isthmus and mid-sagittal section.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

[014] The method for autonomous establishment of machine-independent pedicle principal sections to obtain optimum pedicle-screw axis in direction, length, and safety margin in accordance with the present invention are in more detail hereinafter.

[015] Understanding of invention
[016] As described herein, the present invention is a method for autonomous establishment of machine-independent principal planes of pedicle structure. The principal planes are so formed that its intersection contains the pedicle-screw axis. FIG. 1 shows an example of a computer tomography (CT) system, in this case a stationary CT system 0, by which the method according to an embodiment of the invention can be carried out. The CT system has a gantry housing 6 in which a CT machine coordinate system {C} is attached at isocentre 3, while the patient 5 in prone position is moved on a movable patient couch 6 along the system \widehat{\mathbit{Z}_\mathbit{c}}-\ axis 7. Few valid standard practices about CT machine coordinate frame are that the \widehat{\mathbit{Y}_\mathbit{c}}- axis aligns with gravity axis and therefore XZ plane is normal to gravity axis. In DICOM, the measurements are done with respect to machine coordinate frame {C}. The three sections of CT-machine coordinate system can be thought of as approximately correspond to body principal planes (a) XY plane is an axial plane (b) YZ plane is sagittal plane and (c) ZX plane is a Coronal plane (Note: the classical names are strictly in reference to the body, only for the purpose of gross reference these correspondences are used).

[017] Each slice of the CT scan will be referred to the {C}. The notation correspondence of body based coordinate system in relation to DICOM are, XY referred as an axial, YZ as a sagittal, and XZ as a coronal plane. Stacking planes along any of the three axes will result in 3D visualisation. In surgery, knowing the true orientation of the region of interest (ROI), for example vertebra, is critical. Therefore, a body-feature based coordinate system becomes extremely beneficial for accurate perception.

[018] FIG. 3 depicts an ideal case, case 1, where vertebral longitudinal axis aligns with longitudinal \widehat{Z_c}-\ axis of CT scan. FIG. 3, case 2 depicts a general case, where a vertebra is tilted in space with respect to frame {C}. The present invention autonomously generates the sections which are along true vertebral axes and also principal section of each pedicle which takes care of true orientation of pedicle and geometrical properties. These sections are based on body-feature landmarks and henceforth, are independent to frame {C}.
[019] Steps included in invention
[020]The present invention includes the following processes: importing data collected from imaging device like stationary CT scanner, mobile CT scanner; pre-processing the data; isolating the region of interest; extraction of boundaries of the anatomic body on each axial section; optimizing implant direction with consideration of geometrical aspect; generating machine-independent planes based on local coordinate system; computing a minimum pedicle diameter, a maximum implant diameter with consideration of recommended safety margin; optimizing implant maximum length with consideration of geometrical aspect.
STEP 1
[021] Importation of medical data
[022] Medical data is first obtained from imaging device, for proof of present invention stationary computed tomography scan (CT) is chosen, for the study of the spine area of interest 8, as shown in FIG.2A. The data is obtained from imaging device like a stationary computed tomography scan (CT), mobile computed tomography scan, CT capable fluoroscopy (3D) or similar imaging system. FIG. 1 shows a similar device. FIG. 2B illustrates the data consisting of serially stacked 2D images(sections) 9 of L2 vertebra using DICOM tags. The method of present invention recommends that a thin cut section should be obtained to increase accuracy of result and geometrical detail of ROI.
[023] Pre-processing of data
STEP 2:
[024] Isolation of ROI
[025] FIG. 4 depicts an intensity-based data in form of image of size 512×512 pixels, obtained from a CT scan of L2 vertebra used in accordance with the present invention. A CT image coordinate system {I} is attached to left corner of image, where u-axis represents the increasing columns, and v-axis represents the increasing rows, of the image pixels. A rectangular box 14 depicts the smallest enclosed area of vertebral region (ROI).
[026] Now, a stack 9 of 2D images is created which includes the pedicle of vertebra, say L2 vertebra. Once the serially stacked 2D images of the vertebra is obtained, autonomous isolation of vertebral region is performed over each section, using multiple filters, and finally a rectangular region 14, as shown in FIG 4, is defined. The coordinates (u, v) of the all the corner of rectangle is temporarily recorded and cropped.

STEP 3:
[027] Detection of bony region
[028] FIG. 5 illustrates an image with intensity value of either zero or one. Zero intensity value is shown by black pixels while the intensity value, one is shown by white pixels. White pixels represent the denser part like cortex bone and black pixels represents the porous portion like air, tissue, etc. The image has same size as that of rectangle in {I} frame.
FIG. 5 comprise of illustration of the left pedicle region \mathbit{P}_\mathbit{L} and the right pedicle region \mathbit{P}_\mathbit{R}, spinal canal SC and vertebral body VB. The image is obtained after an edge detection technique, for example canny edge detection, is applied within the enclosed region of rectangle box. The image is further padded with zero intensity value to make size 512×512 pixels.

STEP 4:
[029] Anatomical boundary extraction
[030] In the present invention, a method has been invented to extract the outer boundary of the vertebral body and the spinal canal. FIG. 7 shows the resulting image illustrating the extraction of boundary of cortical bone, which comprise segmented vertebral boundary VBB, and the spinal canal boundary SCB.
[031] The method suggests that detection of pixels corresponding to boundary of anatomic body can be performed by discrete scanning as if a ray is emerging from the origin of the coordinate system. The method first sorts the pixels lexicographically, first capturing v-coordinate extents, then capturing u-coordinate extents.
[032] To extract a vertebral boundary VBB, a working coordinate system {W1} is attached at approximate centre of VB (shown in FIG. 6A). The method selects an interior point to be the origin of frame {W1}. Confirmation of detecting an interior point of VB is performed using Hough’s man circle detection technique. (Note: Any point inside the VB can be chosen as the origin but for the reason of geometrical simplicity and to obtain the resulting axis in alignment with vertebra sagittal symmetry, the centre of the circle that closely aligns with vertebral body is chosen as the origin.) Similarly, to extract a spinal canal boundary SCB, a working coordinate system {W2} (shown in FIG 6B) is attached at an approximate centre of SC. A similar ray scan technique is used to detect boundary pixels but in this case the pixels that are closest to centre are chosen to be the boundary pixels.
STEP 5:
[033] Segmentation of pedicle region, partition of pedicle: left and right, and creation of pedicle point cloud
[034] Segmentation of SBB facilitates in defining location of extreme top-most pixel and bottom-most pixels along v-axis of the image. Further, the area-centre of segmented SCB is defined. Then, a line 17 is constructed, joining area-centre of SCB and an approximate centre of vertebra (centre of fitted circle). FIG. 8(A) is the illustration of the area-centre of SCB and an approximate centre of vertebra, and a line 17. A rectangle enclosing the pedicle region, and oriented along the line. The rectangle is constrained to pass (u, v) coordinates of extreme pixels of spinal canal (FIG. 6B). Initially, the line, segments the pedicle into two parts: the left and the right. As per the patient’s left and right, one side of the line covered the left pedicle region, while the other side of the line covered the right pedicle region. This procedure is applied to each of axial section of chosen stack 9 of L2 vertebra.
[035] Once left pedicle and right pedicle is defined independently, the location of each pixel of pedicle in frame {I} is recorded and transformed to frame {C}. These locations are termed as: planar point cluster of the pedicle of XY image plane. The planar point cluster is stacked together along the Z-axis of frame {C}. This stack of planar points forms a 3D pedicle cluster referred as pedicle point cloud, as shown in the FIG. 8(B).
[036] Defining the axis of pedicle screw or similar implant
After completing the pre-processing of images, the image data are represented in point cloud with respect to physical frame, in this case CT machine frame {C}. Now, the present invention discloses a methodology to find optimum pedicle-screw axis. The methods constitute of following steps:
finding of area-centroid of pedicle region in XZ plane of frame {C},
finding of seed axis of pedicle,
iterative solution scheme to discover optimum pedicel-screw axis.
STEP 6:
[037] Autonomous triangulation of pedicle region, computation of the area and its area-centroid.
[038] FIG. 9(A) shows an illustration of section constructed normal to Y axis of frame {C}. These sections are separated by spacing provided in DICOM tag along Y axis of frame {C}, and will be referred as XZ plane (coronal layers) of frame {C}.
[039] In accordance with present invention, the method defines initially the pedicle-screw axis as seed axis using area-centroid of pedicle region in successive coronal layers. The method is executed in following way. Multiple parallel XZ plane (layers) is reconstructed which segregate all the points belong to a particular Y value (or point cluster in particular XZ plane), from a 3D pedicle point cloud. Accurate pedicle region is constructed by triangulating the 3D pedicle point cloud layer by layer. The triangulated pedicle region is utilized to accurately determine the physical and geometrical features of the pedicle. The triangulation methods are described herein. Firstly, a largest inner convex hull (FIG. 13) for each layer is formulated. Secondly, identify the mean of the convex hull, and third, arrange the data in a counter-clockwise direction. Thereafter, constructing triangles such that one vertex of all triangles is at the mean. FIG. 9(B) illustrates a layer of coronal The triangulated region thus constructed contributes in finding many important data like area, area centroid, eigenvectors, etc., accurately.
[040] Further, area and its corresponding area-centroid is calculated using standard mathematical equation. Suppose that the boundary of the pedicle is the piecewise linear path formed by traversing the points \left[\left(x_i,y_i\right)\right]_{i=1}^m, where m is total no. of points in a plane (FIG. 9B). A triangulated 2D surface is created by joining each edge with a centre of convex hull (FIG. 13). This procedure is repeated for both left and right pedicle, separately. The location of each area in the layer is assigned by location of its area-centroid (FIG. 9(B)) in frame {C}. These locations are sorted lexicographically by Y-coordinate.
STEP 7:
[041] Identification of pivot point for the axis
[042] FIG. 10 illustrates the plot of pedicle area as a function of Y-coordinate of its area-centroid in the case study of L2 vertebra. The ordinate is the area corresponding to the triangulated region of each XZ plane. The abscissa corresponds to the Y-coordinate of each of the area-centroid. The transverse area of right pedicle is shown by the curve 15 while the transverse area of left pedicle is shown by the curve 16. It is important to note that the label at which the transverse area goes minimum is not the same for both the pedicle. The result suggests the possibilities that: (a) the geometry of left and the right pedicle is not mirror image of each other, or (b) the vertebra is rotated about Z -axis of frame {C}, or (c) combined effect of both (a) and (b). It is crucial to identify and account for such differences during surgery. The location 15A and 16A corresponding to minimum sectional area occurs at area-centroid and referred to as the pivot point for the initial seed axis of the pedicle-screw. The initial seed axis of the pedicle-screw is parallel to {\hat{Y}}_C of frame {C}
STEP 8:
[043] Iterative solution scheme to discover optimum pedicel-screw axis
[044] The pedicle volume has an asymmetric structure, and its principal axis is not aligned with any of the orthogonal axes of frame {C}. In order to overcome this difficulty and to provide better perspective, an axis is defined to approximate local symmetries of the pedicle structure. In this case, innovators have divided the pedicle structure into 3 regions. The central region is the narrowest region and the other two on either side has diverging and bigger region.
[045] Step 6 and step 7 discloses computation of minimal area and its location with respect to frame {C}. An iteration is initiated along {\hat{Y}}_C , the minimal area at narrowest region of the central region is found. Four layers adjoining the minimal area on either side in the narrow region is considered for further trending axis (FIG. 12). The trending axis is obtained by passing a line through the area centroid, noted as the pivot point of the minimal area and pass through area-centroids of adjacent layers within the narrow region on either side in the least perpendicular distance. The method of invention selects only those area-centroid that are within that narrow region. Thereafter, a covariance matrix is formulated. A solution to a covariance matrix yields 3 vectors. The vector with the highest variation serves as the seed axis of pedicle. The axis so determined is noted as {\hat{Y}}_L(1) (FIG. 12). Table 1 shows the magnitude of minimal area obtained in each iteration for left pedicle of L2 vertebra. The least area thus obtained is called as pedicle isthmus and its location is called as geometric-centre of the pedicle. The axis along which pedicle isthmus is found is called a optimum pedicle screw axis, in this case {\hat{Y}}_L(2).
STEP 10
[046] Determination of minimum diameter of pedicle and maximum diameter of pedicle screw
[047] Once the pedicel isthmus is defined, a circle is drawn at the pedicle isthmus. The radius is increased till it touches the either side of the isthmus boundary, largest inner convex hull of the isthmus. FIG. 13 illustrates the procedure. The midpoint of the isthmus is obtained by iterating the position of the centre from the boundary wall. The positioning of the centre is carried out till the circle is tangential to either side of the boundary. The maximum diameter of pedicle screw is determined by allowing an optimum safety margin.
STEP 11
[048] Method to localize Vertebral and Pedicle Coordinate Systems (VCS & PCS) in DICOM image space
[049] CT Machine Coordinate System:
C (\widehat{X_c},\widehat{Y_c}\ and\ \widehat{Z_c}) is a Machine Coordinate System (MCS), and Co is the origin of the MCS. Few valid standard practices about CT Machine Coordinate System are that the Y- axis aligns with gravity axis and therefore XZ plane is normal to gravity axis. In DICOM, the measurements are done with respect to machine coordinate system (MCS).
The three sections of CT-machine coordinate system can be thought of as correspond to body principal planes (a) XY plane is an axial plane (b) YZ plane is sagittal plane and (c) ZX plane is a coronal plane (Note: the classical names are strictly in reference to the body, only for the purpose of gross reference these correspondences are used).
[050] Vertebra Coordinate System (VCS)
[051] FIG. 14 illustrates the establishment of VCS, also illustrates the mid sagittal as well as the mid coronal section of the pedicle.
[052] {_^\mathbit{C}}\mathbit{V}\left(\mathbit{i}\right)\left({_^C}{\hat{X}}_{V_i\ },{_^C}{\hat{Y}}_{V_i\ },{_^C}{\hat{Z}}_{V_i\ }\right): {_^\mathbit{C}}\mathbit{V}\left(\mathbit{i}\right) stands for ith Vertebra Coordinate System (VCS) described with respect to {C} in image space. Vo is the origin of corresponding vertebra. {\hat{X}}_{V_i\ },{\hat{Y}}_{V_i\ }and\ {\hat{Z}}_{V_i\ } are the unit vectors along the principal directions of VCS. When written with respect to MCS, {C}, they are represented as {_^C}{\hat{X}}_{V_i\ },{_^C}{\hat{Y}}_{V_i}\ and\ {_^C}{\hat{Z}}_{V_i} .A coordinate frame is attached to the vertebra, based on the vertebra body-feature landmarks, by autonomous operation of the processor. The establishment of {_^\mathbit{C}}\mathbit{V}\left(\mathbit{i}\right) is based on:
S_C\left(i\right) is the position vector of centre of vertebral canal of the vertebra with respect to MCS. It is a unique feature point, identified and measured autonomously in image space. This point is taken as the origin, Vo, of the vertebra.
B_C\left(i\right) is the centre of vertebral body of the vertebra with respect to MCS. It is a unique virtual feature point, identified and measured autonomously.
L_C\left(i\right) and R_C\left(i\right) are the two feature points, they are position vectors of the right and left centres of pedicle isthmus with respect to MCS. They are identified and measured autonomously.
For the ith vertebra frame:
The coordinates of feature points, S_C\left(i\right) and B_C\left(i\right) are measured in the image and recorded. A unit vector, directing along S_C\left(i\right) to B_C\left(i\right) to define {_^C}{\hat{Y}}_{V_i\ }is processed. Similarly, establish a temporary unit vector directing along S_C\left(i\right) to R_C\left(i\right) to define a temporary unit vector, \ {_^C}{\widehat{X^\prime}}_{V_i} . Following this establish {_^C}{\hat{Z}}_{V_i} by taking the cross product {_^C}{\widehat{X^\prime}}_{V_i}\times{_^C}{\hat{Y}}_{V_i\ }. Finally, formulated the orthogonal VCS with respect to {C} by {_^C}{\hat{X}}_{V_i\ }= {_^C}{\hat{Y}}_{V_i\ }\times{_^C}{\hat{Z}}_{V_i\ }
When we stack three unit vectors as column of 3x3 matrix, in order {_^C}{\hat{X}}_{V_i\ },{_^C}{\hat{Y}}_{V_i}\ and\ {_^C}{\hat{Z}}_{V_i} , we will get rotation matrix which describes the ith VCS with respect to {C} and represented as {{_V_i^C}ROT}_ .
{{_V_i^C}ROT}_=\left[{_^C}{\hat{X}}_{V_i\ }{_^C}{\hat{Y}}_{V_i}\ \ {_^C}{\hat{Z}}_{V_i}\right]=\left[\begin{matrix}{{_^}X}_{V_i}^ \bullet{{_^}X}_C&{{_^}Y}_{V_i}^ \bullet{{_^}X}_C&{{_^}Z}_{V_i}^ \bullet{{_^}X}_C\\{{_^}X}_{V_i}^ \bullet{{_^}Y}_C&{{_^}Y}_{V_i}^ \bullet{{_^}Y}_C&{{_^}Z}_{V_i}^ \bullet{{_^}Y}_C\\{{_^}X}_{V_i}^ \bullet{{_^}Z}_C&{{_^}Y}_{V_i}^ \bullet{{_^}Z}_C&{{_^}Z}_{V_i}^ \bullet{{_^}Z}_C\\\end{matrix}\right]
[053] Definition of {_^\mathbit{C}}\mathbit{L}\left(\mathbit{i}\right)\left({\hat{X}}_{L_i\ },{\hat{Y}}_{L_i\ },{\hat{Z}}_{L_i\ }\right) and {_^\mathbit{C}}\mathbit{R}\left(\mathbit{i}\right)\left({\hat{X}}_{R_i\ },{\hat{Y}}_{R_i\ },{\hat{Z}}_{R_i\ }\right) :
are the left and right pedicle coordinate system (PCS) with respect to {C} respectively specific to the pedicles. The method for defining the PCS is same for both left and right pedicles. The discussion is elaborated for left PCS. The origins, Lo and Ro of the each of the PCS are autonomously, uniquely defined at area-centroids of the pedicle isthmus (critical cross-sectional area). Two axes ( {\hat{X}}_{L_i\ },{\hat{Z}}_{L_i\ }) lie in the plane containing the critical cross-sectional area and the third axis, {\hat{Y}}_{L_i\ } which is normal to the critical sectional area, is the cross-product of the unit vectors of the first two axes. The plane {\hat{Z}}_{L_i\ }-\ {\hat{X}}_{L_i\ } can be seen as the pedicle coronal, {\hat{Y}}_{L_i\ }-\ {\hat{Z}}_{L_i\ } as the pedicle sagittal and {\hat{X}}_{L_i\ }-{\widehat{\ Y}}_{L_i\ }as the pedicle axial sections of the left specific pedicle. These coordinate axes are referred as the principal axes of the pedicle which defines Pedicle Coordinate System (PCS). The mathematical formulation for describing PCS comprises
PCS with respect to ith vertebra frame is as follows.
{_^\mathbit{C}}\mathbit{L}\left(\mathbit{i}\right)\left({\hat{X}}_{L_i\ },{\hat{Y}}_{L_i\ },{\hat{Z}}_{L_i\ }\right) = Vo + {{_V_i^C}ROT}_ {{_^}\left[{_^\mathbit{V}}\mathbit{L}\left(\mathbit{i}\right)\left({\hat{X}}_{L_i\ },{\hat{Y}}_{L_i\ },{\hat{Z}}_{L_i\ }\right)\right]}_^T
{{_^}\left[{_^\mathbit{V}}\mathbit{L}\left(\mathbit{i}\right)\left({\hat{X}}_{L_i\ },{\hat{Y}}_{L_i\ },{\hat{Z}}_{L_i\ }\right)\right]}_^T = {{_^}\left[{{_V_i^C}ROT}_\right]}_^{-1} [{_^\mathbit{C}}\mathbit{L}\left(\mathbit{i}\right)\left({\hat{X}}_{L_i\ },{\hat{Y}}_{L_i\ },{\hat{Z}}_{L_i\ }\right) - Vo(x,y,z) ]
{{_^}\left[{_^\mathbit{V}}\mathbit{L}\left(\mathbit{i}\right)\left({\hat{X}}_{L_i\ },{\hat{Y}}_{L_i\ },{\hat{Z}}_{L_i\ }\right)\right]}_^T: ith Left pedicle coordinate system with respect to its vertebra frame {V}.
Similarly, {{_^V}\mathbit{R}\left(\mathbit{i}\right)\left[\left({\hat{X}}_{L_i\ },{\hat{Y}}_{L_i\ },{\hat{Z}}_{L_i\ }\right)\right]}_^T: ith Right pedicle coordinate system with respect to its its vertebra frame {V}.
,CLAIMS:CLAIMS:
1. Autonomous establishment of machine independent pedicle principal sections and to obtain optimum pedicle-screw axis in direction, length and safety margin, comprises:
- accurately determining, by the processor, a pedicle-screw axis for facilitating the placement of pedicle screws in the patient, by constructing a computer implement technique.
- performing, by the processor, Machine-independent Multiplanar Reconstruction (MiMPRs) of patient pedicle with the use of a first computer implement technique
- establishing, by the processor, a body-feature Vertebra Coordinate System (VCS) and determine the pose of the vertebra of a patient with respect to tomographic Machine Coordinate System (MCS) with the use of a second computer implement technique
- determining, by the processor, the diameter of the pedicle screw allowing an optimum safety margin with the use of a third computer implement technique.
2. The method as claimed in claim 1, wherein the computer implement technique is used for the assessment of medical imaging data, wherein the medical imaging data comprises an image of patient’s spine and the like.
3. The method as claimed in claim 2, wherein the medical imaging data are chosen from the set such as stationary computed tomography (CT) system, mobile computed tomography (CT) system, positron emission tomography- computed tomography (PET-CT) system.
4. The method as claimed in claim 2, wherein the medical imaging data are intensity-based data, wherein the intensity data consists of number of pixels, which are located with respect to tomographic machine coordinate system (MCS).
5. The method as claimed in claim 1, wherein determining, the pedicle screw axis for facilitating the placement of pedicle screws in patient comprises:
a) Segmenting, by the processor, individual vertebrae boundaries
b) creating, by the processor, a point cloud of the outer cortical wall of the vertebra by taking axial sections of the vertebra with respect to MCS and stacking them in sequence,
6. The method as claimed in claim 5, wherein further the segmentation and creation of point cloud comprises:
a) Isolating a stack of serially arranged axial sections of imaging data of the vertebra,
b) Isolating further, by the processor, a vertebral region of interest of imaging data,
c) identifying, by the processor, the dense cortical bone of the vertebra
d) extracting, by the processor, the boundary of the cortical region
e) representing, by processor, the vertebra in 3D point cloud with respect to MCS
7. The method as claimed in claim 1, wherein determining, the pedicle screw axis for facilitating the placement of pedicle screws in patient comprises:
a) a processing steps and a technical solution to partition the left-right pedicles of the vertebra
b) a technical solution to autonomously identify pedicle isthmus of left-right pedicles of the vertebra
c) a processing steps and a technical solution to identify the center of the pedicle isthmus which acts as a pivot point for the pedicle screw axes

8. The method as claimed in claim 1, wherein establishing Vertebra Coordinate System comprises autonomously annotating body feature landmarks, wherein the second computer implement technique comprises:
a) identifying, by processor, left and right pedicle isthmus, area centroid of minimal area with respect to MCS,
b) identifying, by processor, center of spinal canal with respect to MCS,
c) identifying, by processor, geometric center of vertebral body with respect to MCS.

9. The method as claimed in claim 1, wherein performing a Machine-independent Multiplanar Reconstruction (MiMPRs) of the patient pedicle, by the processor, comprises:
a) Determining the pose of the pedicle axis with respect to VCS
b) Autonomous generation of the principal planes that contains pedicle screw axis. The principal planes, XY, YZ and ZX of Pedicle Coordinate System (PCS) represents true-axial, true-sagittal and true-coronal sections respectively of the pedicle,

10. The method as claimed in claim 9, wherein the determining the pose comprises novel processing solution
a) calculating, by the processor, the area and geometric center of the pedicle isthmus; using a third computer implement technique.
b) The solution provides isolated representation of pedicle structure.