Modeling method of respiratory model and synchronous respiratory tracking method of CyberKnife

Abstract

The present application discloses a method for modeling a breathing model and a synchronous breathing tracking method of a CyberKnife, the modeling method comprising: obtaining a first breathing curve of a target object when breathing freely, and constructing a first breathing model based on free breathing; obtaining a second breathing curve corresponding to the target object when breathing with the assistance of at least one assisted respirator, and constructing a second breathing model based on each assisted respiration; obtaining corresponding statistical parameters based on the breathing models of all breathing modes; comparing the statistical parameters of all breathing modes, obtaining an optimal breathing model based on a selection rule, and using the breathing mode corresponding to the optimal breathing model as the optimal breathing mode. The present application selects the breathing mode that best suits the target object from the breathing models established by free breathing and multiple assisted respirations, avoids inaccurate construction of the breathing model caused by involuntary deep breathing, can accurately predict the target position, reduce rigid body errors, and shorten treatment time.

Description

Technical Field

[0001] The present application relates to the technical field of stereotactic radiotherapy, and more specifically, to a modeling method of a respiratory model and a synchronous respiratory tracking method of a CyberKnife. Background Art

[0002] Cyber ​​Knife, also known as "stereotactic cyber surgery platform", "network knife" or "computer knife", is a new type of whole body stereotactic radiosurgery treatment equipment. Cyber ​​Knife robotic radiosurgery system is a technology that can replace surgery and perform non-invasive treatment on malignant and non-malignant tumors (hereinafter referred to as target tissues) in any part of the body (including prostate, lungs, brain, spine, liver, pancreas and kidneys).

[0003] During CyberKnife treatment, the position of the target tissue changes with the patient's respiratory cycle. Therefore, during treatment, a respiratory model needs to be constructed to achieve synchronous respiratory tracking of the target tissue so that the radiation beam can accurately reach the target tissue.

[0004] The current CyberKnife synchronous breathing tracking technology is a free breathing tracking technology, in which the target subject has a high degree of freedom in breathing. However, for target subjects with poor lung function, it is difficult to avoid involuntary deep breathing during free breathing. In this case, involuntary deep breathing during the construction of the breathing model will lead to unstable breathing amplitude, large mutations, irregular breathing curves, and large rigidity errors in the constructed breathing model. When the threshold is exceeded, the model needs to be remodeled, which increases the treatment time. During the treatment process, the breathing prediction value obtained based on the breathing model is inaccurate, the treatment rigid body error is large, and the actual amplitude of the target subject's free breathing may exceed the amplitude of the breathing curve, resulting in treatment interruption. Summary of the invention

[0005] The present application provides a method for modeling a respiratory model and a synchronous respiratory tracking method of a CyberKnife. By comparing the statistical parameters between the respiratory models established by free breathing and multiple assisted breathing, the most suitable respiratory mode for the target object is selected to avoid inaccurate respiratory model construction caused by involuntary deep breathing. At the same time, personalized customization of the target object's respiratory mode is achieved. When the target object needs it, the respiratory rhythm is controlled through external intervention to stabilize the amplitude and obtain the optimal respiratory curve, thereby establishing the optimal respiratory model, which can accurately predict the target position, reduce rigid body errors, shorten treatment time, and improve treatment accuracy and efficiency.

[0006] The present application provides a modeling method of a respiratory model, comprising:

[0007] Acquire the first breathing curve of the target object during free breathing, and construct a first breathing model based on free breathing;

[0008] Acquire a second breathing curve corresponding to the target object breathing with the assistance of at least one assisted breathing device, and construct a second breathing model based on each assisted breathing;

[0009] Obtain corresponding statistical parameters based on the breathing models of all breathing modes;

[0010] The statistical parameters of all breathing modes are compared, and the optimal breathing model is obtained according to the optimal selection rule, and the breathing mode corresponding to the optimal breathing model is taken as the optimal breathing mode.

[0011] Preferably, for each breathing mode, constructing a breathing model includes:

[0012] Repeat the following steps until the coverage is greater than the preset value:

[0013] In each respiratory cycle, first CT slice and contrast images and first orthogonal images are acquired at a plurality of target phase points according to the real-time body position of the target object, and a first digitally reconstructed radiographic image is generated according to the first CT slice and contrast images;

[0014] calculating a first error between a position of the tumor marker in the first orthogonal image and a position of the tumor marker in the first digitally reconstructed radiographic image at each target phase point in the same respiratory cycle;

[0015] Calculate the coverage rate of the breathing cycle according to the proportion of target phase points whose first error is less than or equal to the threshold in the same breathing cycle to all target phase points;

[0016] If the coverage rate is less than the preset value, for the target phase point whose first error is greater than the threshold, the shooting position of the orthogonal planar film system at the target phase point in the next respiratory cycle is corrected according to the first error at the target phase point;

[0017] The first error curve of the entire breathing cycle is fitted according to the first errors of all target phase points in the last breathing cycle as the breathing model of the breathing mode.

[0018] The phase point is determined by the breathing curve.

[0019] Preferably, the tumor marker is a body surface marker positioned on a flat portion of the body surface projection area of ​​the treatment site of the target subject or a gold marker implanted in the treatment site.

[0020] Preferably, the optimization rule is:

[0021] First, find a qualified breathing model;

[0022] Among the qualified respiratory models, the optimal respiratory model is screened in the priority order of standard error > correlation error. If there is only one respiratory model with the smallest standard error, then this respiratory model is selected as the optimal respiratory model. If there are two respiratory models with equal and smallest standard errors, the one with the smallest correlation error is selected as the optimal respiratory model.

[0023] Preferably, the ventilator assist device comprises a sound ventilator assist device and a visual ventilator assist device.

[0024] The present application also provides a method for synchronous breathing tracking of a CyberKnife, comprising:

[0025] According to the treatment part of the target object and the confirmed CT image of the target object, the target object is fixed in a personalized position;

[0026] Construct an optimal breathing model based on the breathing curve of the optimal breathing method;

[0027] During the treatment, the optimal breathing model is used to predict the irradiation position of the CyberKnife beam at the next moment;

[0028] The irradiation position is sent to CyberKnife so that the irradiation position of the radiation beam moves synchronously with the treatment part caused by the breathing of the target subject.

[0029] Preferably, the CyberKnife synchronous breathing tracking method further comprises:

[0030] During treatment, the breathing curve is updated according to the preset update cycle, and the breathing model is also updated at the same time.

[0031] Preferably, determining the optimal breathing mode specifically includes:

[0032] Acquire the first breathing curve of the target object during free breathing, and construct a first breathing model based on free breathing;

[0033] Acquire a second breathing curve corresponding to the target object breathing with the assistance of at least one assisted breathing device, and construct a second breathing model based on each assisted breathing;

[0034] Obtain corresponding statistical parameters based on the breathing models of all breathing modes;

[0035] The statistical parameters of all breathing modes are compared, and the optimal breathing model is obtained according to the optimal selection rule, and the breathing mode corresponding to the optimal breathing model is taken as the optimal breathing mode.

[0036] Preferably, obtaining a breathing curve specifically includes:

[0037] When the target subject is wearing a synchronized breathing tracking top, a plurality of infrared detection tracking markers are attached to the synchronized breathing tracking top;

[0038] Detects displacement changes of infrared detection tracking markers as the target object breathes;

[0039] The breathing curve of the target object is fitted based on the displacement changes of the infrared detection tracking marker.

[0040] Preferably, the optimization rule is:

[0041] First, find a qualified breathing model;

[0042] Among the qualified breathing models, the optimal breathing model is screened in the priority order of standard error > correlation error. If there is only one breathing model with the smallest standard error, the breathing model is used as the optimal breathing model. If there are two breathing models with equal and smallest standard errors, the one with the smallest correlation error is selected as the optimal breathing model. Other features and advantages of the present application will become clear through the following detailed description of exemplary embodiments of the present application with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the application and, together with the description, serve to explain the principles of the application.

[0044] Figure 1 A flow chart of a modeling method for a respiratory model provided in the present application;

[0045] Figure 2 This is a flow chart of the synchronous breathing tracking method of CyberKnife provided in this application. DETAILED DESCRIPTION

[0046] Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present application.

[0047] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present application, its application, or uses.

[0048] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0049] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0050] The present application provides a method for modeling a respiratory model and a synchronous respiratory tracking method of a CyberKnife. By comparing the statistical parameters between the respiratory models established by free breathing and multiple assisted breathing, the most suitable respiratory mode for the target object is selected to avoid inaccurate respiratory model construction caused by involuntary deep breathing. At the same time, personalized customization of the target object's respiratory mode is achieved. When the target object needs it, the respiratory rhythm is controlled through external intervention to stabilize the amplitude and obtain the optimal respiratory curve, thereby establishing the optimal respiratory model, which can accurately predict the target position, reduce rigid body errors, shorten treatment time, and improve treatment accuracy and efficiency.

[0051] It should be noted that the entire process of CyberKnife's synchronous breathing tracking treatment is: preparation and position fixation of the target object (i.e., the patient), determination of the optimal breathing model, and treatment by CyberKnife through synchronous breathing tracking according to the breathing model. These three stages are continuous, and the target object cannot adjust its position during this period. The breathing model is used to predict the target position of the CyberKnife beam and transmit it to the CyberKnife in real time, so that the CyberKnife moves synchronously with the breathing of the target object.

[0052] As an embodiment, the preparation and position fixation of the target object are as follows:

[0053] Step 1: Based on the target object's treatment site, select a relatively flat area for body surface marking. The body surface marking uses 6 0.5mm lead wires, which are arranged into a regular hexagon and attached to the body surface projection area using 3M dressings as tumor markers. Alternatively, based on the target object's treatment center location, a gold marker is implanted under CT / ultrasound guidance as a tumor marker.

[0054] Step 2: Wear a synchronized breathing tracking top of the corresponding size according to the target's body shape. The top is made of elastic fabric, round neck, cardigan, vest design, and the placket of the clothes is attached with Velcro, which fits the body surface and is easy to wear. The shape fit and high elasticity (moves with the target) can be provided in three sizes, namely S: height <160CM weight <60KG. M: height <175CM weight <80KG. L: height >180CM weight >80KG. A strip of Velcro is set on the surface of the top for sticking infrared detection and tracking MARK.

[0055] Step 3: According to the treatment site of the target object and the confirmed CT image of the target object, a negative pressure vacuum pad is used to fix the target object in a personalized position (supine or prone position) to ensure the stability of the detection target. At the same time, the pre-treatment CT tomography and contrast images with tumor markers are obtained according to the real-time position of the target object (the layer thickness is conventionally selected to be 1.5MM). Based on the CT tomography and contrast images, an orthogonal DRR image (digital reconstruction radiographic image) based on the pre-treatment CT is generated.

[0056] The modeling system includes a tracking marker harness, an interface box, and a synchronous breathing camera matrix, which cooperates with the numerical control system and the orthogonal flat-plate system on the CyberKnife device to establish the breathing model. One end of the tracking marker harness is connected to the infrared detection tracking MARK (preferably a transparent head with a transparent cable that will not interfere with the reference tracking), and the other end is connected to the numerical control system through the interface box to obtain power input.

[0057] The synchronous breathing camera matrix contains 3 LED detectors to detect and track the movement of the infrared detection tracking mark. When the infrared detection tracking mark moves with the patient's breathing, the signal from the synchronous breathing camera matrix can be used to determine the breathing curve. The breathing curve will be displayed on the display screen of the CNC system. The synchronous breathing camera matrix is ​​an adjustable matrix, which can be adjusted in time according to the position of the infrared detection tracking mark until the LED detector can fully detect the signal of the infrared detection tracking mark. The adjustment can be confirmed by the extremely fast flashing of each LED detector, or by viewing the breathing curve displayed by the CNC system.

[0058] In the present application, the numerical control system is mainly used to display the respiratory curve, model building, etc. It can be understood that the numerical control system is also used to perform the functions required in the CyberKnife device.

[0059] The orthogonal film system mainly includes two pairs of tube-image plate systems, each pair of which is in a 45-degree correspondence, ultimately forming an orthogonal film system for obtaining orthogonal images with tumor markers.

[0060] like Figure 1 As shown, the modeling method of the respiratory model provided in this application includes:

[0061] S110: Acquire a first breathing curve of the target object during free breathing, and construct a first breathing model based on free breathing.

[0062] S120: Acquire a second breathing curve corresponding to the target object breathing with the assistance of at least one assisted breathing device, and construct a second breathing model based on each assisted breathing.

[0063] As an embodiment, the ventilator includes a sound ventilator (eg, a metronome) and a visual ventilator (eg, a visual real-time breathing waveform), so that the target subject can breathe following the rhythm of the ventilator.

[0064] The breathing curve is composed of multiple breathing cycles similar to sine waves, and each breathing cycle contains eight equally distributed acquisition points (i.e., phase points) such as peaks and troughs. The breathing model collects at least 7 points and at most 15 points, and the breathing model is updated in real time starting from the 16th point.

[0065] S130: Obtain corresponding statistical parameters according to the breathing models of all breathing modes.

[0066] As an embodiment, the statistical parameters include breathing amplitude, error values ​​of three translational degrees of freedom (i.e., SUP / INF (up and down translation), LEF / RGT (left and right translation), ANT / POST (forward and backward translation)), error values ​​of three rotational degrees of freedom (i.e., LEF / RGT (left and right rotation), UP / Down (head and feet rotation), CW / CCW (clockwise and counterclockwise rotation)), standard error of the breathing model, correlation error, and model rigidity error.

[0067] The breathing amplitude, the error values ​​of the three translational degrees of freedom (i.e., SUP / INF (up and down translation), LEF / RGT (left and right translation), and ANT / POST (front and back translation)), and the error values ​​of the three rotational degrees of freedom (i.e., LEF / RGT (left and right rotation), UP / Down (head and feet rotation), and CW / CCW (clockwise and counterclockwise rotation)) are used to construct and correct the breathing model in real time.

[0068] The standard error is the error value of evaluating a model point during the modeling process. std The calculation formula is as follows:

[0069] (1)

[0070] e i is the difference between the model curve and the model points, and n is the number of model points.

[0071] The correlation error is used to evaluate the quality of the model during normal treatment. This value accurately indicates the degree to which the latest model point matches the current breathing model.

[0072] The threshold for both the standard error and the correlation error is 5 mm, and the clinical requirement is 2 mm, the smaller the better. If the threshold is exceeded, the respiratory model is unqualified.

[0073] The model rigidity error is the root mean square of the rigidity errors of each pair of tumor markers. The rigidity error of each pair of tumor markers is the difference between the corresponding distances in the CT volume and the real-time orthogonal image.

[0074] Taking three tumor markers as an example, the rigid error RBE of each pair of tumor markers is as follows:

[0075] F1-2ICT: distance between tumor markers 1 and 2 in the CT volume;

[0076] F1-3ICT: distance between tumor markers 1 and 3 in the CT volume;

[0077] F2-3.|CT: distance between tumor markers 2 and 3 in the CT volume;

[0078] F1-2ILive: distance between tumor markers 1 and 2 in real-time orthogonal images;

[0079] F1-3ILive: distance between tumor markers 1 and 3 in real-time orthogonal images;

[0080] F2-3ILive: distance between tumor markers 2 and 3 in real-time orthogonal images.

[0081] RBE1-2= F1-2ICT-F1-2Ilive;

[0082] RBE1-3= F1-3ICT-F1-3Ilive;

[0083] RBE2-3= F2-3|CT-F2-3|Live.

[0084] Rigidity error is used to determine whether the distance between tumor markers has changed. The range of rigidity error RBE threshold is 0.5-5mm, and the maximum threshold can be adjusted. The clinical requirement is to set the maximum threshold to 2.5mm. The rigidity error threshold setting is the maximum allowable rigidity error of each pair of tumor markers. If it exceeds 2.5mm, the clinic considers that the tumor marker has moved and cannot be used. If the value of any pair of tumor markers exceeds this threshold, an error will occur and the treatment will be interrupted.

[0085] S140: Compare the statistical parameters of all breathing patterns, obtain the optimal breathing model according to the optimal selection rule, and use the breathing pattern corresponding to the optimal breathing model as the optimal breathing pattern. Apply the optimal breathing pattern and the optimal breathing model to stereotactic synchronous breathing tracking therapy.

[0086] As an embodiment, the selection rule is: among the qualified breathing models, the optimal breathing model is screened in the priority order of standard error > correlation error. If there is only one breathing model with the smallest standard error, then this breathing model is used as the optimal breathing model. If there are two breathing models with equal and smallest standard errors, then the one with the smallest correlation error is selected as the optimal breathing model.

[0087] As an embodiment, for each breathing mode, building a breathing model includes:

[0088] P1: In each respiratory cycle, the first CT slice and contrast images and the first orthogonal images at multiple target phase points are acquired according to the real-time body position of the target object, and the first digitally reconstructed radiological image is generated according to the first CT slice and contrast images.

[0089] P2: Calculate the first error between the position of the tumor marker in the first orthogonal image and the position of the tumor marker in the first digitally reconstructed radiographic image at each target phase point in the same respiratory cycle.

[0090] The tumor marker is a body surface marker located on a flat part of the body surface projection area of ​​the treatment site of the target object or a gold marker implanted in the treatment site.

[0091] P3: Calculate the coverage rate of the breathing cycle according to the proportion of target phase points whose first error is less than or equal to the threshold in the same breathing cycle to all target phase points.

[0092] P4: Determine whether the coverage is greater than or equal to the preset value. If yes, execute P6. Otherwise, execute P5.

[0093] P5: For a target phase point whose first error is greater than a threshold, correct the shooting position of the orthogonal planar imaging system at the target phase point in the next respiratory cycle according to the first error at the target phase point, and then return to P1.

[0094] P6: Fit the first error curve of the entire breathing cycle according to the first errors of all target phase points in the last breathing cycle as the breathing model of the breathing mode. The first error curve is a straight line or a curve, that is, the breathing model is a linear model, a curve model or a hyperbolic model, etc.

[0095] Based on the above, if Figure 2 As shown, the present application also provides a method for synchronous breathing tracking of a CyberKnife, comprising:

[0096] S210: According to the treatment part of the target object and the confirmed CT image of the target object, the target object is fixed in a personalized position. Please refer to the above description of the preparation work and position fixation of the target object.

[0097] S220: Constructing an optimal breathing model according to the breathing curve of the optimal breathing method, wherein the optimal breathing method and the optimal breathing model are determined according to the modeling method of the above-mentioned breathing model.

[0098] S230: During the treatment process, predict the irradiation position of the CyberKnife radiation beam at the next moment according to the optimal breathing model.

[0099] S240: Sending the irradiation position to CyberKnife so that the irradiation position of the radiation beam moves synchronously with the treatment part caused by the breathing of the target object.

[0100] Preferably, during the treatment process, the breathing curve is updated according to a preset update cycle, and the breathing model is updated at the same time.

[0101] Although some specific embodiments of the present application have been described in detail by way of example, it should be understood by those skilled in the art that the above examples are only for illustration, not for limiting the scope of the present application. It should be understood by those skilled in the art that the above embodiments may be modified without departing from the scope and spirit of the present application. The scope of the present application is defined by the appended claims.

Claims

1. A method for modeling a respiratory model, characterized in that: include: Acquire the first breathing curve of the target object during free breathing, and construct a first breathing model based on free breathing; Acquire a second breathing curve corresponding to the target object breathing with the assistance of at least one assisted breathing device, and construct a second breathing model based on each assisted breathing; Obtain corresponding statistical parameters based on the breathing models of all breathing modes; Comparing the statistical parameters of all breathing modes, obtaining the optimal breathing model according to the optimal selection rule, and taking the breathing mode corresponding to the optimal breathing model as the optimal breathing mode; The optimization rule is as follows: First, find a qualified breathing model; Among qualified respiratory models, the optimal respiratory model is screened in the order of standard error > correlation error. If there is only one respiratory model with the smallest standard error, the respiratory model is selected as the optimal respiratory model. If there are two respiratory models with equal and smallest standard errors, the one with the smallest correlation error is selected as the optimal respiratory model. For each breathing mode, a breathing model is constructed, including: Repeat the following steps until the coverage is greater than the preset value: In each respiratory cycle, first CT slice and contrast images and first orthogonal images are acquired at a plurality of target phase points according to the real-time body position of the target object, and a first digitally reconstructed radiographic image is generated according to the first CT slice and contrast images; calculating a first error between a position of the tumor marker in the first orthogonal image and a position of the tumor marker in the first digitally reconstructed radiographic image at each target phase point in the same respiratory cycle; Calculate the coverage rate of the breathing cycle according to the proportion of target phase points whose first error is less than or equal to the threshold in the same breathing cycle to all target phase points; If the coverage rate is less than the preset value, for the target phase point whose first error is greater than the threshold, the shooting position of the orthogonal planar film system at the target phase point in the next respiratory cycle is corrected according to the first error at the target phase point; Fitting a first error curve of the entire breathing cycle according to the first errors of all target phase points in the last breathing cycle as a breathing model of the breathing mode; The phase point is determined by the breathing curve.

2. The method for modeling a respiratory model according to claim 1, characterized in that: The tumor marker is a body surface marker positioned on a flat portion of the body surface projection area of ​​the treatment site of the target subject or a gold marker implanted in the treatment site.

3. The method for modeling a respiratory model according to claim 1, characterized in that: The assisted breathing device includes a sound assisted breathing device and a visual assisted breathing device.

Images