JP7048038B2 - Robot moving body phantom system - Google Patents

Description

The present invention relates to a robotic moving body phantom system, and more particularly to a robotic moving body phantom system including a robot manipulator, a moving body tracking device and a phantom, and verifying a treatment plan for radiotherapy.

With the advent of a super-aging society, when treating malignant neoplasms (cancer), minimally invasive treatment has come to be selected, and the need for radiation therapy is extremely increasing. Radiation therapy has made remarkable progress due to advances in computer and other technologies, and is positioned as one of the three pillars of cancer treatment. This radiation therapy is used not only for the complete cure of the tumor, but also for the purpose of alleviating the pain caused by the tumor.

Currently, the mainstream of radiotherapy is three-dimensional radiotherapy, which uses CT (Computed Tomography) images to grasp the anatomical structure of the human body and the position of the tumor in three dimensions. However, radiation therapy for tumors with respiratory movement such as the chest and abdomen needs to be treated including the entire area where the tumor moves due to respiratory movement. Therefore, although radiation therapy increases the concentration of radiation on the tumor, there is a problem that extra radiation is administered to the surrounding normal tissues and side effects due to radiation increase.

Furthermore, with the help of a computer, the intensity of radiation can be adjusted by combining multiple beams formed by a multi-leaf diaphragm to reduce the dose to normal tissues, and the intensity of the radiation can be modulated as a radiation therapy technology that can intensively irradiate the tumor. Radiation therapy is known. Intensity-modulated radiotherapy, when the treatment is applied to a site with respiratory movement, will administer a completely different radiation dose than planned, which causes problems such as a decrease in tumor control rate and an increase in side effects on normal tissues. be.

In order to solve these problems, in recent years, radiation therapy has been researched and developed toward four-dimensionalization by adding a time element to a three-dimensional space, and is expected to be applied to clinical practice. In clinical practice, moving body tracking radiation therapy using a moving body tracking device ( ^SyncTraXTM , Shimadzu Corporation), which is one of the four-dimensional radiotherapy, is performed.

As an example of such a moving body tracking device, Patent Document 1 describes a metal placed near a tumor under bronchoscopy or CT or ultrasonic guidance by performing pattern recognition image processing on a two-way X-ray fluoroscopic image. A moving object tracking device capable of calculating three-dimensional position coordinates of a marker in real time and tracking a metal marker (tumor) moved by respiratory movement is disclosed. The X-ray fluoroscopy position of the moving object tracking device can be selected from three directions, and the metal marker can be tracked without being restricted by the irradiation angle of the medical linear accelerator (regardless of the gantry or couch angle). Is possible. The moving body tracking radiotherapy is a treatment in which radiation is emitted only when the three-dimensional position of the metal marker calculated in real time using the moving body tracking device reaches a certain position, and the moving body tracking device and the medical linear accelerator are used. Treatment can be performed by combining them.

In moving body follow-up radiotherapy, after seeing a doctor, a fixture dedicated to the patient is created and CT imaging for treatment planning is performed. Based on the anatomical information of the CT image, a treatment plan is performed by simulating the irradiation direction and irradiation amount on a computer. Once the treatment plan is complete, a motion tracking device is used to confirm the visibility of the metal markers placed in the patient and determine the fluoroscopy position to use in the treatment. At this time, the three-dimensional coordinates of the metal marker moved by breathing are recorded as a log. After that, in order to evaluate the validity of the treatment plan in simulation, the quality of the treatment plan is guaranteed.

As an example of radiation therapy based on a treatment plan, Patent Document 2 describes a phantom that has a radiation absorption rate equivalent to that of the human body and is simulated as a water-equivalent tissue, and is measured by a dosimeter inserted in the phantom. A quality assurance procedure is disclosed that verifies that there is no difference between the radiation dose given and the calculated value at the time of treatment planning. When this quality assurance ensures the accuracy of radiation therapy, treatment is started for the patient.

Japanese Patent No. 3053389 Japanese Patent No. 4115675

However, the quality assurance of the current moving body tracking radiotherapy is based on the coordinate data of one axis having the three-dimensional coordinates of the metal marker acquired by using the moving body tracking device as disclosed in Patent Document 1. A phantom as disclosed in Document 2 is driven by one axis. Therefore, there is no device using a phantom that reproduces the movement of a tumor that moves three-dimensionally by the respiration of the human body, and the quality of moving body tracking radiotherapy cannot be guaranteed with high accuracy.

The present invention has been made in view of the above-mentioned conventional problems, and an object of the present invention is to have a degree of freedom in which three-dimensional translational movement is possible in order to reproduce an actual tumor movement, and is utilized for quality assurance of radiotherapy. It is an object of the present invention to provide a moving body phantom system having a follow-up accuracy that can be achieved.

In order to solve the above problems, the robot moving body phantom system according to the present invention has a robot manipulator having three-dimensional translational motion degrees of freedom, a robot control device for controlling the robot manipulator, and a human body fixed to the tip of the robot manipulator. It is equipped with a phantom having the same radiation absorption rate and a moving body tracking device that measures changes in the marker position of a patient placed near the tumor in real time, and the robot control device has a three-dimensional motion trajectory of the marker position of the patient. It has a target trajectory generation unit that generates a target trajectory of the robot manipulator, and is characterized by controlling the robot manipulator so that the marker position in the phantom follows the target trajectory.

According to the robot moving body phantom system according to the present invention, the accuracy of quality assurance of radiotherapy can be improved. The effects described here are not necessarily limited, and may be any of the effects described in the present technology.

It is a schematic diagram which shows the coordinate system of the robot moving body phantom system which concerns on 1st Embodiment of this invention, (A) shows the treatment apparatus coordinate system of a medical linear accelerator, (B) shows the working coordinate system of a robot manipulator. show. It is a graph which shows the lung tumor locus of a patient measured by the moving body tracking apparatus which concerns on 1st Embodiment of this invention. It is a graph which shows the lung tumor locus of a patient measured by the moving body tracking apparatus which concerns on 1st Embodiment of this invention. It is a graph which shows the lung tumor locus of a patient measured by the moving body tracking apparatus which concerns on 1st Embodiment of this invention. It is a graph which shows the lung tumor locus of a patient measured by the moving body tracking apparatus which concerns on 1st Embodiment of this invention. It is a schematic diagram which shows the structure of the robot moving body phantom system which concerns on 1st Embodiment of this invention. It is a schematic diagram which shows the moving body tracking apparatus installed together with the medical linear accelerator which concerns on 1st Embodiment of this invention. It is a figure explaining the method of setting the target value between the sample points of the measurement result of the moving body tracking apparatus which concerns on 1st Embodiment of this invention. It is a graph which shows the operation of this algorithm in the case of linear interpolation by the control device which concerns on 2nd Embodiment of this invention. It is a schematic diagram which shows the introduction of the additional feedback loop using the external controller which concerns on 2nd Embodiment of this invention. It is a figure explaining an example of the calculation method when the angle calculation part determines the rotation angle of a tumor. This is an example of an image in which a conductor tracking device actually captures a plurality of embedded markers. This is an example of the tumor rotation angle obtained by the method according to the present invention. It is a schematic diagram which shows the installation example of the robot manipulator which concerns on one Example of this invention. It is a schematic diagram which shows the installation example of the robot manipulator which concerns on other Examples of this invention. It is a graph which shows the plot of the time change of the three-dimensional error by the control device which concerns on 1st Embodiment of this invention.

<1. First Embodiment>
First, the configuration of the robot moving body phantom system according to the first embodiment of the present invention will be described with reference to FIGS. 1 to 6. The embodiment of the present invention is not limited to the embodiment shown below, and any of the embodiments can be combined. Further, in the embodiment shown below, the "water equivalent phantom" is used as the moving body phantom, but the present invention is not limited to this, and a phantom other than the "water equivalent" may be used.

<2-1. Overview of moving phantom system>
In radiation therapy, devices that reproduce the movement of tumors (markers) that move under the influence of breathing and heartbeat are important for quality assurance of radiation therapy for moving tumors such as moving body tracking radiation therapy and moving body tracking irradiation. Is. As such a device, a moving object phantom system is known. Here, the moving phantom system is a phantom drive device that simulates changes in the tumor position in the patient's body in real time, a water-equivalent phantom having a radiation absorption rate equivalent to that of living tissue (human body), a dosimeter, and a radiation-sensitive film. It is a quality assurance system for radiotherapy, which consists of a combination of.

In the water equivalent phantom installed on the phantom drive device, a marker to be positioned by the moving object tracking device is embedded, and a dosimeter or a radiation sensitive film is installed as needed. The phantom drive device moves the water-equivalent phantom with a pre-recorded tumor movement trajectory at rest as a target value. While monitoring this operation in the same way as during treatment, irradiation is performed according to the treatment plan, and the dose is measured with a dosimeter to confirm whether the irradiation dose specified in the treatment plan can be administered.

At present, it is known that the moving object phantom system is capable of linear reciprocating motion with one degree of freedom and transiently causes a large tracking error of about several mm. The movement of the tumor in the body is a movement in a three-dimensional space, and it is desired to have a moving body phantom system that can reproduce it and that the three-dimensional tracking error of the entire time can be suppressed to about 1 mm for high-precision radiotherapy.

Regarding the configuration of a moving body phantom system using a robot manipulator and a similar positioning device, a stepping motor and a 6-axis robot manipulator are used to simulate the respiratory deformation of the thorax and the movement of the tumor in the thorax. It is known that a moving body phantom system for radiation therapy has been constructed. However, in this moving object phantom system, in addition to the particle beam detector fixed to the tip of the robot manipulator, the three-dimensional target trajectory used to evaluate the tracking accuracy of the tumor trajectory is a sine wave. Only the error evaluation for the target value, which is considered to be relatively easy to improve the tracking accuracy, is performed. The peculiarity and difficulty of setting a problem when using a robot manipulator for a moving object phantom system is that although the target trajectory is periodic to the large marker, the amplitude and phase, and in some cases the waveform shape itself, also fluctuates. This moving object phantom system does not touch on that point.

On the other hand, a phantom drive device using an XYZ table configured by combining a linear acting actuator has been developed. According to this phantom drive device, it was reported that the result of evaluating the tumor trajectory tracking performance by the μ + 2σ value of the three-dimensional error signal was within 0.8 mm for a total of 20 cases of lung cancer, liver cancer and pancreatic cancer. Has been done. The performance evaluation of the phantom drive device is performed using the accuracy of following the tumor motion trajectory of the cancer patient measured in advance as an index. As will be described later, there is a large difference in tumor movement between cases, but the present inventors also decided to evaluate the performance using the above values as a guide.

Therefore, the present inventors have constructed a robot moving body phantom system using an industrial small 6-axis robot manipulator system (MZ07-01, Nachi-Fujikoshi Co., Ltd.). Table 1 outlines the specifications of the selected robot manipulator as an example.

The motion degrees of freedom required for a robot moving body phantom are only three-dimensional translational motion degrees of freedom (translational three degrees of freedom), but a total of water equivalent phantoms and jigs for fixing to the tip of the robot manipulator are required. Since it has a mass of 7 kg, the model was selected in consideration of the payload of the tip guaranteed by the robot manipulator. When the only motion required for the robot manipulator is translational motion, it is undeniable that the rotary joint type robot manipulator is mechanically and controlally disadvantageous to the positioning device based on the linear actuator. However, this specification proposes a method for correcting the target trajectory given to the robot manipulator, and even if a rotary joint type robot manipulator whose load applied to the actuator changes greatly depending on the posture is used, the follow-up performance to the patient's tumor trajectory is clinical. We are proposing a robot moving body phantom system that can meet the required accuracy in the field.

<2-2. Configuration example of robot moving phantom system>
When using a robot manipulator in a normal production site, its operation can be programmed using the attached teaching device or dedicated software, whether it is point-to-point position control or following a continuous trajectory. As long as it is used for the same work in the same place, it is not necessary to change the trajectory and operation once programmed. However, when the robot manipulator is used as a phantom drive device for a moving phantom, the tumor trajectory that the phantom should follow differs from patient to patient, and high-precision tracking is required for the entire time during operation.

FIG. 1 is a schematic diagram showing a coordinate system of a robot moving body phantom system according to the first embodiment of the present invention. FIG. 1 (A) shows how to take the coordinate system of the medical linear accelerator 10 used for measuring the tumor position in quality assurance and treatment, and FIG. 1 (B) shows the working coordinates used in the work plan of the robot manipulator. Shows the system settings. In FIG. 1A, a radiation therapy / measurement system is formed by a medical linear accelerator 10 and a bed 11, and an origin ic (iso-center) of a treatment device coordinate system is provided. In FIG. 1B, the robot manipulator 12 has a water equivalent phantom 13 attached to the tip thereof, and the rear end thereof is installed at the origin O of the working coordinate system.

When the robot moving body phantom system is used in combination with the radiotherapy / measurement system of FIG. 1 (A), the robot manipulator 12 is placed on the bed 11 visible in front of FIG. 1 (A) in the X-axis direction of the robot manipulator 12. Is fixed so as to coincide with the y-axis direction of the medical linear accelerator 10 which is a treatment device. Therefore, coordinate conversion is required when passing the coordinate data of the tumor trajectory measured by the coordinate system of the treatment device to the robot moving body phantom system. The operation results of the three-dimensional locus data and the robot manipulator 12 shown in the present specification are all expressed in the working coordinate system of the robot manipulator 12 of FIG. 1 (B), and the origin O is the coordinates of the treatment device. A plot is created by moving in parallel so that it overlaps with the origin ic of the system.

2 to 5 are graphs showing lung tumor trajectories in the X-axis, Y-axis, and Z-axis directions of four lung cancer patients A to D measured by the moving body tracking device according to the first embodiment of the present invention. be. The waveforms shown in FIGS. 2 to 5 show the lung tumor trajectories of patients A to D taken up as target trajectories in the accuracy verification of the robot moving phantom system of the present embodiment. In the case of lung tumor, the behavior is such that the vibration of a high frequency component caused by the beat of the adjacent heart is superimposed on the slow vibration with a period of about 4 s called respiratory movement, but the amplitude and the degree of overlap are different. It turns out that there are very large individual differences. It is also important that it reflects the fluctuations of the rhythmic movements of the living body and is unlikely to be a periodic signal in the strict sense.

<2-3. Robot control system configuration example>
FIG. 6 is a schematic diagram showing a configuration of a robot moving body phantom system according to the first embodiment of the present invention. The robot moving body phantom system 60 includes a robot manipulator 61, a robot control device 62, a teaching pendant 63, and an external controller 64. Further, a water equivalent phantom 65 is attached to the tip of the arm of the robot manipulator 61.

The robot control device 62 has a target trajectory generation unit 66 that converts and outputs a three-dimensional motion trajectory of a patient tumor measured in advance using a moving object tracking device into a target trajectory for the robot manipulator 61. Does the external controller 64 hold information on the patient's marker position in order to match the time interval of tumor position tracking of the moving body tracking device with the control cycle of the tracking motion of the robot control device 62 with the target trajectory of the robot manipulator 61? Or, it has a "sampling cycle conversion" function that interpolates the measurement result of the time interval by the moving object tracking device, retains the information of the patient's marker position, and upsamples. As described above, the robot moving body phantom system 60 is configured to follow the marker position in the water equivalent phantom 65 fixed to the tip of the robot manipulator 61 to the target trajectory generated by the robot target trajectory generating unit 66 with high accuracy. Has been done.

In the present embodiment, the robot moving body phantom system 60 that satisfies the accuracy required in the clinical field is constructed by using the follow-up mode of the external controller 64 included in the robot control device 62 of the selected robot manipulator 61. The "sampling cycle conversion" function is not limited to the case where the external controller 64 has, and may be provided by another configuration such as the robot control device 62.

The robot control device 62 includes a driver for a motor that drives each axis of the robot manipulator 61 and a processing circuit for an encoder signal. The robot control device 62 includes a lower system that performs servo compensation control by a two-degree-of-freedom controller at each axis level, and a higher system that realizes a response to an operation using the teaching pendant 63 by an operator and a function of teaching reproduction. , Has a hierarchical structure. The lower system and the upper system operate while constantly transmitting and receiving data (information signals) necessary for control and operation by communication means.

In the follow-up mode of the external controller 64, the host system of the robot control device 62 performs TCP (Transmission Control Protocol) data communication with the external controller 64 at a fixed cycle of 5 ms. The external controller 64 can command the robot control device 62 the target trajectory of the end effector in real time.

On the other hand, the robot control device 62 is configured to perform control based on the target trajectory information given by the external controller 64, and to feed back information such as the current tip position and posture as a result to the external controller 64. .. Table 2 below shows the information exchanged between these robot control devices 62 and the external controller 64.

As shown in Table 1, the selected robot manipulator 61 achieves a position repetition accuracy (JIS B 8342 compliant) of ± 0.02 mm under the control of the robot control device 62. However, the tracking performance required when used as the robot moving object phantom system 60 is not repetitive positioning to a fixed point, but high-precision tracking in real time to a target trajectory that cannot be explicitly described as a function of time.

When the robot moving body phantom system 60 of the present embodiment is used in a clinical setting as a quality assurance tool, it is considered difficult to entrust the system user to adjust the control operation of the robot manipulator 61 in order to improve the trajectory tracking accuracy. .. In addition, the target trajectory is poorly regular and is a tumor trajectory that varies greatly from patient to patient.

Therefore, in the present embodiment, a method of online correction of the tip position target trajectory given to the robot manipulator 61 with the patient's tumor position trajectory measured by the moving body tracking device as a starting point is proposed, and the parameters are readjusted for each patient's tumor trajectory. It was decided to construct a robot moving body phantom system 60 that does not require the above.

<3. Reduction of tracking error by correcting the target trajectory>
Next, the reduction of the tracking error by correcting the target trajectory will be described with reference to FIGS. 7 to 10.

<3-1. Measurement of tumor migration trajectory>
Here, the acquisition of the patient's tumor trajectory data, which is the target trajectory for the robot moving body phantom system 60, will be described. As described above, the present inventors utilize the moving body tracking device of the marker to perform this measurement in the process of clinical practice.

FIG. 7 is a schematic view showing a moving body tracking device 80 installed together with the medical linear accelerator 81 according to the first embodiment of the present invention. As shown in FIG. 7A, the moving object tracking device 80 of the present embodiment includes color receiving devices 84, 85 corresponding to the two X-ray sources 82, 83. A sleeper 86 on which the patient to be treated is placed is installed at a position facing the medical linear accelerator 81. For the markers captured in the two stereo images, the coordinate values in the images are specified using the template matching technique, and the coordinate values of the markers in the treatment device coordinate system are output.

Further, as shown in FIG. 7B, three different spatial combinations are possible between the X-ray sources 82 and 83 and the color image receiving devices 84 and 85, and the tumor site and the treatment can be used. A combination of the sleeping bed 86 to be moved and the gantry that does not interfere with the gantry is selected and used.

As shown in FIG. 7A, the pattern 1 is arranged at a position where the X-ray sources 82 and 83 and the corresponding color image receiving devices 84 and 85 intersect with each other across the medical linear accelerator 81. In pattern 2, two X-ray sources 82b and 83 are arranged on the left side facing the paper surface of FIG. 7 (B), and two color image receiving devices 84b and 85 corresponding to the right side facing the paper surface of FIG. 7 (B). Is arranged at the intersection of each X-ray. In pattern 3, two X-ray sources 82 and 83b are arranged on the right side facing the paper surface of FIG. 7 (B), and two color image receiving devices 84 and 85b corresponding to the left side facing the paper surface of FIG. 7 (B). Is arranged at the intersection of each X-ray.

However, in any case, there is no change in the principle of performing stereo coordinate measurement by template matching. The moving object tracking device 80 guarantees that the localization error of the marker position is within 0.8 mm.

<3-2. Generation of tumor migration trajectory>
The moving body tracking device 80 of the present embodiment has an ability to measure and output the tracking result of the marker position placed in the vicinity of the tumor in a cycle of about 33 ms in order to track the position of the mobile tumor. On the other hand, in the configuration of the robot moving body phantom system 60, the external controller 64 is configured to feed the target value to the robot control device 62 at intervals of 5 ms. Therefore, in order to drive the robot manipulator 61, it is necessary to generate a target value time series for the robot manipulator with a shorter time interval from the time series of the measurement result log file of the moving object tracking device 80. In this target trajectory generation, four methods are shown below using FIG. 8 as a method of determining a target value between sample points of the measurement result of the moving object tracking device 80.

FIG. 8 is a diagram illustrating a method of determining a target value between sample points of the measurement result of the moving body tracking device according to the first embodiment of the present invention. FIG. 8 [S] is a step common to any of the ^methods of determining the target value of the present embodiment. be. The measurement of the moving object tracking device according to the present embodiment is not limited to the case of measuring with the ^SyncTraXTM .

The method of FIG. 8 [1] is the most basic method and is a method of maintaining the current value until the target coordinate value is updated. In FIG. 8 [1], after the step of FIG. 8 [S] a), b) the latest marker position information is held until the next sampling time, and c) the locus of the marker position information held in the previous step is sampled at a sampling period of 5 ms. Upsample with.

The method of FIG. 8 [2] is a method of pre-reading the marker position at the next sampling time. In FIG. 8 [2], after the step of FIG. 8 [S] a), b) the marker position at the next sampling time is pre-read and held, and c) the locus of the marker position held in the previous step is held at a sampling period of 5 ms. Upsample. The method of FIG. 8 [2] is the same as the method of FIG. 8 [1] in that the constant value is maintained within one hour interval, but the target value is obtained in advance of the value one sample ahead of the moving object tracking device 80 log. By setting, the purpose is to reduce the influence of the transient response operation of the robot control device 62 on the error.

The method of FIG. 8 [3] is simply a linear interpolation method of interpolating between two adjacent coordinates with a line segment. In FIG. 8 [3], after the step of FIG. 8 [S] a), b) the coordinate data is interpolated by a line segment, and c) the locus of the coordinate data interpolated in the previous step is upsampled with a sampling period of 5 ms.

The method of FIG. 8 [4] is a method of interpolating the coordinate data with a cubic spline function. In FIG. 8 [4], after the step of FIG. 8 [S] a), b) the coordinate data is interpolated by a cubic spline function, and c) the locus of the coordinate data interpolated in the previous step is upsampled with a sampling period of 5 ms. .. The method of FIG. 8 [4] is an interpolation method that guarantees the continuity of velocity and acceleration at the connection point of time, based on the idea that the velocity and acceleration indicating the movement of a living tissue such as a tumor are continuous. ..

The finally generated trajectory is k (k = 0,1, ..., N) on the moving object tracking device 80 log file regardless of which of the four methods shown in FIGS. 8 [1] to [4] is used. -1) Using the time variable t ∈ [Tst [k], Tst [k + 1]) belonging to the third time interval, it is expressed by the following (Equation 1).

As described above, since it can be represented by a cubic polynomial with respect to t, the target trajectory calculation data corresponding to the one-hour interval of the moving object tracking device 80 log file has 12 coefficients of these three polynomials.

Basically, the length of one sample section is Tst [k + 1] -Tst [k] = 0.033s. However, in addition to some fluctuation, data is lost in the section when the marker recognition fails in the image processing process of the moving object tracking device 80, although it is rare. If data is missing, calculate using the marker position and time obtained next. The process of generating the coefficient set of (Equation 1) from the moving object tracking device 80 log file is executed by creating an m script of MATLAB (Mathworks). As described above, the target trajectory can be expressed as a function of time. Using this, the correction of the target trajectory in real time on the external controller 64 described in detail below and the processing for improving the tracking error can be performed. It becomes feasible.

<4. 2nd Embodiment>
Next, the robot moving body phantom system according to the second embodiment of the present invention will be described.

<4-1. Communication delay reduction correction and online reinterpolation of the target trajectory>
The polynomial coefficient generation from the moving body tracking device 80 log described in the first embodiment is a process necessary for instructing the robot moving body phantom system 60 of the tumor trajectory data of the patient to be treated, and is a process required for the moving body tracking device 80. It does not change the data contained in the log of. However, it is difficult to satisfy the required accuracy in clinical practice simply by simply feeding the target trajectory from the external controller 64 to the robot control device 62 by using the method of FIG. 8 [1]. Therefore, in the second embodiment, on the premise of the configuration of the robot moving body phantom system 60 shown in FIG. 6, the target trajectory correction and the target trajectory information using the tracking error information for constructing a system with higher tracking accuracy are used. The online reinterpolation method will be described.

<4-2. Correction of target trajectory for the purpose of reducing the effect of communication delay>
As shown in FIG. 6, in the robot moving body phantom system 60, network communication is used for exchanging data between each subsystem, and a delay may occur due to the communication between each system. Since the delay of data in the transfer of the amount related to the control is directly linked to the deterioration of the tracking accuracy, the robot moving body phantom system 60, which requires high tracking accuracy in real time, needs to reduce the influence of this delay.

を以下の(式２)により算出して目標値を拡張する。

Therefore, with the intention of this compensation, the correction amount is used for xk (t), yk (t), and zk (t) in (Equation 1) using Tr = 0.005s.

Is calculated by the following (Equation 2) to expand the target value.

However, the constants αx, αy, αz, βx, βy, and βz included in (Equation 2) were selected so as to have positive values. In addition, in the step of generating the target value series of the robot control device 62 at short intervals from the moving object tracking device 80, the constant value is maintained within one hour section of the moving object tracking device 80 log, as shown in FIGS. 8 [1] and [2]. As for the method, since there is no change in velocity and acceleration within the section, this correction is not applied in the experimental evaluation.

When linear interpolation is performed as shown in FIG. 8 [3], the acceleration term takes a pulse-like non-zero value in one sample after the connection point of the sampling period conversion, but becomes 0 in the other steps, and the robot Not only is it a cause of inducing vibration of the manipulator 61, but the effect of correction is not sufficiently exhibited. For the speed control system of the robot manipulator 61, assuming that the transmission characteristic from the speed target value to the speed target can be approximated as a first-order lag, the acceleration changes like the impulse response of the first-order lag system.

If the transfer function can be expressed as 1 / (Ts + 1) (T> 0), if the acceleration signal α (t) is observed at regular time intervals Tr> 0, α (t) will be as follows (Equation 3-1). ) Decays.
α ((i + 1) Tr) = γ ・ α (i ・ Tr) (γ = e ^{-Tr / T} ) ・ ・ ・ (Equation 3-1)

Therefore, in the case of linear interpolation, the time ti = i · Tr (i) in the interval t ∈ [Tst [k], Tst [k + 1]), where αk is the acceleration value of the non-zero x-axis calculated at the beginning of the interval. = 1,2, ...) The value of the acceleration at the time is attenuated according to the following (Equation 3-2).

The y-axis and z-axis can be processed in the same manner as the x-axis processing described above. The value of γ is adjusted by trial and error during repeated experiments, and γ = 0.9 is used, but other values may be used.

<4-3. Feedback target value correction by dynamic reinterpolation>
Next, the feedback target value correction by dynamic reinterpolation using the tracking error information in the robot control device 62 will be described with reference to FIGS. 9 and 10.

The correction target value for each axis defined in the above (Equation 2) is determined only from the tumor motion trajectory observed by the moving body tracking device 80, and the tracking error that occurs when the robot manipulator 61 is actually moved. It is not a correction that takes into account. Therefore, using the tip position information of the robot manipulator 61 fed back from the robot control device 62 to the external controller 64, additional feedback correction using the tracking error information is performed on the target value given to the robot control device 62.

As described above, in the external controller 64 follow-up mode of the robot control device 62, the current value of the tip position / posture of the robot manipulator 61 is fed back to the external controller. Using the value of the tracking error, which is the difference between the two, the interpolation work for generating the target value to be sent to the robot control device 62 at a cycle of 5 ms is redone, and the interpolation trajectory is regenerated. Specifically, the above-mentioned tracking error is added to the coordinate value of the starting point of interpolation that has already become past data, and the target coordinate value to be sent to the robot control device 62 in the future is increased or decreased according to the sign of the error. I do.

Here, the calculation algorithm of dynamic reinterpolation is shown below. Since the same processing is performed independently for all the X-axis, Y-axis, and Z-axis, the calculation procedure for the X-axis will be described here as a representative. Assuming that the time interval for performing reinterpolation is the kth time interval t ∈ [Tst [k], Tst [k + 1]) of the moving object tracking device log, as described above, before the correction in this interval. The target value is determined by the following cubic polynomial of (Equation 4).
xk (t) ＝ ax [k] ＋ bx [k] (t-Tst [k]) ＋ cx [k] (t-Tst [k]) ² ＋ dx [k] (t-Tst [k]) ³・ ・ ・(Equation 4)

The sample number of the external controller 64 within this time interval is represented by i (i = 1, 2, ...), And from the current position of the X axis fed back from the robot control device 62 to the external controller 64, for example. It is assumed that an error of ex [i] is observed in step i. At this time, the target trajectory generation formula that replaces (Equation 4), which is valid only within this section, is represented by the following (Equation 5).
x ⁽ⁱ⁾ k (t) ＝ ax [k, i] ＋ bx [k, i] (t-Tst [k]) ＋ cx [k, i] (t-Tst [k]) ²
＋ dx [k, i] (t-Tst [k]) ³・ ・ ・ (Equation 5)

Then, the coefficients ax [k, i] to dx [k, i] are determined so that (Equation 5) satisfies all of the following constraints.

The above (Equation 7) requests that the value of the target value at the end time of this section matches the initial value of the target value of the next section. The above (Equation 8) and (Equation 9) are conditional expressions for ensuring the continuity of velocity and acceleration at the connection point, which is a feature of cubic spline interpolation, and the method shown here is used in this section. Even if the target value generation polynomial of is changed to the above (Equation 5), the continuity of position, velocity and acceleration will be maintained.

として、(式５)の係数は、(式６)と以下の(式１０) により定められるので、これらの式をオンラインで計算することで再補間実施後の補間多項式を求めることができる。

here,

As the coefficient of (Equation 5) is determined by (Equation 6) and the following (Equation 10), the interpolation polynomial after reinterpolation can be obtained by calculating these equations online.

FIG. 9 is a graph showing the operation of this algorithm in the case of linear interpolation by the control device according to the second embodiment of the present invention. 9 (A) shows the next target value sent to the robot controller, FIG. 9 (B) shows the current position of the tip calculated by the encoder information of the robot, and FIG. 9 (C) shows the uncorrected state. Represents the target trajectory of and the next target value sent to the robot controller. Subsequently, FIG. 9 (D) shows the current position of the new tip calculated by the encoder of the robot, and FIG. 9 (E) shows the target trajectory before the correction and the target trajectory after the correction.

Even when the target value trajectory is determined by the cubic spline function, the concept of reinterpolation calculation is exactly the same as the algorithm shown in FIG. In addition, in the case where the target value is set to a constant value within the section shown in FIGS. 8 [1] and [2], (Equation 6) and (Equation 7) cannot be satisfied at the same time. Even in this case, it is possible to increase or decrease the target value used in the section using only (Equation 6), but if the robot manipulator 61 is operated in a test operation with such a setting, it may induce violent vibration. be. It is considered that this is because the change of the target value is stepped and a large deviation temporarily occurs, so that the value of ex [i] added by (Equation 6) also becomes large. As a result of the trial operation, it was judged that it is dangerous to carry out the experiment with this setting, so in the following evaluation experiment, when using a certain target value, it is not applicable as well as the correction using the previous speed and acceleration. There is.

は、以下の(式１１)～(式１３)で与えられる。

When the target value obtained as a result of applying the dynamic reinterference algorithm of the target value trajectory and the target value correction represented by the above (Equation 2) using the velocity and acceleration are used at the same time, the robot control device from the external controller 64 Orbit command value sent to 62

Is given by the following (Equation 11) to (Equation 13).

In the above (Equation 11) to (Equation 13), an additional feedback loop is introduced outside the feedback control system completed between the robot manipulator 61 and the robot control device 62 as shown in FIG. 10 below. Corresponds to.

FIG. 10 is a schematic diagram showing the introduction of an additional feedback loop using the external controller according to the second embodiment of the present invention. A feedback control system is formed between the robot manipulator 61 and the robot control device 62 of the present embodiment. Further, the external controller 64 of the present embodiment has a target trajectory generation unit 66, a target trajectory correction unit 111, and a target trajectory regeneration unit 112. The target trajectory generation unit 66 can include a position acquisition unit 201 that acquires a three-dimensional motion locus of the patient's marker position, and an angle calculation unit 202 that calculates the rotation angle of the marker that has acquired the three-dimensional motion trajectory. .. If it is also desired to grasp the rotation angle of the tumor, it is necessary to continue to acquire the three-dimensional motion locus of at least three marker positions of the patient by the position acquisition unit 201. Then, the angle calculation unit 202 performs a process of calculating the rotation angle of the tumor based on the acquired marker position to generate a target trajectory of the robot manipulator 61.

を用いて

と関係づけられるので、この３つの式から回転行列Ｒを数値的に一意に定めることができ、定まった回転行列からロボットの手先に固定されているファントムの回転角度の目標軌道を生成することができる。FIG. 11 is a diagram illustrating an example of a calculation method when the angle calculation unit 202 determines the rotation angle of the tumor. FIG. 11A shows a state when the moving object tracking device 80 detects the positions of the _three markers m1, _m2 , and m3 at time k- ₁ . FIG. 11B shows a state when the same three marker positions are detected at the next time k. FIG. 12 is an example of an image in which the conductor tracking device 80 actually captures a plurality of embedded markers. It is assumed that the position change of the markers m ₁ , m ₂ , and m ₃ is sufficiently represented only by the three-dimensional position change of the tumor center of gravity and the rotation in the three-dimensional space. The angle of rotation of the tumor between these two images can be defined using angles such as roll pitch yaw or Euler angles. FIG. 13 is an example of the tumor rotation angle obtained by the method according to the present invention. Between each vector v ₁ , v ₂ , n _{k-1 at} time k-1 and the corresponding vector v'1, _v'2 , n _k at time k, the rotation angle in the above three-dimensional space is used. A simultaneous transformation matrix defined using the rotation matrix R defined in the above and the vector δ representing the three-dimensional position change of the tumor center of gravity position determined by using the position acquisition unit 201.

Using

Since it is related to, the rotation matrix R can be numerically and uniquely determined from these three equations, and the target trajectory of the rotation angle of the phantom fixed to the hand of the robot can be generated from the determined rotation matrix. can.

For actual tumor movements, especially lung tumors, the lung itself repeatedly expands and contracts as the patient breathes. To express this transformation more strictly, the above rotation transformation matrix is expressed using the column vectors r1, r2, r3. R = [r1, r2, r3]
Introduced scalar real numbers a, b, and c that represent expansion and contraction in the axis direction, and R = [ar1, br2, cr3].
In addition to the matrix R, a, b, and c may be defined. The unknown increases due to the introduction of a, b, and c. In that case, the change between the time k-1 and k of the vector from m2 to m3 may be used.

The target trajectory correction unit 111 has a function of correcting the robot target trajectory using the velocity or acceleration of the motion trajectory of the robot manipulator 61 defined by interpolating the trajectory output as a result of the tumor position tracking of the moving body tracking device 80. Have. The target trajectory correction unit 111 may include a position correction unit 203 that corrects the three-dimensional coordinate position of the robot manipulator 61, and an angle correction unit 204 that corrects the rotation angle of the robot manipulator 61 that corrects the three-dimensional coordinate position. can. When correcting the robot target trajectory in consideration of the rotation angle, the position correction unit 203 corrects the three-dimensional coordinate position of the robot manipulator 61, and the angle correction unit 204 corrects the three-dimensional coordinate position of the robot manipulator 61 based on the corrected three-dimensional coordinate position. By performing the process of calculating the rotation angle, the target trajectory of the robot manipulator 61 is corrected.

The target orbit regeneration unit 112 has a function of calculating the follow-up error of the target orbit at each control sampling time from the control results for the target orbit given by the target orbit generation unit 66 and transmitting the information to the target orbit correction unit 111. Have. The target trajectory correction unit 111 has a function of adding the received tracking error of the target trajectory to the target value information of the start point of the interpolation section corresponding to the time, and re-interpolating in real time during control to regenerate the target trajectory. Have.

In the external controller 64 of the present embodiment, the information of the target trajectory of the working coordinate system generated by interpolation by the target trajectory generation unit 66 is transmitted to the target trajectory correction unit 111 and the target trajectory regeneration unit 112. After that, the target trajectory correction unit 111 makes a delay correction using the velocity and the acceleration, and the target trajectory regeneration unit 112 makes a correction by feedback re-storage. Then, the information of the target value after each correction is transmitted to the robot manipulator 61.

<5. Example>
Next, an example of quality assurance of radiotherapy using the robot moving body phantom system 60 of the above embodiment will be described with reference to FIGS. 14 to 16.

FIG. 14 is a schematic diagram showing an installation example of a robot manipulator according to an embodiment of the present invention. The robot manipulator 61 of FIG. 14 is fixed to a fixed base 121, and a water equivalent phantom 65 is attached to the tip of an arm.

FIG. 15 is a schematic diagram showing an installation example of a robot manipulator according to another embodiment of the present invention. The robot manipulator 61 of FIG. 15 is fixed to a sleeper 86 as shown in FIG. 8A with a fixture 131, and a water equivalent phantom 65 is attached to the tip of an arm.

<5-1. Follow-up accuracy evaluation experiment to tumor trajectory>
In this embodiment, in order to evaluate the effect of the proposed target trajectory correction method, the lung tumor trajectory 4 shown in FIGS. 2 to 5 is shown in a state where the water equivalent phantom 65 is actually fixed to the tip of the robot manipulator 61. The robot moving body phantom system 60 according to the above embodiment developed with the case as the target trajectory was driven, and the tracking accuracy thereof was evaluated.

In the method proposed in the present specification, a sampling cycle for generating a target value to be supplied to the robot control device 62 in a 5 ms cycle from the measurement result of the tumor trajectory using the moving body tracking device 80 obtained at a time interval of about 33 ms. There are four choices for conversion. There are two choices for the presence or absence of correction intended for delay compensation represented by (Equation 2). There are two choices for whether or not to apply dynamic reinterpolation for generating the target trajectory using the tracking error that can be acquired by the robot control device 62, which was described using (Equation 4) to (Equation 10). Therefore, there are 16 combinations for all.

However, as described above, when a constant target value is used within one sampling cycle of the moving object tracking device 80, the velocity / acceleration of the orbit becomes 0 within one sampling cycle, so that the delay compensation term becomes 0. In addition, since (Equation 6) and (Equation 7) cannot be satisfied at the same time, feedback reinterpolation cannot be applied. Therefore, in this evaluation, an experiment was conducted targeting a total of 10 different combinations excluding 6 combinations that the correction does not work or cannot be realized from all 16 combinations, and the robot moving body phantom system 60 using the robot manipulator 61 was used. The follow-up accuracy with respect to the target trajectory before correction was evaluated. The tracking error is determined as time-series data, and the RMS value of the error and the μ + 2σ value, which is an index related to the maximum error that can occur, are calculated and summarized.

The experiment was performed using two robot manipulators 61 in two different environments. The first environment is a laboratory having a concrete floor, and as shown in FIG. 14, the robot manipulator 61 is installed and fixed according to the manufacturer's specifications. The information on the tip position of the robot manipulator 61 obtained in this environment is only the tip position calculated by the robot control device 62 from the encoder information and fed back to the external controller 64.

The second environment is the clinical environment in which the patient is actually treated. Here, in addition to the tip position information by the robot control device 62, the measured value using the moving object tracking device 80 can also be obtained. Details will be described later, but in a clinical environment, as shown in FIG. 15, the robot manipulator 61 is fixed to the bed 86 of the medical linear accelerator 81 with the fixture 131, and the vibration-absorbing urethane foam is placed between the bed and the base of the bed. I installed it and experimented.

<5-2. Experimental results in the laboratory>
Table 3 shows the calculation of the root mean square error value and the value of μ + 2σ with respect to the target tumor trajectory from the experimental results conducted in the laboratory.

The error in this case is calculated as the difference between the uncorrected value of the target value sent from the external controller 64 to the robot control device 62 in a certain sample cycle and the marker position returned from the robot control device 62 in the next sample cycle. ing. As described above, network communication is used between the external controller 64 and the robot control device 62, and for data transfer in the robot control device 62, and there is a delay associated therewith, and the delay is in the buffer of the network. Considering that it changes randomly depending on the state, the error calculated here is not an accurate error at a certain sample time point. However, as shown in the evaluation results in the clinical environment below, if the statistical error index value calculated using the actual value of the robot control device 62 is large, the statistical error index value evaluated using the moving object tracking device 80 is also large. Become.

From the numerical values in Table 3, it can be seen that the correction of the target trajectory for the purpose of reducing the communication delay proposed above works extremely effectively. It can be seen that in all four cases, linear interpolation combined with delay compensation and feedback reinterpolation is not used gives the smallest μ + 2σ value. However, when comparing the results of linear interpolation, the difference in μ + 2σ values between the case where feedback reinterpolation is performed with delay compensation and the case where feedback reinterpolation is performed is the static error of this robot manipulator 61 shown in Table 1 except for case D. It can be seen that the accuracy of repeated positioning is within 0.02 mm, which is indistinguishable from the error that occurs accidentally.

On the other hand, when the third-order spline interpolation is used for the sampling period conversion, the μ + 2σ value is worse than the linear interpolation by 0.1 mm or more. It can be seen that the use acts to increase the error compared to linear interpolation.

The feedback re-interpolation method exhibits a large error reduction effect as compared with the case where the robot manipulator 61 is driven by the sampling period conversion alone, but the effect is small as compared with the error reduction amount when the delay compensation is used. It was also confirmed that when delay compensation and feedback reinterpolation are used together, the result is slightly deteriorated in this environment. However, the difference is within the error of repeated positioning to the fixed point guaranteed by the robot manipulator 61 used.

<5-3. Evaluation experiment in clinical environment>
Next, an evaluation experiment was conducted using the clinical equipment of the hospital and also using the moving body tracking device 80. The content of the experiment is exactly the same as the experiment in the laboratory, and the robot moving body phantom system 60 is used for the tumor trajectories of the four cases shown in FIGS. 2 to 5 using exactly the same parameter values as in the experiment in the laboratory. Driven. In the clinical environment, as the operation result of the robot manipulator 61, two measurement results of the position information output by the robot control device 62 and the measured value obtained by the moving object tracking device 80 can be obtained. The definition of the error when the position information output by the robot control device 62 is used is the same as described above.

Table 4 below shows the results of processing the error from the uncorrected tumor position target trajectory in the above case. In addition, Table 5 below shows the results of statistical processing of the error between the marker position measurement result of the moving body tracking device 80 and the pre-correction tumor target position.

Further, as an example of error evaluation, FIG. 16 shows a graph in which the tumor locus in FIG. 5 is set as a target value and the time change of the three-dimensional error is plotted for the three target value generation schemes in Table 5.

First, as in the laboratory environment, the combination that gives the minimum μ + 2σ value in Table 4, which is the result of error analysis using the tip position obtained by the robot control device 62, is the same as in Table 3, and the delay compensation term is added based on linear interpolation. It is the one that was put in. The minimum value is also almost the same as that in Table 3. Feedback reinterpolation has a certain degree of error reduction effect when used alone, but when combined with delay compensation, it has the same tendency to worsen the results, although it is not large.

On the other hand, in Table 5, which is an error using the measurement result using the moving object tracking device 80, different results are obtained. First, the best value of μ + 2σ is about 0.2 mm to 0.25 mm in cases A, B, and D with large tumor movement, and about 0.06 mm in case C with small movement compared to Tables 3 and 4. It's getting bigger. In addition, 3 out of 4 cases using cubic spline interpolation gave the minimum μ + 2σ value. The point that the result becomes worse when the delay compensation and the dynamic reinterpolation are used together is the same as the analysis using the position output of the robot control device 62.

The marker position output by the robot control device 62 is relative position information from the work coordinate origin O on the fixed base 121 of the robot manipulator 61, whereas the measurement result by the moving body tracking device 80 is a linear accelerator for medical use. It is an absolute coordinate value in a coordinate system fixed in a room determined with respect to the origin ic of 81. If the working coordinate origin O of the robot manipulator 61 does not move on the coordinate system of the medical linear accelerator 81, the difference between the two measured values is due to the difference in the measurement system. In this case, it is an error that can occur in the three-dimensional position measurement using the stereo X-ray image by the moving object tracking device 80. Template matching is used for marker capture in the fluoroscopic image, and an error may occur in the calculation of the marker center of gravity position in the process.

For the robot manipulator 61 used in this embodiment, anchor bolts are struck on the floor to fix the fixing base 121, and bolt tightening is satisfactory for fixing the robot manipulator 61 to the fixing base 121. The torque value is indicated.

On the other hand, since it was impossible to install anchor bolts in a clinical environment and the metal pedestal could not be used because it adversely affected radiation, the robot manipulator 61 was fixed to the sleeper 86 of the medical linear accelerator 81. Staying. In the process of the preliminary experiment, the bed and the robot manipulator sometimes vibrated to a size that could be clearly confirmed visually. Therefore, as described above, the vibration isolator was used to reduce the vibration. As a result, the vibration that can be visually recognized has disappeared, but it is considered that fine vibration that cannot be visually confirmed remains, and the measured value by the moving object tracking device 80 can also capture the vibration. Therefore, the robot control device. It is considered to be a major factor as the reason why the value becomes larger than the error value at 62.

It is possible to explain by the vibration of the sleeper 86 and the robot manipulator 61 that the evaluation using the moving body tracking device 80 showed the highest accuracy when the third-order spline interpolation was used in 3 out of 4 cases. Conceivable. In the cubic spline, the continuity of the velocity and the acceleration at the connection point of the interpolation is guaranteed, so that the continuity of the manipulated variable generated by the robot control device 62 seems to be high, but in the case of linear interpolation, the velocity at the connection point. Since the acceleration changes in steps and impulses, it is considered that transient vibration is more likely to be induced than in spline interpolation.

When the robot moving body phantom system 60 of the above embodiment is used for quality assurance of radiotherapy, the value when the moving body tracking device 80 is used is meaningful as an evaluation of the influence on the quality. In that sense, in the measurement results in the clinical environment, as a result of error analysis using the moving body tracking device 80, the best value of μ + 2σ was less than 0.8 mm in all four cases, which is good for the robot moving body phantom system 60. It shows the accuracy.

In quality assurance in radiotherapy for tumors showing respiratory movement, treatment is performed on the water equivalent phantom 65 while actually measuring the position of the marker in the phantom using the moving body tracking device 80. In the actual treatment, a marker placed near the tumor is planned for the treatment that irradiates radiation when the tumor showing respiratory movement comes near the specified range of the pre-planned position, which is generally called ambush irradiation. It is a protocol to administer radiation by entering the inside of a cube with a side of 4 mm centered on the position. The result of μ + 2σ <0.8 means that there is almost no situation where administration is not performed due to the positioning error of the water equivalent phantom 65 even though the tumor trajectory is in the center of the cube at the planned position. This is a significant result showing that the accuracy of quality assurance can be improved by using the robot moving object phantom system 65.

In this example, for the purpose of improving the accuracy of quality assurance of radiotherapy, we attempted to construct a 4D robot moving phantom system 60 that can follow a three-dimensional tumor motion trajectory in a patient's body with high accuracy. From the results shown in this example, it was found that the developed robot moving body phantom system 60 can be used for quality assurance in clinical practice.

10 Medical Linear Accelerator 11 Sleeper 12 Robot Manipulator 13 Water Equivalent Phantom 60 Robot Dynamic Phantom System 61 Robot Manipulator 62 Robot Control Device 63 Teaching Pendant 64 External Controller 65 Water Equivalent Phantom 66 Target Trajectory Generator 80 Motion Tracking Device 81 Medical Linear Accelerator 82, 83, 82b, 83b X-ray source 84, 85, 84b, 85b Color image receiving device 86 Sleeper 111 Target trajectory correction unit 112 Target trajectory regeneration unit 121 Fixing table 131 Fixing tool 201 Position acquisition unit 202 Angle calculation unit 203 Position correction Part 204 Angle correction part ic Treatment device Origin of coordinate system O Origin of working coordinate system

Claims (8)

A robot manipulator with three-dimensional translational degrees of freedom, a robot control device that controls the robot manipulator, a phantom fixed to the tip of the robot manipulator and having a radiation absorption rate equivalent to that of the human body, and placed near the tumor. It is equipped with a moving body tracking device that measures changes in the marker position of the patient in real time.
The robot control device has a target trajectory generation unit that generates a target trajectory of the robot manipulator from a three-dimensional motion trajectory of the marker position of the patient, and the marker position in the phantom follows the target trajectory. A robot moving phantom system characterized by controlling a robot manipulator. Further equipped with an external controller for transmitting and receiving information signals to and from the robot control device.
The external controller retains the information on the marker position of the patient and upsamples the trajectory of the marker position, thereby controlling the time interval of the tumor position tracking of the moving body tracking device and the control cycle of the tracking operation with the target trajectory. The robot moving body phantom system according to claim 1, wherein the robot moving body phantom system is made to match. Further equipped with an external controller for transmitting and receiving information signals to and from the robot control device.
The external controller interpolates the measurement result of the time interval and upsamples the locus of the marker position to control the time interval of the tumor position tracking of the moving body tracking device and the control cycle of the follow-up operation with the target trajectory. The robot moving body phantom system according to claim 1, wherein the robots are matched. The target trajectory generation unit includes a position acquisition unit that acquires a three-dimensional motion locus of the marker position of the patient, and an angle calculation unit that calculates the rotation angle of the tumor in which the marker that has acquired the three-dimensional motion trajectory is embedded. The robot moving body phantom system according to any one of claims 1 to 3, wherein the robot moving body phantom system is provided. Further having a target trajectory correction unit for correcting the target trajectory,
The target trajectory correction unit is characterized in that the robot target trajectory is corrected using the velocity or acceleration of the robot manipulator motion trajectory defined by interpolating the trajectory output as a result of tumor position tracking of the moving body tracking device. The robot moving body phantom system according to any one of claims 1 to 4. It also has a target trajectory regeneration unit that regenerates the target trajectory, and calculates the tracking error of the target trajectory at each control sampling point from the control results for the target trajectory given by the target trajectory generation unit, and the information thereof. Is transmitted to the target orbit regeneration unit,
The target trajectory regeneration unit regenerates the target trajectory by re-interpolating in real time during control by adding the received tracking error of the target trajectory to the target value information of the start point of the interpolation section corresponding to the time. The robot moving body phantom system according to claim 5, wherein the robot moving body phantom system is characterized in that. The target trajectory correction unit is characterized by including a position correction unit that corrects the three-dimensional coordinate position of the robot manipulator and an angle correction unit that corrects the rotation angle of the robot manipulator that corrects the three-dimensional coordinate position. The robot moving body phantom system according to claim 5 or 6. The robot moving body phantom system according to any one of claims 1 to 7, wherein the robot control device and the moving body tracking device transmit and receive information signals by using a communication means.

Images