Accelerator and particle beam therapy system

Abstract

This accelerator 1 includes: an ion source 12 for generating ions; a pair of magnetic poles 123 for exciting a magnetic field therebetween; and a de-electrode 221 for applying a high-frequency electric field. The high-frequency electric field is applied to ions circulating in the magnetic field by the de-electrode 221 to accelerate the ions, a plurality of annular beam circulation trajectories which are formed by the pair of magnetic poles 123 and in which ions having different energies circulate are aggregated on one side, and the frequency of the high-frequency electric field is modulated by the beam circulation trajectories. The magnetic field is modulated along the beam circulation trajectories so that a secondary Fourier component of a magnetic field gradient along the beam circulation trajectories formed by the magnetic field is smaller than an average of the magnetic field gradient. As a result, provided are an accelerator and a particle beam therapy system to which a more robust optical design is applied with respect to an irregular component by determining a guideline for the distribution of a magnetic field applied to a beam that is accelerated while circulating inside a variable energy accelerator.

Description

Accelerator and Particle Beam Therapy System

[0001] The present invention relates to an accelerator for accelerating protons or heavy ions such as helium ions and carbon ions, and a particle beam therapy system.

[0002] Patent Document 1 discloses an accelerator used in a particle beam therapy system that can obtain an irradiation beam suitable for particle beam therapy by the spot scanning method, is small, inexpensive, and easy to adjust. The accelerator includes an acceleration device that accelerates a charged particle beam to a predetermined energy and emits it, an irradiation device that emits the charged particle beam to an irradiation target, a deflection electromagnet that deflects the charged particle beam, a beam transport system that guides the charged particle beam emitted from the acceleration device to the irradiation device, and a beam blocking device installed in the beam transport system that blocks the supply of the charged particle beam to the irradiation device. The beam blocking device includes a blocking electromagnet installed upstream of the deflection electromagnet in the traveling direction of the charged particle beam, and a beam dump installed downstream of the deflection electromagnet in the traveling direction of the charged particle beam.

[0003] Patent Document 2 discloses a variable energy accelerator that can precisely control the emission of a charged particle beam from a circular accelerator. In a circular accelerator that accelerates a charged particle beam by applying a first high frequency in a main magnetic field while increasing the orbital radius, a second high frequency having a different frequency from the first high frequency is applied in an aggregation region of the orbit of the charged particle beam inside the circular accelerator or on the maximum energy orbit of the charged particle beam, thereby emitting the charged particle beam.

[0004] Japanese Patent No. 4691576, Japanese Patent No. 7002952

[0005] In particle beam therapy, an ion beam with energy corresponding to the depth from the body surface is irradiated to a cancer tumor to be irradiated.

[0006] The kinetic energy of the ions to be irradiated is required to be around 200 MeV per nucleon in the case of protons, and an accelerator is used to obtain an ion beam with such energy. <00During irradiation, the three-dimensional dose to be applied to the affected area is planned in advance. To apply the dose to the affected area according to the plan, there are methods such as the spot scanning irradiation method, which involves scanning the ion beam with an electromagnet to match the shape of the affected area, as described in Patent Document 1.

[0008] In the spot scanning irradiation method, the affected area is divided into layers according to the depth from the body surface, and an ion beam is irradiated to each layer with a predetermined energy corresponding to the depth from the body surface.

[0009] From the perspective of reducing the burden on patients, it is desirable to complete irradiation in a short period of time. Therefore, the beam output from the accelerator requires energy reproducibility and the maintenance of high beam output.

[0010] Furthermore, there are several types of accelerators used in particle beam therapy, including synchrotrons, cyclotrons, synchrocyclotrons, and variable energy accelerators as described in Patent Document 2.

[0011] In particle beam therapy, when delivering a predetermined radiation dose to the affected area according to the treatment plan, it is desirable to complete the irradiation in a shorter time, in terms of reducing the physical burden on the patient and enabling more treatments to be performed.

[0012] In the particle beam therapy system described in Patent Document 1, the spot scanning irradiation method involves gradually extracting the beam stored in the accelerator and irradiating the irradiation spot. If a larger amount of beam can be stored in the accelerator, the number of spots that can be irradiated with a single acceleration increases, and the time spent accelerating and decelerating the beam that cannot be irradiated during the treatment period can be reduced.

[0013] However, the amount of beam that can be accelerated is determined by the size of the aperture in phase space and the magnitude of the beam trajectory deviation caused by the imperfections in the magnetic field excited along the accelerator's beam path. Increasing the physical aperture increases costs, such as by using larger magnets, while reducing the imperfections in the magnetic field inevitably increases costs for adjusting the magnetic field and manufacturing the magnets themselves.

[0014] Therefore, it is desirable to reduce the sensitivity of the beam trajectory to the imperfections in the magnetic field.

[0015] However, in the variable-energy accelerator described in Patent Document 2, the behavior of the beam in response to imperfections is not clear, and optical design guidelines for beam stabilization and the resulting storage of high-current beams are not yet established, revealing that there is room for improvement.

[0016] The object of the present invention is to provide a variable-energy accelerator and particle beam therapy system to which an optical design more robust to imperfections than conventional designs is applied.

[0017] The present invention includes multiple means for solving the above problems, but to give one example, it includes an ion source that generates ions, a pair of magnetic poles that excite a magnetic field between them, and a high-frequency electrode that applies a high-frequency electric field, and accelerates ions orbiting in the magnetic field by applying the high-frequency electric field with the high-frequency electrode, and converges a plurality of annular beam orbits formed by the pair of magnetic poles, each of which ions of different energies orbit, and modulates the frequency of the high-frequency electric field with respect to the beam orbits, wherein the magnetic field is modulated along the beam orbits such that the second-order Fourier component of the magnetic field gradient along the beam orbit formed by the magnetic field is smaller than the average of the magnetic field gradient.

[0018] According to the present invention, it is possible to provide an optical design that is more robust to imperfections than conventional designs. Other problems, configurations, and effects will be clarified by the following description of the examples.

[0019] This figure shows the overall configuration of the particle beam therapy system of the example. This figure shows the overall configuration of the accelerator of the example. This figure shows the internal structure of the accelerator of the example. This figure shows the relationship between the beam energy and orbital frequency during orbit in the accelerator of the example. This figure shows the design trajectory shape of the accelerator of the example. This figure shows the relationship between the beam energy and the magnetic field on the design trajectory during orbit in the accelerator of the example. This figure shows the β function along the trajectory of the beam during orbit in the accelerator of the example and the accelerator of the comparative example. This figure shows the magnetic field strength along the trajectory of the beam during orbit in the accelerator of the example. This figure shows the magnetic field strength in the trajectory plane of the accelerator of the example as a contour plot. This figure shows an example of a configuration for exciting the characteristic magnetic field in the accelerator of the example. This figure shows another example of a configuration for exciting the characteristic magnetic field in the accelerator of the example. This figure shows yet another example of a configuration for exciting the characteristic magnetic field in the accelerator of the example. This figure shows a schematic of the high-frequency modes excited at the D-electrode in the accelerator of the example. This figure shows the timing chart of each device during operation in the particle beam therapy system of the example. This figure shows the connection of the control device in the particle beam therapy system of the example. This figure shows the control flow in the particle beam therapy system of the example.

[0020] Examples of the accelerator and particle beam therapy system of the present invention will be described with reference to Figures 1 to 16. In the drawings used herein, the same or corresponding components are denoted by the same or similar reference numerals, and repeated descriptions of these components may be omitted.

[0021] First, the overall configuration of the particle beam therapy system 1000 will be explained using Figure 1. Figure 1 is an overall configuration diagram of the particle beam therapy system 1000 in this embodiment.

[0022] The particle beam therapy system 1000 shown in Figure 1 includes an accelerator 1 having electrodes that accelerates the beam and extracts it to the outside, a blocking device (not shown) that blocks the accelerated beam before transporting it to the downstream side, a beam transport device 2 that transports the beam extracted from the accelerator 1, an irradiation device 3, a treatment table 4, an overall control device 40 that controls the accelerator 1 and the beam transport device 2, an irradiation control device 50, a treatment plan database 60, and a treatment planning device 70.

[0023] In the particle beam therapy system 1000, the beam generated by the accelerator 1 is transported to the irradiation device 3 via the beam transport device 2. The irradiation device 3 controls the position of the beam using the magnetic field of a scanning electromagnet and irradiates a predetermined affected area of ​​the patient 5 who is lying supine on the treatment table 4.

[0024] The treatment planning device 70 is a device that calculates and determines various control parameters related to the irradiation of the beam to the patient 5, and the created treatment plan is recorded in the treatment planning database 60.

[0025] The overall control unit 40 controls each device based on the treatment plan obtained from the treatment plan database 60. The overall control unit 40 also determines the target energy to be accelerated based on the treatment plan and transmits predetermined command values ​​to each device.

[0026] During the irradiation of the patient 5 with the beam, the irradiation device 3 constantly measures the irradiated dose and outputs the measurement results to the irradiation control device 50. The irradiation control device 50 is a control device for monitoring the irradiation dose and irradiation position, and when irradiation of a predetermined planned dose is completed at a certain spot, it performs irradiation at the next spot. By repeating this process, the dose distribution specified in the treatment plan predetermined by the treatment planning device 70 can be applied to the appropriate position and depth.

[0027] The accelerator 1 in this embodiment is a frequency-modulated variable-energy accelerator in which the center of the circulating beam trajectory does not coincide with the circular center of the accelerator 1. This accelerator is a circular accelerator that has a time-constant magnetic field as its main magnetic field and accelerates protons circulating in the main magnetic field with a high-frequency electric field. Its external appearance is shown in Figure 2.

[0028] The accelerator 1 in this embodiment is a frequency-modulated variable-energy accelerator. This accelerator is a circular accelerator that has a time-constant magnetic field (static magnetic field) as its main magnetic field and accelerates protons orbiting within the main magnetic field using a high-frequency electric field. Its external appearance is shown in Figure 2.

[0029] As shown in Figure 2, the accelerator 1 uses electromagnets 11, which can be divided into upper and lower sections, to excite a main magnetic field within the region through which the accelerating and circulating beam passes (hereinafter referred to as the beam passage region 20 (see Figure 3)). The beam passage region 20 inside the electromagnets 11 is evacuated by a vacuum pump (not shown).

[0030] The electromagnet 11 is provided with multiple through-holes that connect the outside to the beam passage region 20. For example, various through-holes such as an extraction beam through-hole 111 for extracting the accelerated beam, coil connection through-holes 112 and 113 for drawing out coil conductors placed inside the electromagnet 11 to the outside, and a high-frequency power input through-hole 114 are provided on the surface of the upper and lower divided connection surfaces.

[0031] A high-frequency accelerating cavity 21 is installed through a high-frequency power input through-hole 114 to form an accelerating electric field for accelerating ions into a beam. As will be described later, the high-frequency accelerating cavity 21 is equipped with a D-electrode 221 (see Figure 3) for applying a high-frequency electric field to ions orbiting in a magnetic field to accelerate them, and a rotary variable capacitance capacitor 212 for modulating the frequency of the accelerating electric field according to the beam orbit.

[0032] An ion source 12 for supplying hydrogen ions and a beam injection port 115 for guiding the beam extracted from the ion source 12 into the beam passage region 20 are installed at a position offset from the center of the upper part of the electromagnet 11. Ions are injected between the electromagnets 11 inside the accelerator 1 through the beam injection port 115 and the injection section 130 (see Figure 3). Power necessary for the injection of ions into the beam passage region 20 is supplied to the injection section 130 from the outside through the beam injection port 115.

[0033] Next, the internal structure of accelerator 1 will be explained using Figures 2 and 3. Figure 3 shows the arrangement of equipment when the electromagnet 11 is divided into upper and lower sections and viewed from above.

[0034] As shown in Figure 2, the upper and lower parts of the electromagnet 11 each have a cylindrical return yoke 121 and a top plate 122, and inside them, as shown in Figure 3, there is a pair of cylindrical magnetic poles 123 that excite a magnetic field. The beam passage region 20 is located within the cylindrical space sandwiched between the opposing magnetic poles 123. The surfaces on which these upper and lower magnetic poles 123 face each other in mirror image are defined as the magnetic pole surfaces. The planes that are equidistant from the upper and lower magnetic pole surfaces sandwiched between them are called the orbital planes.

[0035] In the recess formed between the magnetic pole 123 and the return yoke 121, an annular coil 13 is installed along the outer wall of the magnetic pole 123. By passing an electric current through the coil 13, the opposing magnetic poles 123 are magnetized, and a magnetic field is excited in the beam passage region 20 with a predetermined distribution, which will be described later.

[0036] The high-frequency accelerating cavity 21 excites an accelerating high-frequency electric field in the acceleration gap 223 to accelerate ions by a λ / 4 type resonant mode. The portion of the high-frequency accelerating cavity 21 that is fixedly installed relative to the accelerator 1 is defined as the D-electrode 221. The high-frequency accelerating cavity 21 is installed through a high-frequency power input through-hole 114 and forms a D-electrode 221 that surrounds a portion of the beam passage region 20.

[0037] Ions are accelerated by a high-frequency electric field excited in the acceleration gap 223, which is the region sandwiched between the D electrode 221 and the ground electrode 222 positioned opposite the D electrode 221. In order for the high-frequency electric field to synchronize with the beam's reciprocating frequency, the frequency of the high-frequency electric field must be an odd multiple of the beam's reciprocating frequency. In this accelerator 1, the frequency of the high-frequency electric field is set to 1x the beam's reciprocating frequency.

[0038] In this type of accelerator 1, ions generated in the ion source 12 are extracted into the beam passage region 20 in a low-energy ion state by a voltage applied to the extraction electrode of the injection section 130. Each time the incident ions pass through the acceleration gap 223, they are accelerated by the high-frequency electric field excited by the high-frequency accelerating cavity 21, becoming a high-energy ion beam.

[0039] Furthermore, as shown in Figure 3, in order to extract the beam outside the accelerator 1, two additional magnetic field generating shims 311 that excite a quadrupole magnetic field or a multipole magnetic field of six or more poles, and a disturbance high-frequency electrode 313 that applies a high-frequency electric field are installed electrically insulated from a part of the magnetic pole surface. An extraction outlet is installed at one end of the magnetic pole surface.

[0040] The disturbance-generating high-frequency electrode 313 can be used to apply a high-frequency (RF) electric field (disturbance-generating high frequency) with a small amplitude, which kicks the orbiting particle in the direction of its orbital plane, causing the particle to deviate from its designed orbit. The particle whose orbit has deviated from the designed orbit passes near the additional magnetic field generating shim 311.

[0041] The magnetic field generated by the additional magnetic field generating shim 311 restricts the stable region of motion in the horizontal plane for the ion beam orbiting within the beam passage region 20, and introduces particles that have been kicked out of the stable region by the disturbing high-frequency signal into the extraction septum electromagnet 312. In the accelerator 1 of this embodiment, the pair of additional magnetic field generating shims 311 each superimpose and excite magnetic fields of opposite polarity onto the main magnetic field formed by the magnetic poles 123.

[0042] By applying a high-frequency voltage of an appropriate frequency to the disturbance high-frequency electrode 313, disturbance is introduced to the beam, and the on / off control of the beam becomes possible in synchronization with the on / off of the RF electric field applied to the disturbance high-frequency electrode 313, according to the principle described later. Details of the additional magnetic field generating shim 311 and the disturbance high-frequency electrode 313 will be described later.

[0043] In accelerator 1, the shapes and arrangements of the upper and lower magnetic poles 123, coil 13, shim 311 for generating additional magnetic field, extraction septum electromagnet 312, and disturbance high-frequency electrode 313 are designed so that the in-plane component of the main magnetic field is almost zero in the orbital plane, resulting in a symmetrical arrangement and current distribution with respect to the orbital plane. Furthermore, as shown in Figure 3, the shapes of the magnetic poles 123, D-electrode 221, coil 13, and disturbance high-frequency electrode 313 are symmetrical with respect to the line segment connecting the center of the high-frequency power input through-hole 114 and the center of the coil connection through-hole 112 when accelerator 1 is viewed from above.

[0044] Next, the orbit and motion of the beam circulating in the accelerator 1 of this embodiment will be described.

[0045] The beam is accelerated while circulating in the beam passage region 20. The kinetic energy of the extractable beam in the accelerator 1 of this embodiment is at least 70 [MeV] and at most 235 [MeV]. The higher the kinetic energy, the lower the beam's circulating frequency. The beam with the kinetic energy immediately after incidence has a circulating frequency of 76 [MHz], and the beam that has reached 235 [MeV] circulates in the beam passage region 20 at 59 [MHz]. The relationship between these energies and the circulating frequency is as shown in FIG. 4.

[0046] The designed orbits of the beams with each energy are shown in FIG. 5. In FIG. 5, the designed orbit is a circular orbit with a radius of 0.497 [m] corresponding to the orbit with the maximum energy of 250 [MeV] on the outermost side. Also, the 0 [MeV] orbit is defined as a single point at the position of the ion source 12. The designed orbits of the energies between them are defined as circles. The dotted line is a line connecting the same circulating phases of each orbit and is called an equal-circulating phase line.

[0047] As shown in FIG. 5, in the accelerator 1 of this embodiment, as the beam is accelerated, the orbit center (designed orbit) of the beam moves in one direction within the orbit plane. As a result of the movement of the designed orbit, there are regions where the orbits with different kinetic energies are close to each other (regions where the circulating orbits converge) and regions where they are far apart (regions where the circulating orbits are discrete). That is, the designed orbit of the beam is eccentric.

[0048] Connecting each point of the designed orbits that are closest to each other results in a line segment perpendicular to all the designed orbits. Also, connecting each point of the designed orbits that are farthest from each other results in a line segment perpendicular to all the designed orbits, and these two line segments exist on the same straight line. If this straight line is defined as the axis of symmetry, the shape of the designed orbit is symmetric with respect to the plane passing through the axis of symmetry and perpendicular to the orbit plane.

[0049] The equal-circulating phase lines shown in FIG. 5 are those with a circulating phase of ±π / 2 from the converging region. As shown in FIG. 5, the acceleration gap 223 formed between the dee electrode 221 and the grounded electrode 222 facing the dee electrode 221 is installed along this equal-circulating phase line.

[0050] In order to create the orbital configuration described above and generate stable beam oscillations around the orbit, accelerator 1 in this embodiment reduces the magnitude of the average magnetic field on the orbit as it moves outward in the direction of the deflection radius of the design orbit, resulting in the relationship between the beam energy and the average magnetic field on the orbit shown in Figure 6. In other words, it creates a magnetic field distribution in which the magnetic field on the orbit decreases as the beam energy increases, and where the magnetic field decreases on the radially outward side.

[0051] In this system, particles that deviate slightly radially from the design trajectory receive a restoring force that returns them to the design trajectory, while particles that deviate perpendicularly to the trajectory plane also receive a restoring force from the main magnetic field that returns them to the trajectory plane. In other words, if the magnetic field is appropriately reduced relative to the beam energy, particles that deviate from the design trajectory will always experience a restoring force acting in the direction that tries to return them to the design trajectory, causing them to oscillate near the design trajectory.

[0052] This mechanism allows the beam to oscillate stably in both the radial direction within the orbital plane and in the direction perpendicular to the orbital plane. This oscillation around the design orbit is called betatron oscillation. As a result of betatron oscillation, particles that deviate from the design orbit can remain near the design orbit, making it possible to stably orbit and accelerate the beam.

[0053] The betatron oscillation in accelerator 1 of this embodiment will be described in more detail.

[0054] In synchrotrons and cyclotrons, as well as in variable-energy accelerators like accelerator 1 in this embodiment, the focusing force on the orbit generally changes depending on the position on the orbit. In a synchrotron, focusing elements such as quadrupole magnets and deflection magnets are placed on the beam orbit. In a cyclotron, when adopting a Hill-Valley type magnetic field distribution, the beam is focused by an effect called edge focusing at the transition point between the Hill region and the Valley region.

[0055] In a system where the converging force changes along the orbit, the betatron oscillation does not become a simple harmonic motion. Specifically, there exists a beta function denoted as β(s) as a function of the position along the orbit, and each particle oscillates with an amplitude proportional to its square root. Although the phase of the oscillation varies among individual particles, the betatron oscillation phase that advances over one orbit takes a constant value within the range of linear optics, independent of the initial state of the particle. The value obtained by dividing the betatron oscillation phase advance over one orbit by 2π is called the tune.

[0056] Under a uniform magnetic field, the horizontal tuning is 1 and the vertical tuning is 0. By weakening the magnetic field on the outer edge, a convergence force is generated due to the magnetic field gradient, causing the vertical tuning to increase and the horizontal tuning to a value slightly less than 1 (for example, 0.97). This state of tuning is called weak convergence.

[0057] Furthermore, the reciprocal of the β function (1 / β) represents the phase lead of the betatron oscillation per unit length. Therefore, the line integral of 1 / β(s) over one orbit is 2π × tune. In the weakly convergent state, the horizontal tune is approximately 1, making it technically impossible to significantly change the average value of 1 / β, and thus the average value of β cannot be significantly changed either. Therefore, changing the distribution of the magnetic field gradient essentially does not change the average value of the β function much, only the state of oscillation (balance of amplitude and frequency components) changes.

[0058] Now, in accelerators with eccentric orbits, such as accelerator 1 in this embodiment, conventionally, examples have been given in which the magnetic field on the orbit is made uniform, as described in Patent Document 2. However, due to the eccentric orbit, there is a large bias in the distribution of the gradient along the design orbit.

[0059] For example, if a uniform magnetic field is adopted along the orbit, the magnetic field gradient will be large in the orbital convergence region where the deflection angle is 0, and conversely, the gradient will be small in the region where orbits are discrete and the deflection angle is close to π. In a system where the gradient is localized in this way, the beta function along the orbit will oscillate.

[0060] When oscillations occur in the beta function, the following technical difficulties arise due to the properties of the beta function.

[0061] Since the β function is a quantity related to the beam size during orbit, regions with a larger beam size will occur during orbit compared to the case where the β function is uniform. Therefore, if there are imperfect magnetic field components in those regions that are proportional to the square or cube of the displacement r, the amount of kick the beam experiences will increase.

[0062] Furthermore, if there are error magnetic fields at other locations in orbit, the beam will shift significantly at locations where the beta function is large.

[0063] As described above, the β function exhibits properties that indicate sensitivity to error magnetic fields. Therefore, when the maximum value of the β function, i.e., the amplitude, is large, the tolerance value for error magnetic fields becomes small.

[0064] Furthermore, a problem specific to the weakly convergent state was revealed to be that when the horizontal tune is close to 1, and the beta function contains a component that oscillates at half the wavelength of the orbital circumference, it causes the following instability under certain error magnetic fields.

[0065] When the beta function oscillates at a wavelength half the orbital length, the beta function becomes a function with two peaks on the orbit, as shown in Figure 7. If an error magnetic field is generated at one of the peaks, causing the beam to "kick," in a system with a tuning close to 1, the beam displacement will be maximum at the other peak position, which is isoanisotropic to the peak position where the kick occurred.

[0066] Furthermore, if a kick opposite to the above occurs at another peak, the beam displacements due to the error magnetic fields are superimposed in a direction that reinforces each other. When an error magnetic field is generated at a certain point in the orbit and an error magnetic field in the opposite direction is generated 180 degrees to the opposite side, when this error magnetic field is viewed globally, the Z component of the magnetic field is distributed with a gradient in a certain direction within the orbital plane.

[0067] There are many causes for the occurrence of such a large-scale error magnetic field. For example, such an error magnetic field can be caused by various factors such as the relative positional misalignment of the upper and lower divided parts of the electromagnet 11, the positional misalignment or tilt of the opposing magnetic poles 123, the installation error of the coil 13, and the manufacturing error of the magnetic poles.

[0068] For example, a common example of a magnetic field distribution used as the main magnetic field in conventional variable energy accelerators was a uniform magnetic field distribution along the orbit. However, with a uniform magnetic field along the orbit, the magnetic field gradient is large in the aggregated region and small in the discrete region. As a result, large oscillations of the beta function are observed, as shown in Figure 7.

[0069] Therefore, the aforementioned response to the error magnetic field occurs, resulting in behavior sensitive to the error magnetic field. Consequently, it has been found that conventional variable energy accelerators have room for improvement, as the allowable error of the magnetic field becomes small, making adjustment difficult, or the area in the phase space where a predetermined focusing force can be obtained and acceleration can be stably achieved decreases due to the orbital shift caused by the error magnetic field, and the amount of beam that can be accelerated and irradiated decreases.

[0070] Therefore, in accelerator 1 of this embodiment, the oscillation of the β function is suppressed by devising the magnetic field distribution, thereby realizing a system that is more robust to error magnetic fields. Specifically, the magnetic field distribution oscillates along the orbit as shown in Figure 8. That is, it has the following characteristics with respect to the distribution of the magnetic field along the beam trajectory.

[0071] As shown in Figure 8, the average magnitude of the magnetic field in the range of ±π / 8 from the concentration region (section "A" in Figure 8) is greater than the average magnitude of the magnetic field for the entire orbit. The average magnitude of the magnetic field in the range of ±π / 8 from the position deflected π / 2 upstream and downstream from the position where the beam orbit is closest (sections "B" and "C" in Figure 8) is smaller than the average magnitude of the magnetic field for the entire orbit. The average magnitude of the magnetic field in the range of ±π / 8 from the position deflected π upstream and downstream from the position where the beam orbit is closest (section "D" in Figure 8) is greater than the average magnitude of the magnetic field for the entire orbit.

[0072] In this way, the magnetic field itself oscillates, causing oscillations in the converging force. By making these oscillations have an appropriate distribution, the oscillations of the β function can be suppressed, as shown in Figure 7.

[0073] This principal magnetic field is schematically shown in Figure 9 as the distribution of the gap-direction component of the magnetic field on the orbital plane. As the beam energy increases, the average magnetic field strength on the orbit decreases, and the magnetic field distribution oscillates along the beam trajectory, resulting in elliptical contour lines within the orbital plane.

[0074] Furthermore, the distribution is bell-shaped in both the left-right and right-right directions, as shown in Figure 9, with the beam incidence point being the peak. Comparing the bell-shaped distribution in the left-right and up-down directions, the fluctuation range is larger in the left-right direction (deflection angle ±π / 2) than in the up-down direction (deflection angle 0 or π). In such a magnetic field, unlike conventional accelerators, the magnetic field gradient along the orbit is not maximum in the concentrated region (deflection angle 0).

[0075] If a large magnetic field gradient exists only near the convergence region, as in conventional methods, the β function oscillates at a wavelength of half the orbital length, as mentioned above. The amplitude of this oscillation is determined by the component that oscillates at a wavelength of half the orbital length when the magnetic field gradient is Fourier-decomposed over one revolution of the orbit. In conventional examples, the magnetic field gradient distribution is localized in the convergence region, so when Fourier-decomposed, numerous harmonic components are generated.

[0076] In contrast, in the accelerator 1 of this embodiment, the magnetic field is modulated along the beam orbit so that the second-order Fourier component of the magnetic field gradient along the beam orbit formed by the magnetic field is smaller than the average of the magnetic field gradient. Furthermore, it is desirable that the magnetic field that excites the magnetic pole 123 has the largest amplitude among the first-order and higher Fourier components of the magnetic field gradient along the beam orbit.

[0077] If we can create a magnetic field distribution with a gradient at a deflection angle of ±π / 2, as shown in this magnetic field distribution, it is possible to cancel out the second harmonic component of the magnetic field gradient. This cancels out the oscillation component of the beta function, resulting in a flatter beta function.

[0078] The main magnetic field distribution described above is excited when a predetermined excitation current is passed through the coil 13, thereby magnetizing the magnetic pole 123. Furthermore, the shape of the magnetic pole 123 is symmetrical with respect to the orbital plane, and on the orbital plane, it has only a magnetic field component perpendicular to the orbital plane.

[0079] For the magnetic field excited by the electromagnet 11 to satisfy the aforementioned characteristics, for example, the average value of the magnetic pole 123 gap size formed by the magnetic pole 123 in a range of ±π / 8 from the position deflected π / 2 upstream and π / 2 downstream from the position where the beam orbit is closest can be made larger than the average value of the magnetic pole 123 gap size for the entire orbit, and the average value of the magnetic pole 123 gap size formed by the magnetic pole 123 in a range of ±π / 8 from the position deflected π upstream and π downstream from the position where the beam orbit is closest can be made smaller than the average value of the magnetic pole 123 gap size for the entire orbit.

[0080] Specifically, the magnetic poles 123 can have irregularities on their pole surfaces. In particular, the magnetic field of the accelerator 1 in this embodiment is characterized by the formation of magnetic field minima in the concentrated and discrete regions, and in order to form a magnetic field minima at a deflection angle of ±π / 2, more magnetic material can be placed on the magnetic poles 123 directly above and below the positions where a high magnetic field is required, and conversely, fewer magnetic material can be placed on the magnetic poles 123 directly above and below the positions where a small magnetic field is required, thereby increasing the gap size.

[0081] More specifically, as shown in Figure 10, the magnetic members 140 are arranged on the surface of the magnetic pole 123 such that the gap size is large in the region where the deflection angle from the convergence point is ±π / 2, and small in the region where the deflection angle from the convergence point is 0 or π, along the trajectory.

[0082] In addition, coils 141 and 142 (see Figures 11 and 12) with current paths arranged to enclose at least two of the following ranges: a range of ±π / 8 from the position where the beam orbit is closest, a range of ±π / 8 from a position deflected π / 2 upstream and π / 2 downstream from the position where the beam orbit is closest, and a range of ±π / 8 from positions deflected π upstream and downstream from the position where the beam orbit is closest, can be provided within the gap of a pair of magnetic poles 123.

[0083] Specifically, this means for forming a magnetic field distribution not only involves creating irregularities on the magnetic pole 123 by attaching a magnetic member 140 to the magnetic pole 123, but also, as shown in Figure 11, by arranging coils 141 around concentrated and discrete regions where the magnetic field should be relatively increased, and by passing an appropriate current through the coils 141 to excite an additional magnetic field, thereby realizing a predetermined magnetic field distribution.

[0084] Alternatively, as shown in Figure 12, coils 142 can be placed around deflection angles ±π / 2 where the magnetic field should be relatively reduced, and by passing an appropriate current through the coils 142, a magnetic field opposite to the main magnetic field can be excited, thereby achieving a predetermined magnetic field distribution.

[0085] Furthermore, it is also possible to incorporate magnetic materials with different permeability (such as magnetic alloys like permalloy and permendur, or magnetic materials other than iron, such as nickel).

[0086] Regardless of the method used, to increase the charge amount of a beam that can be stably accelerated, it is effective to reduce the second-order Fourier component of the magnetic field gradient, that is, the component that oscillates at half the wavelength of the orbit. For this purpose, the magnetic field is made to be maximum in the concentrated and discrete regions, and minimum at the deflection angle ±π / 2.

[0087] Therefore, there are magnetic field distributions other than the one shown in this embodiment that satisfy this characteristic, and the effects of the present invention can be obtained even if the positions of the magnetic field maxima and minimum are shifted by about π / 8. Furthermore, even in magnetic field distributions where there are no clear maxima and minimum, the effects of the present invention can be obtained if the average value of the magnetic field in the range of ±π / 8 from the four points with deflection angles of 0, ±π / 2, and π is greater than or less than the average.

[0088] To achieve such a magnetic field, in the design examples shown in Figures 10 to 16, magnetic materials are arranged in the range from four points with deflection angles of 0, ±π / 2, and π to ±π / 8 (regions "A", "B", "C", and "D" in Figure 10), or coils are arranged so that region "A" and region "C", or region "B" and region "D" are included in their inner circumference.

[0089] Thus, the average magnitude of the magnetic field within ±π / 8 of the position where the beam orbits with infinitesimally different energies are closest together can be set to be greater than the average magnitude of the magnetic field for the entire orbit. The average magnitude of the magnetic field within ±π / 8 of the position deflected π / 2 upstream and π / 2 downstream from the position where the beam orbits are closest together can be set to be smaller than the average magnitude of the magnetic field for the entire orbit. The average magnitude of the magnetic field within ±π / 8 of the position deflected π upstream and π downstream from the position where the beam orbits are closest together can be set to be greater than the average magnitude of the magnetic field for the entire orbit.

[0090] As described above, the high-frequency accelerating cavity 21 excites an electric field in the acceleration gap 223. To this end, high-frequency power is introduced from an external high-frequency power supply (see Figure 9, low-level high-frequency generator 42 and high-frequency amplifier 43) through the input coupler 211, and a high-frequency electric field is excited in the acceleration gap 223 between the D electrode 221 and the ground electrode 222.

[0091] Generally, the electromagnetic field excited by a D-electrode is an electromagnetic field with a specific resonant frequency and spatial distribution determined by the electrode shape. Electromagnetic fields with a specific frequency and spatial distribution are called eigenmodes, and there are multiple types of eigenmodes. The mode excited for acceleration is called the fundamental mode.

[0092] Figure 13 shows the electromagnetic field distribution and surface current distribution in the fundamental mode. In Figure 13, the outline of the resonator is shown by a thick line, the electric field distribution by a thick arrow (E), the magnetic field distribution by a dotted arrow (B), and the general shape of the current distribution on the resonator surface by a solid arrow. In the fundamental mode, an electric field in phase is generated from the dee electrode 221 to the ground electrode 222 at all points in the gap.

[0093] In the accelerator 1 of the present invention, the frequency of the electric field is modulated in accordance with the energy of the circulating beam in order to excite the high-frequency electric field in synchronization with the revolving beam. In a high-frequency accelerating cavity 21 using a resonant mode as used in the present invention, it is necessary to sweep the high-frequency range over a wider range than the resonance width. For this reason, the resonant frequency of the high-frequency accelerating cavity 21 also needs to be changed.

[0094] This control is performed by changing the capacitance of a rotary variable capacitance capacitor 212 installed at the end of the high-frequency accelerating cavity 21. The rotary variable capacitance capacitor 212 controls the capacitance generated between a conductive plate directly connected to the rotating shaft 213 and an external conductor by the rotation angle of the rotating shaft 213. In other words, the rotation angle of the rotating shaft 213 is changed as the beam accelerates.

[0095] Next, we will describe the behavior of the beam from beam injection to extraction in the accelerator 1 of this embodiment.

[0096] First, low-energy ions are output from the ion source 12, and the beam is guided into the beam passage region 20 through the beam injection port 115 and the injection section 130.

[0097] As the beam is injected into the beam passage region 20, it is accelerated by the high-frequency electric field, increasing its energy and the radius of rotation of its trajectory. Subsequently, the beam is accelerated while ensuring directional stability with the high-frequency electric field.

[0098] In other words, the center of gravity of the beam does not pass through the acceleration gap 223 at the time when the high-frequency electric field is at its maximum, but rather passes through the acceleration gap 223 when the high-frequency electric field is decreasing over time. As a result, since the frequency of the high-frequency electric field and the reciprocating frequency of the beam are synchronized in an integer multiple ratio, particles accelerated at a predetermined phase of the accelerating electric field will be accelerated at almost the same phase in the next turn.

[0099] On the other hand, particles accelerated at a phase earlier than the acceleration phase receive a greater acceleration than particles accelerated at the acceleration phase, and therefore receive acceleration at a later phase in the next turn. Conversely, particles accelerated at a phase later than the acceleration phase receive a smaller acceleration than particles accelerated at the acceleration phase, and therefore receive acceleration at an advanced phase in the next turn.

[0100] Thus, particles whose timing is off from a predetermined acceleration phase move in the direction of returning to the acceleration phase, and this action allows them to oscillate stably within the phase plane (direction of motion) consisting of momentum and phase. This oscillation is called synchrotron oscillation. In other words, while accelerating, particles are gradually accelerated while undergoing synchrotron oscillation, reaching the predetermined energy at which they are extracted. While undergoing stable synchrotron oscillation, individual particles rotate within a stable region called a high-frequency bucket on the phase plane.

[0101] In order to extract a predetermined extraction beam at the target energy, the high-frequency electric field applied to the high-frequency accelerating cavity 21 is gradually reduced, and the output from the external high-frequency power supply is controlled by the overall control device 40 so that the amplitude of the high-frequency electric field becomes approximately zero when the target energy is reached.

[0102] Through this process, the beam reaches near its target energy and then orbits stably. Then, high-frequency current is applied to the disturbance-generating high-frequency electrode 313. The frequency of the high-frequency current matches the frequency of the beam's betatron oscillation, and the beam is subjected to disturbances that depend on its position in the direction of propagation, i.e., the time it passes through the disturbance-generating high-frequency electrode 313.

[0103] When focusing on a specific particle, the frequency of the disturbance electric field and the frequency of the orbital betatron oscillation coincide, causing them to resonate and increasing the amplitude of the betatron oscillation of that particle. As the betatron oscillation amplitude continues to increase, the betatron oscillation rapidly diverges due to the action of a kick magnetic field excited by the additional magnetic field generating shim 311, which is installed outside the design trajectory, and the beam is displaced outward from the design trajectory. As a result, the beam is introduced into the extraction septum electromagnet 312. This boundary between the stable and unstable regions is called the separatrix.

[0104] From the time the target energy is reached until extraction, the individual particles constituting the beam orbit in a phase space determined by the beam's horizontal position and inclination, divided into a region where they can orbit stably and a region where the orbital deviation continues to increase unstably, due to the quadrupole magnetic field and a multipole magnetic field of six or more poles originating from the additional magnetic field generating shim 311.

[0105] In accelerator 1, particles inside the separatrix continue to stably undergo betatron oscillations, but particles outside the separatrix accumulate kick effects from the additional magnetic field generating shim 311 with each orbit, resulting in large horizontal displacements relative to the design trajectory.

[0106] Particles that have undergone a large horizontal displacement are transported out of the accelerator 1 by passing along the extraction trajectory 322, due to the disturbance electric field formed by the disturbance high-frequency electrode 313 (described later) and the magnetic field formed by the extraction septum electromagnet 312 on the pre-installed extraction trajectory 322.

[0107] In this process, to increase the number of particles that can be stably accelerated to a predetermined energy within the separatrix, i.e., the beam volume, it is crucial that the center position and angle of the beam orbiting within the separatrix do not deviate significantly from the design trajectory.

[0108] Furthermore, when the electric field applied to the disturbance high-frequency electrode 313 is cut off, the increase in the betatron oscillation amplitude of the beam stops, and the beam circulates within the stable region, thus allowing the beam extraction to be stopped.

[0109] The above describes the beam's behavior from injection to extraction during one cycle, but accelerator 1 repeats similar operations in subsequent cycles.

[0110] In other words, accelerator 1 operates in a pattern consisting of an injection process, an acceleration process, and an extraction process. The extraction process is completed when most of the circulating beam has been extracted and it is determined that the remaining amount of the circulating beam has fallen below a specified value, or when irradiation in a certain layer is completed. Once the extraction process is complete, it returns to the injection process.

[0111] The next injection process begins when the rotation angle of the rotary variable capacitance capacitor 212 reaches a predetermined angle and the resonant frequency of the D electrode 221 becomes a value suitable for beam injection.

[0112] In the treatment plan defined by the treatment planning device 70, beams of multiple energies are typically irradiated onto the affected area of ​​the patient 5. In the particle beam therapy system 1000 of this embodiment, when irradiating with beams of multiple energies, the energy of the beams generated by the accelerator 1 is changed and irradiated onto the affected area sequentially.

[0113] When irradiation of one layer is complete and it is time to irradiate a different layer, accelerator 1 generates a beam with a different energy than the one used during previous operation.

[0114] Accelerator 1 discards the orbiting beam when performing a layer switch.

[0115] One method of discarding the beam is to energize a shielding electromagnet (not shown) installed in the beam transport device 2 and then extract the beam from the accelerator 1 in that state. In other words, beam discarding can be achieved by applying a high-frequency electric field to the disturbance high-frequency electrode 313 while the shielding electromagnet is energized.

[0116] Another method of beam discarding is to apply an accelerating electric field to the D-electrode 221 instead of the disturbing high frequency of the disturbing high-frequency electrode 313, thereby achieving a similar effect. To extract the beam with an accelerating electric field, the frequency of the accelerating electric field must match the betatron frequency of the beam. Since the frequency of the accelerating electric field is periodically modulated, almost the entire beam is discarded at a specific timing in the modulation period.

[0117] The series of processes described above—beam injection, acceleration, extraction, and beam discarding—are shown as a timing chart in Figure 14.

[0118] As shown in Figure 14, at a certain time t １ If the irradiation is performed with a first-energy beam up to this point, and then with a second-energy beam thereafter, then the application of a high-frequency electric field in the acceleration gap 223, or the application of a disturbance-type high-frequency electric field, shall be performed at the timing when the irradiation of the first-energy beam is completed.

[0119] Subsequently, after the extraction and disposal of the remaining first-energy beam within accelerator 1 is completed, control of the acceleration of the second-energy beam is initiated. For example, as control of the acceleration of the second-energy beam, it is desirable to start applying an accelerating electric field (high-frequency electric field in the acceleration gap 223) to accelerate the circulating beam.

[0120] Furthermore, without predicting or measuring the remaining amount of the first energy beam, discard control may be promptly executed when it is determined that an energy switch has occurred.

[0121] Figure 15 shows a control block diagram for realizing the above processing. As shown in Figure 15, a high-frequency power supply 46 is connected to the disturbance high-frequency electrode 313. The high-frequency power supply 46 is controlled by a disturbance high-frequency control device 47. High-frequency power, amplified from the high frequency generated by the low-level high-frequency generator 42, is input to the high-frequency accelerating cavity 21 via a high-frequency amplifier 43. A motor control device 41 is connected to the servo motor 214 and controls it to rotate at a constant angular velocity.

[0122] These disturbance high-frequency control devices 47, low-level high-frequency generators 42, and motor control devices 41 control their respective targets based on commands from the overall control device 40. The overall control device 40 controls each device based on the treatment plan obtained from the treatment plan database 60. The overall control device 40 also determines the target energy to be accelerated based on the treatment plan and transmits predetermined command values ​​to each device.

[0123] Next, Figure 16 shows the process from the start to the end of irradiation as a flowchart.

[0124] As shown in Figure 16, after irradiation begins, the overall control device 40 sets the acceleration energy (S101) and the spot position (S102).

[0125] Next, the incident beam from the ion source 12 and the acceleration of the beam in the beam passage region 20 are controlled (S103). Once the acceleration to a predetermined energy is complete, i.e., the irradiation preparation is complete, the beam irradiation is controlled (S104).

[0126] When beam irradiation begins, the irradiation control device 50 receives a signal from the dose measuring device (not shown) in the irradiation device 3 and constantly monitors the irradiation dose and irradiation position. Once irradiation of a predetermined spot is complete, the device moves to the next spot and irradiates it (Yes in S105, S106).

[0127] In this embodiment, once all spots in a certain layer have been irradiated (Yes in S107), the overall control device 40 checks whether there is a plan for irradiation in the next layer, i.e., irradiation at a different energy (S108). If it is determined that there is a plan for irradiation in the next layer, i.e., irradiation at a different energy (No in S108), the overall control device 40 makes the above-mentioned beam discarding means (S109), and controls the symmetrical equipment based on the result of that determination to discard the beam. Then, the process returns to step S101 and new setting values ​​are sent to each control device in order to irradiate the next layer.

[0128] The irradiation is completed by performing each of the above steps until irradiation is finished on all layers (Yes in S108).

[0129] Furthermore, the irradiation method is not limited to spot scanning irradiation; it can also be suitably applied to raster scanning irradiation, which continues irradiation without stopping the beam when moving between spots, and line scanning irradiation, which keeps the beam moving while irradiating without stopping.

[0130] Next, the effects of this embodiment will be described.

[0131] The accelerator 1 of this embodiment described above comprises an ion source 12 for generating ions, a pair of magnetic poles 123 for exciting a magnetic field between them, and a D-electrode 221 for applying a high-frequency electric field. The accelerator 1 accelerates ions orbiting in the magnetic field by applying a high-frequency electric field with the D-electrode 221, and converges a plurality of annular beam orbits formed by the pair of magnetic poles 123, each orbiting ions of different energies, on one side, and modulates the frequency of the high-frequency electric field with respect to the beam orbits. The magnetic field is modulated along the beam orbits such that the second-order Fourier component of the magnetic field gradient along the beam orbit formed by the magnetic field becomes smaller than the average of the magnetic field gradient.

[0132] As described above, our investigations have revealed that in a conventional accelerator of the type that aggregates multiple annular beam orbits, each formed by a pair of magnetic poles 123 and carrying ions of different energies, the displacement of the beam orbit is sensitive to error magnetic fields.

[0133] In contrast, the configuration and operating procedure of accelerator 1 described above, by possessing the characteristics mentioned above, allows for the modulation of the magnetic field in orbit. As the magnetic field in orbit oscillates, the β function of the beam is flattened, and the beam's sensitivity to positional errors in the magnetic field becomes less sensitive than before. This allows for the acceleration of a larger beam quantity in a single acceleration process than in conventional examples, and increases the number of spots that can be irradiated in a single acceleration process. Consequently, the number of acceleration and deceleration operations performed during layer irradiation is reduced, and the irradiation time can be shortened. Therefore, the irradiation time can also be shortened.

[0134] From a different perspective, if the magnetic field can be adjusted with the same level of precision as before, the extra margin can be used to suppress the current drawn from the ion source 12, rather than to increase the amount of charge that can be accelerated. In this case, the lifespan of the ion source 12 can be extended, and by eliminating the need for maintenance-related shutdowns and ensuring stable operation, it is possible to improve the system's uptime.

[0135] Furthermore, the magnetic field excited by magnetic pole 123 has the largest amplitude among the first and higher Fourier components of the magnetic field gradient along the beam orbit, thus more reliably achieving smoothing of the beam's β function. Therefore, the effects described above can be obtained more reliably.

[0136] Furthermore, the average magnitude of the magnetic field within ±π / 8 of the position where the beam orbits with slightly different energies are closest is greater than the average magnitude of the magnetic field for the entire orbit. Conversely, the average magnitude of the magnetic field within ±π / 8 of the position deflected π / 2 upstream and π / 2 downstream from the position where the beam orbits are closest is smaller than the average magnitude of the magnetic field for the entire orbit. Conversely, the average magnitude of the magnetic field within ±π / 8 of the position deflected π upstream and π downstream from the position where the beam orbits are closest is greater than the average magnitude of the magnetic field for the entire orbit. This increases the number of magnetic field adjustment patterns, thereby improving design flexibility.

[0137] Furthermore, because the average size of the magnetic pole 123 gap formed by the magnetic pole 123 in a range of ±π / 8 from the position deflected π / 2 upstream and π / 2 downstream from the position where the beam orbit is closest is larger than the average size of the magnetic pole 123 gap for the entire orbit, and the average size of the magnetic pole 123 gap formed by the magnetic pole 123 in a range of ±π / 8 from the position deflected π upstream and π downstream from the position where the beam orbit is closest is smaller than the average size of the magnetic pole 123 gap for the entire orbit, the above-mentioned smoothing of the beam's β function can be achieved by adjusting the structure, such as adjusting the shape of the magnetic pole 123, thus eliminating the need for adjustment during operation.

[0138] Furthermore, by providing coils 141 and 142 within the gap between the pair of magnetic poles 123, with current paths arranged to enclose at least two ranges from the following locations: a range of ±π / 8 from the position where the beam orbit is closest, a range of ±π / 8 from a position deflected π / 2 upstream and π / 2 downstream from the position where the beam orbit is closest, and a range of ±π / 8 from positions deflected π upstream and downstream from the position where the beam orbit is closest, the smoothing of the beam's β function can be achieved without adjusting the shape of the magnetic poles 123 or other structural adjustments, making fine-tuning extremely easy.

[0139] <Other> The present invention is not limited to the embodiments described above, and various modifications and applications are possible. The embodiments described above are explained in detail for the purpose of clearly illustrating the present invention, and are not necessarily limited to those having all the configurations described.

[0140] 1...Accelerator 2...Beam transport device 3...Irradiation device 4...Treatment table 5...Patient 11...Electromagnet 12...Ion source 13...Coil 20...Beam passage region 21...High-frequency accelerating cavity 40...Overall control device 41...Motor control device 42...Low-level high-frequency generator 43...High-frequency amplifier 46...High-frequency power supply 47...Disturbance high-frequency control device 50...Irradiation control device 60...Treatment plan database 70...Treatment planning device 111...Penetration port for extraction beam 112,113...Penetration ports for coil connection 114...Penetration port for high-frequency power input 115...Penetration port for beam injection 121...Return yoke 122...Top plate 123...Magnetic pole 130...Injection section 140...Magnetic member 141,142...Coil 211...Input coupler 212...Rotating variable capacitance capacitor 213...Rotating shaft 214...Servo motor 221...D electrode 222...Ground electrode 223...Acceleration gap 311...Shim for generating additional magnetic field 312...Septum electromagnet for extraction 313...High-frequency electrode for disturbance 322...Extraction orbit 1000...Particle beam therapy system

Claims

1. An accelerator comprising an ion source for generating ions, a pair of magnetic poles for exciting a magnetic field between them, and a high-frequency electrode for applying a high-frequency electric field, wherein the high-frequency electric field is applied to ions orbiting in the magnetic field using the high-frequency electrode to accelerate them, and a plurality of annular beam orbits formed by the pair of magnetic poles, each orbiting ions of different energies, are brought together on one side, and the frequency of the high-frequency electric field is modulated by the beam orbits, wherein the magnetic field is modulated along the beam orbits such that the second-order Fourier component of the magnetic field gradient along the beam orbits formed by the magnetic field is smaller than the average of the magnetic field gradient.

2. The accelerator according to claim 1, wherein the magnetic field excited by the magnetic pole is such that the second-order Fourier component of the magnetic field gradient along the beam orbit has the largest amplitude among the first-order and higher-order Fourier components.

3. An accelerator according to claim 1, wherein the average magnitude of the magnetic field in a range of ±π / 8 from the position where the beam orbits with infinitesimally different energies are closest is greater than the average magnitude of the magnetic field of the entire orbit, the average magnitude of the magnetic field in a range of ±π / 8 from a position deflected π / 2 upstream and π / 2 downstream from the position where the beam orbits are closest is smaller than the average magnitude of the magnetic field of the entire orbit, and the average magnitude of the magnetic field in a range of ±π / 8 from a position deflected π upstream and π downstream from the position where the beam orbits are closest is greater than the average magnitude of the magnetic field of the entire orbit.

4. An accelerator according to any one of claims 1 to 3, wherein the average value of the magnetic pole gap size formed by the magnetic poles in a range of ±π / 8 from a position deflected π / 2 upstream and π / 2 downstream from the position where the beam orbit is closest is greater than the average value of the magnetic pole gap size of the entire orbit, and the average value of the magnetic pole gap size formed by the magnetic poles in a range of ±π / 8 from a position deflected π upstream and π downstream from the position where the beam orbit is closest is smaller than the average value of the magnetic pole gap size of the entire orbit.

5. An accelerator according to any one of claims 1 to 3, wherein a coil having a current path is provided within the gap between a pair of magnetic poles, the coil having a current path arranged to enclose at least two of the following ranges: a range of ±π / 8 from the position where the beam orbit is closest; a range of ±π / 8 from a position deflected π / 2 upstream and π / 2 downstream from the position where the beam orbit is closest; and a range of ±π / 8 from positions deflected π upstream and downstream from the position where the beam orbit is closest.

6. A particle beam therapy system comprising an accelerator according to any one of claims 1 to 3, and an irradiation device for emitting a beam extracted from the accelerator.

Images