# US 2018/0021594 A1 · Volumetric modulated arc therapy (vmat) with non-coplanar trajectories

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2018/0021594 A1
- Document ID: US20180021594A1
- Status: Published application
- Kind: A1
- Application number: 15/540,051
- Filed: 2016-01-20
- Published: 2018-01-25
- Earliest claimed: 2015-01-22
- Assignees: Koninklijke Philips N.V.; Massachusetts General Hospital
- Applicants: Koninklijke Philips N.V; Massachusetts General Hospital
- Inventors: DADIV PAPP; JAN UNKELBACH; THOMAS BORTFELD; Mattieu Frédéric BAL
- Web page: https://patentgrep.com/patent/US20180021594A1

## CPC

- A61N 5/103: {Treatment planning systems}
- A61N 5/1047: {with movement of the radiation head during application of radiation, e.g. for intensity modulated arc therapy or IMAT}
- A61N 5/1081: {Rotating beam systems with a specific mechanical construction, e.g. gantries}
- A61N 2005/1041: {using a library of previously administered radiation treatment applied to other patients}
- A61N: ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY (measurement of bioelectric currents A61B; surgical instruments, devices or methods for transferring non-mechanical forms of energy to or from the body A61B18/00; anaesthetic apparatus in general A61M; incandescent lamps H01K; infrared radiators for heating H05B)

## Abstract

A method includes determining a set of candidate beam directions. The radiation therapy method further includes selecting a sub-set of non-coplanar beam directions of interest from the set of candidate beam directions based on a fluence optimization using a beam angle selection algorithm. The radiation therapy method further includes determining a set of delivery options based on a beam trajectory algorithm, wherein the delivery options at least include a non-coplanar trajectory during radiation treatment delivery. The radiation therapy method further includes optimizing the delivery options to generate a VMAT radiation plan with non-coplanar beam trajectories. The optimizing of the delivery options includes optimizing at least one machine parameter.

## Claims

### Claim 1

A method, the comprising:
determining a set of candidate beam directions;
selecting a sub-set of non-coplanar beam directions of interest from the set of candidate beam directions based on a fluence optimization using a beam angle selection algorithm;
determining a set of delivery options based on a beam trajectory algorithm, wherein the delivery options at least include a non-coplanar trajectory during radiation treatment delivery; and
optimizing the delivery options to generate a Volumetric Modulated Arc Therapy (VMAT) radiation plan with non-coplanar beam trajectories, wherein the optimizing of the delivery options includes optimizing at least one machine parameter
wherein the fluence optimization is based on: minimize F(d), where d represents a dose distribution and F is an objective function quantifying deviation of the dose distribution d from treatment goals, wherein the dose distribution d is determined by: d=Σb∈C⊂BDb(xb−xrefb)+M(xref), where B represents the set of candidate beam directions, C represents the sub-set, b represents a beam index, xb represents a vector of beamlet intensities for the beam b, Db represents a corresponding dose-fluence matrix, M represents a dose computation method, for instance collapsed cone convolution or Monte Carlo based method approach, and xref represents a set of reference fluence intensities, which are indexed with b,
wherein the treatment plan consists of a beam set Cn, which includes n beams, where n is positive an integer, and further comprising:
adding, separately, each remaining candidate beam b⊂B/Cn to the beam ensemble Cn;
solving minimize F(d);
identifying a candidate beam that yields a lowest objective function value; and
adding the candidate beam that yields the lowest objective function value the beam ensembles,
wherein minimize F(d) is only approximated by performing a predetermined first number of iterations of a gradient descent algorithm, wherein the objective function value reached after the predetermined first number of iteration represents a score for a quality of a candidate beam, or
wherein the method further comprises:
selection a beam angle based on only a norm of a gradient of the objective function in aa iteration to obtain a beam score;
adding the beam to the beam ensemble Cn; and
fixing fluence maps of all beams of the beam ensemble Cn at their optimal values while a new candidate beam b has zero fluence.

### Claim 2 (depends on 1)

The method of claim 1, wherein the at least one machine parameter includes multi-leaf collimator leaf positions.

### Claim 3 (depends on 1)

The method of claim 1, further comprising:
transmitting the plan to a controller that controls a radiation treatment system to continuously deliver radiation based on non-coplanar trajectories while rotating the beam and the subject support based on the signal.

### Claim 4 (depends on 1)

The method of claim 1, wherein the set of candidate beam directions is determined taking into account predetermined criteria, wherein the predetermined criteria includes at least one or more of avoiding a collision of a rotating gantry of the treatment system with the subject support of the treatment system or avoiding a collision of the rotating gantry with a subject supported by the subject support.

### Claim 5

(canceled)

### Claim 6 (depends on 1)

The method of claim 1, further comprising: iteratively selecting beam angles by identifying and adding beam to the treatment plan successively.

### Claim 7

(canceled)

### Claim 8

(canceled)

### Claim 9

(canceled)

### Claim 10 (depends on 1)

The method of claim 1, further comprising:
calculating the norm by:

-
∇
F

(
d
)
+

=
∑
j


[
max

(
0
,
-
∂
F
∂
x
j
b
)
]
2
,
where −∇F(d)+ is a projected negative gradient of the objective function, which is a steepest feasible descent direction at the current solution; and
adding the candidate beam with a largest value of the projected gradient norm to the beam ensemble Cn.

### Claim 11 (depends on 10)

The method of claim 10, further comprising:
finding a fastest beam trajectory that passes through prescribed beam positions by defining a least amount of time needed to reposition the rotating gantry and the subject support from one position to the other, wherein the beam trajectory of interest is a fastest path that visits each beam position.

### Claim 12 (depends on 11)

The method of claim 11, further comprising:
determining a delivery time of every trajectory in seconds;
determining a travel time between two positions (c1, g1) and (c2, g2)
finding a beam trajectory of a predetermined maximum time by identifying beam positions one by one; and
adding the beam to the ensemble;
determining a shortest trajectory visiting each of these angles;
finding a final beam trajectory using anchor points and a sequence in which they are visited; and
determining the plan using dose-influence matrices along the beam trajectory for a predetermined resolution.

### Claim 13 (depends on 12)

The method of claim 12, further comprising:
finalizing the plan optimizer by adapting a VMAT algorithm previously developed for coplanar VMAT.

### Claim 14 (depends on 13)

The method of claim 13, further comprising:
optimizing the plan using a sliding window VMAT optimization algorithm by dividing the arc into K arc sectors, and in each arc sector, the collimator leaves move unidirectionally across the field, delivering an intensity-modulated field.

### Claim 15

A radiation treatment system, comprising:
a candidate beam direction determiner configured to determine a set of candidate beam directions;
a beam direction selector configured to select a sub-set of non-coplanar beam directions from the set of candidate beam directions based on a fluence optimization using a beam angle selection algorithm;
a beam trajectory determiner configured to determine a set of delivery options that at least include a delivery that involves both beam rotation and subject support rotation during volumetric arc delivery; and
a plan optimizer configured to optimize the delivery options, which generates a VMAT radiation plan with non-coplanar beam directions
wherein the fluence optimization is based on: minimize F(d), where d represents a dose distribution and F is objective function quantifying deviation of the dose distribution d from treatment goals, wherein the dose distribution d is determined by: d=Σb∈C⊂BDbxb, where B represents the set of candidate beam directions, C represents the sub-set, b represents a beam index, xb represents a vector of beamlet intensities for the beam b, and Db represents a corresponding dose-fluence matrix,
wherein the beam direction selector approximates minimize F(d) by performing a predetermined first number of iterations of a gradient descent algorithm, wherein the objective function value reached after the predetermined first number of iterations represents a score for a quality of a candidate beam, or
wherein the beam direction selector selects the beams angle based on a norm of a gradient of the objective function in an iteration to obtain a beam score, add the beam to the beam ensemble Cn, fix the fluence maps of all beams the beam ensemble Cn at their optimal values while a new candidate beam b has zero fluence.

### Claim 16

(canceled)

### Claim 17

(canceled)

### Claim 18

(canceled)

### Claim 19 (depends on 15)

The radiation treatment system of claim 15, wherein the beam trajectory determiner finds a shortest beam trajectory that passes through prescribed beam angles by defining a distance as a least amount of time needed to reposition the rotating gantry and the subject support from one angle to the other, wherein the beam trajectory of interest is a shortest path that visits each beam angle, without returning to the first beam angle.

### Claim 20 (depends on 15)

The radiation treatment system of claim 15, wherein the beam plan optimizer optimizes the plan using a sliding window VMAT optimization algorithm by dividing the arc into K arc sectors, and in each arc sector, the collimator leaves move unidirectionally across the field, delivering an intensity-modulated field.

### Claim 21

A non-transitory computer readable medium encoded with computer executable instructions, which when executed by a processor, causes the processor to:
determine a set of candidate beam directions;
select a sub-set of non-coplanar beams from the set of candidate beam directions based on a fluence optimization using a beam angle selection algorithm;
determine a set of delivery options based on a beam trajectory algorithm, wherein the delivery options at least include: a delivery that involves both beam rotation and a subject support rotations during volumetric arc delivery;
optimize the delivery options, which generates a VMAT radiation plan with non-coplanar beam directions; and
control a radiation treatment system in VMAT mode to continuously deliver radiation based on non-coplanar trajectories while rotating the beam and rotating the subject support based on the signal,
wherein the fluence optimization is based on: minimize F(d), where d represents a dose distribution and F is an objective function quantifying deviation of the dose distribution d from treatment goals, wherein the dose distribution d is determined by: d=Σb∈C⊂BDbxb, where B represents the set of candidate beam directions, C represents the sub-set, b represents a beam index, xb represents a vector of beamlet intensities for the beam b, and Db represents a corresponding dose-fluence matrix,
wherein minimize F(d) is approximated by performing a predetermined first number of iterations of a gradient descent algorithm, wherein the objective function value reached after the predetermined first number of iterations represents a score for a quality of a candidate beam, or
wherein the beam angle is selected based on a norm of a gradient of the objective function in an iteration to obtain a beam score, the beam is added to the beam ensemble Cn, the fluence maps of all beams of the beam ensemble Cn are fixed at their optimal values while a new candidate beam b has zero fluence.
