The subject of this disclosure (and its claims) is a method of multi-objective optimization (MCO).
A cancer patient is to receive radiation therapy (RT). This involves a minimum dose for the tumor tissue and maximum doses for surrounding healthy tissue structures, which should ideally not be exhausted. In principle, there are multiple technical options (technologies) with which such a treatment can be carried out for such a patient.
Examples include the following:
Each of these technologies can be performed on one or more different therapy devices. Each device is addressed differently and requires specific planning, whereby the individual planning objectives remain identical across all of the technologies used.
Relevant from a clinical standpoint is the planning of all treatment cases (overall planning), taking into account all available therapy devices. Besides taking into account the individual medical aspects of a therapy, the overall planning also needs to schedule the treatments.
Mathematically, a multi-criteria planning problem arises with the following objectives: high probability of successful treatment while avoiding side effects with each individual therapy and making the most effective use of the available therapy devices, with which as many patients as possible can receive therapy without having to endure long waiting times (time to treatment).
The state of the art is a two-stage decision-making approach. To begin with, the technology is defined, and then this technology is planned. Multi-criteria planning methods can be used in the second step. Doctors decide on the choice of technology based on experience and the results of clinical studies. In order to decide which technology is best suited to a treatment case, it is necessary to undertake separate planning for each possible technology, compare the respective plans, and then select the “best” option.
The clinical procedure also needs to be considered: In urgent cases, treatments already scheduled on certain devices will have to be postponed or even rescheduled on other devices, or else the patient in urgent need of treatment will have to wait.
Each technology must be implemented on specific, dedicated devices—treatment with protons cannot be performed using a photon accelerator. The number of devices available for treatment is one of the limiting factors in hospitals. Treating more patients—proton systems in particular are only available to a very limited extent due to their cost and complexity-automatically leads to waiting times or rescheduling of therapy sessions. A multi-criteria optimization problem becomes apparent: treating as many patients as possible in a given time-or with a given, usually limited amount of investment capital.
If a hospital now has at least two devices that can be used for treatment, they can now plan, on a purely organizational basis (by scheduling), which (daily) fractions of a fractionated treatment of a patient are to be carried out on which of the radiation devices mentioned in the example. If one of the devices is already occupied by another patient over a period of time, this device cannot be included in the planning for the new patient. The patient must either wait a considerable amount of time or be turned away. Especially when treating malignant tumors, it is particularly undesirable to keep patients waiting or even turn them away with the excuse that the devices are occupied. It is particularly important here to start the treatment planned for the fractionation period as quickly as possible. Treating a malignant tumor, for example, is not something that can be put off for long. Such patients at particularly high risk, with the risk here being the period of time until treatment begins.
Another difficulty is when technical radiation equipment which was manufactured more recently is used in the same way as older technical radiation equipment. It turns out that the state-of-the-art, recently purchased devices are permanently in use or permanently occupied, and very few patients, if any, are willing to be treated on the older devices-if their doctors' orders even allow for this.
The technical problem of the invention lies in the ability to individually provide each patient with good therapy, whereby the quality of the therapy also includes minimizing the waiting time of each patient (under “time to treatment,” a technical criterion of therapy planning).
The invention is based on the realization that a planned total dose of a planned therapy is spread over multiple treatment sessions and that the fractions originate from radiation devices that are not identical. It is assumed that two radiation devices are not the same and therefore that some daily doses are delivered by one radiation device and other daily doses by another radiation device. Taking into consideration the therapy as a whole, this results in a combinational effect through a sum of the doses per day over the duration of therapy. When a “mixture of technologies,” “mixed therapy,” or “combined radiation emission” is referred to, this means that a certain number of (daily) fractions of the planned therapy (with a total number of fractions) are administered on one device and the remaining number of (daily) fractions on another device. However, this is all part of the planning of this therapy, not the therapy itself. References to this always involve the planning.
If two radiation devices—therapy device A and therapy device B—are used, there is a solution in which only therapy device A emits its beams onto the patient, in which case it would be a pure A treatment. There is another version in which only therapy device B emits its beams onto the patient, in which case it would be a pure B treatment.
Any other type of combined radiation emission from both devices to the same patient during the therapy period is combinational irradiation, which lies between “A only” and “B only” in its effect. This involves the planning of these therapies, not their use.
Reference is made to the claims, which are included here for solving the problem mentioned, in particular independent claims 1 and 7.
Embodiment examples of the invention are explained in more detail with the aid of the Figures. All explanations are equally applicable to the disclosure, but they are not to be interpreted in such a way that they must be included as necessary elements of the claims. All of the following examples remain examples even if they are not explicitly preceded by “for example.”
The description deals with the fact that not only one therapy device (as a radiation device) is available, but several of them. Naturally, the therapy devices have an occupancy rate, i.e., their occupancy is scheduled with existing therapies in a time grid, so that a new patient needing therapy has to fit into the existing system of occupancy.
It leads to resentment and even fear if a patient with a critical case of cancer requiring treatment cannot be treated immediately, i.e., no treatment slot is available and no treatment can be planned for them. In other words, no radiation device is available at the moment when the patient announces his or her need for treatment or when it is assigned by the planner.
One way of making this earliest possible treatment available through its underlying planning is to incorporate a different radiation device or a radiation device of another technology into the treatment plan—a radiation device of an older technology is also possible. This results in a mixture of multiple daily doses from different radiation devices.
Here, “mixture” is not intended to mean that the radiation devices are mixed, but rather that the effects of the radiation devices on the patient requiring therapy are mixed (or combined with daily doses over the course of the therapy).
Again, it bears repeating that no therapy is currently being performed on humans (the patient) here, but is only planned. This planning can already be implemented outside the patent claims as therapy with the radiation devices, not with the patent claims. Nevertheless, when describing the planning, it is sometimes necessary to mention radiation therapy per se.
Below, radiation device A and radiation device B—based on the figures, it is their Pareto functions 601 and 701—are to be considered as the at least two radiation devices whose effects in the planning (for the duration of the fractionated therapy) act together on the patient requiring therapy (correct: “to be assumed to act”). Multiple radiation devices of this kind are possible and able to be integrated into the planning.
For this to be possible, the mixture occurs in fractionated daily doses. The period is considered to be the entire duration of treatment, and the smallest unit of the mixture is one day. This division of the units, the total duration and, discretely, the day are to be understood as an example. It is based on the current standard fractionated planning for a patient, where one day has proved to be a good, suitable unit in which a patient is able to bear function, recovers, and is able to function again.
Therapy planning begins on day 1 and ends after the duration of treatment on day X, where X is given as 30 in the examples.
After a day Y<X, the planning switches to another radiation device. We start with the first radiation device.
The premise for the first device should be “start therapy as soon as possible.” The patient requiring treatment therefore has no appreciable waiting time and has the feeling that he or she can be treated right away. In most cases, the technically best possible radiation device or the radiation device with the most state-of-the-art technology is not immediately available, so the waiting time for the patient requiring treatment seems disruptive, unpleasant, or even frightening. In contrast, offering a therapy that starts immediately is preferrable, even if the immediately available device is not state-of-the-art or does not offer the most technically advanced technology, this radiation device should still be scheduled at the beginning. Scheduling could be used as an input value, but since the quality of treatment also plays a role, mere scheduling is not sufficient.
A second premise, although not a mandatory one, is to reduce the number of changes between radiation devices. It will be perceived as unpleasant by the patient if—according to the planning—he or she has to or is supposed to change the radiation device multiple times, i.e., if the planning is such that he or she is made aware of the availability. On the other hand, it is divided into daily doses, and after a day passes, a patient will no longer remember exactly on which radiation machine he or she received the fractionated daily dose the day before.
What is decisive is the dose distribution in the relevant voxels of the tissue over the entire period of the planned therapy, described here as the therapy to be planned, since only this planning is described here and is also claimed, not the radiation therapy itself that is carried out on the patient.
The criteria, ci, according to which a therapy is planned are of a technical nature—for example, selected from the following “mean dose in the heart,” specific clinical objectives, dose in the target, or duration. The technical question that arises is “what quality of treatment is possible” under the premise of “the earliest possible start.” Mathematically, this involves a mixed-integer, non-convex problem that only works with the non-dominating points (a point on the Pareto frontier corresponds to a plan with its technical setting parameters).
The premise of the planned time to treatment (TTT) sets the framework for the mixing ratio for the total duration of therapy planning.
If there are multiple changes between the radiation devices, meaning planned changes, this will have a restrictive effect on the mixing ratio over the total duration.
Quantity F of the efficient mixtures is regarded as being within the tolerable limits. There is both an upper limit and a lower limit defining a minimum radiation dose. To this end, a therapeutic window, labeled ‘f’ in the figures, is opened; see
Operating area 2 and patch area 1 are functionally linked to one another. This functional coupling bears explaining.
Several Pareto frontiers are shown in patch area 1; in the example, these are Pareto frontiers 101 and 401 along with 601, 701, and so on in
Also shown in patch area 1 are two axes c1 and c2 (criterion 1 and criterion 2), which are perpendicular to each other and represent two criteria c1 and c2. Criterion 1 (c1) and criterion 2 (c2) are shown, which are visible and navigable for the planner in operating area 2. The limitation of the illustration to two criteria here is only by way of an example; many more criteria than these will be factored in. Examples of possible technical criteria are listed above. However, since only three criteria can be visibly displayed in a representable Cartesian space, not n criteria where n is greater than 3, we will leave this example as is and merely state that n criteria are factored in, i.e., c1 to cn. The illustration can also be a projection from n-dimensional space.
The following should be noted for an understanding of the Pareto frontier(s). Each Pareto frontier is a function that is only functionally connected. Each of them has a large number of grid points (here, 101; instead of 201; as shown in
The first radiation device 100 of
The first radiation device stands for technology 100 (or technology A). A first support body 114 is rotatably mounted, and a radiation head 110 is arranged on it, which is rigidly connected to support body 114 via a bridge 112. The angle of radiation head 110 relative to patient P can be adjusted via support body 114. The radiation dose and the distribution of the radiation within the proton beam from radiation head 110, which is not shown, can also be adjusted. All these technical values are represented in a Pareto frontier for a patient over the entire course of a fractional radiotherapy (radiation therapy). In the example, this can correspond to Pareto frontier 101 of
It should again be pointed out that it is not the therapy with radiation device 100 (technology 100) that is claimed here, but rather the planning of this therapy. The reference to the therapy devices as radiation devices, which implement this planning later or functionally separately from the planning, serves as an illustration for the patient, but the disclosure is not intended to refer to the patient's therapy. Planning and therapy can be readily separated or differentiated in terms of time and function.
The same applies to the following figures.
In the sequence of
Along line 801 of the Pareto-optimal points, the mixing ratio can be adjusted, i.e., changed in a navigable manner (during planning). Two points are highlighted, which result from the possible mixing ratio of
This is point M2 marked by a square bracket with approx. 67% of Pareto frontier 601, which here stands for radiation device A. In contrast, point M4 marked by an oppositely aligned square bracket has a proportion of approx. 67% of radiation device B. Accordingly, a proportion of approx. 33% remains in the mixture for the respective other radiation device.
As the points to the right of M4 are closer to Pareto frontier 701, the proportion of Pareto frontier 701 is greater here. Point M3 has a 100% share of radiation device B. Point M1 is located on Pareto frontier 601 and therefore has a 100% share of radiation device A. A point corresponds to a therapy plan with its technical settings on the radiation device.
The two points M2 and M4 yielded from the availability of the therapy devices, here with a view to
In a second example in
The switching back of the therapy device may make medical sense, but is not actually desirable as a first choice. It could make sense from a medical standpoint for two reasons:
It should be mentioned here that point M1 depicted is dominated by the Pareto frontier 601 with regard to criterion c2.
Another possibility for the position of the Pareto frontiers of radiation devices A and B, corresponding to 601 and 701, is for there to be a gap 1401 in the therapeutic window of criterion c1, which can be bridged by line 1301 (in the planning) with a mixing ratio of the two therapies that are to be planned.
If synergy effects result from
The navigation environment of
Mouse device M′ is assigned to display 10, e.g., as the trackball, which enables the functions of pointing and operating (triggering a function) at the location where mouse pointer M is displayed.
Mouse pointer M is also used to operate both operating aids 21 and 22 for criteria c1 and c2.
An alternative display 781 is that of a tablet 780, which is given the same representation of
Technological radiation devices 100 to 300′ are each coupled to bus 701 via one input/output device 731, 732, 733 and 734, respectively. Each of these input/output devices is bidirectional, meaning that it can preset parameters from bus 701 to the respective radiation device.
Preferably, each of these radiation devices 100, 200, . . . has sufficient memory of at least 500 GB so that it can store the preset parameters of the subsequent therapy and set them independently, i.e., autonomously, during the therapy, especially for fractionated sessions of the patient during the planning period.
In the example, I/O1 sends the setting parameters to radiation device 100 for the emission of protons. This is then able to apply the parameters of the radiation device for technology 100 “protons” that are to be set—which are actually already prescribed—during the (fractionated) therapy of patient P spread over days. In other words, the doses and intensities and directions of proton radiation (in radiotherapy) to which this individual patient P is to be exposed. This is where the functional separation of planning and therapy occurs.
This example can also be controlled in such a way that the parameters are stored in memory 750 for the duration of the fractionated therapy and are only transmitted to radiation device 100 at the times before the respective therapy session. This fractionated therapy is fractionated programming of respective radiation device 100 to 300′ (akin to the fractionated data transmission of the currently required time section of the therapy, such as 30 days). The actual therapy is performed automatically by radiation device 100, without the previously completed planning.
In the same way, the other devices 200, 300 and 300′ are programmed, prepared for therapy,: and conditioned with regard to data technology.
It bears emphasizing once again that no therapy is performed during the planning stage; the therapy is already fully planned and functionally prepared (i.e., fully planned) before the respective device performs this radiotherapy on the patient. The latter therapy is not claimed.
Navigation via the patches shown on display 10 can only be completed by the planner in order to transfer them later to the associated radiation devices via respective input/output devices 731 to 734.
Number | Date | Country | Kind |
---|---|---|---|
10 2021 124 814.0 | Sep 2021 | DE | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/IB2022/059141 | 9/26/2022 | WO |