Can you do biological based optimization?

Equivalent Uniform Dose (EUD) is currently supported in version 4.4 and up of our software. During DAO, the user has the option to set up EUD greater than or less than constraints with their choice of EUD exponents. The optimization accounts for these biological constraints as well as the DVH constraints while performing the optimization process.

How is this different than other inverse planning systems?

Other vendors such as Pinnacle, CMS, Plato, and Corvus use a two-step approach to create IMRT treatment plans. First, they optimize beamlet intensities. Next, they apply a leaf sequencing algorithm that translates the optimized intensity maps into deliverable aperture shapes. Our direct aperture approach represents the next generation of IMRT planning. We have demonstrated that we can produce treatment plans with dramatic improvements in efficiency with equivalent or better dose distributions as compared to those produced by other vendors.

Isn't convolution too slow for pencil beam calculations?

Not really. With our convolution calculations you will eventually save plenty of time in creating the plans. Some other treatment planning systems have used dose calculation algorithms for their pencil that stress speed very much at the expense of accuracy. The problem with that approach is that you spend a lot of time coming up with a plan that you like. You then perform a final dose calculation, and the treatment plan degrades significantly. Then your hard work is lost. We use a very accurate pencil beam dose calculation because it means that the DVH that you see during the optimization will match what you measure at your machine. With our tool, what you see is what you get.

Starting with version 4.4 of our software we have introduced a fast Collapsed Cone Convolution Superposition for final dose calculations. This algorithm can calculate doses for most plans in five minutes while accurately accounting for heterogeneities.

Isn't this IMRT-lite? In other words, aren't you sacrificing your treatment plan quality to get plans using only 3 to 5 aperture

Absolutely not! Our collaborators at the University of Maryland have done comparisons of Direct Aperture Optimization with both Pinnacle and Corvus for a number of different treatment sites. Mathematically, the relationship between the number of intensity levels per beam direction and the number of apertures per beam direction can be expressed as: N=2n -1.

Where n is the number of apertures and is the possible number of intensity levels. It is thus possible to have seven intensity levels when just three apertures are used. With five apertures, the number of possible intensity levels jumps to thirty-one. Sixty-three intensity levels are possible when six apertures are used. Consequently, highly modulated intensity patterns can be produced using a small number of apertures per beam direction. This offers a significant advantage over the traditional two-step process where the number of apertures is typically 2 to 3 times the number of intensity levels. That is, for 15 intensity levels, Direct Aperture Optimization requires 4 apertures as compared to the 30 to 45 apertures required by traditional algorithms.

Tell me about your Brachytherapy product.

Prowess supports Brachytherapy pre-plans for permanent implant of the Prostate and Breast. We also support orthogonal film based planning for Low dose rate temporary implants using line sources and iridium ribbons. Prowess supports CT based post-planning. We have an innovative Mixed Integer Programming based Inverse Brachytherapy for inverse optimization of seed positions for cases such as Prostate and Breast. We intend to support High Dose Rate Brachytherapy in the near future

These results seem suspiciously good. Are your plans rigged to produce results that are better than would typically be expected?

Not at all - comparable results are expected in general. Users are encouraged to obtain a demo and enter their own planning data so they can see these results for themselves.

What are the advantages of Direct Aperture Optimization?

Advantages of DAO

  • More efficient treatment plans. With other systems, you often need half an hour or more to deliver the treatment. With this technique, you can treat your patients in conventional 15 minute time slots. This means that you won't have to extend your treatment day to do IMRT.
  • The plans are more intuitive. With DAO, you don't have to take such a dramatic change of going from 1 shape per direction to 20 or 30. We give you complete control over the complexity of the plan. If you want to start slowly, you could use say 2 apertures per beam direction. As you gain confidence, you can move to 4 or 5 which is more than enough for even the toughest cases.
  • Easier QA. Many of the complaints of physicists with regards to IMRT are related to QA. With DAO, QA is a snap. For one thing, the delivery of the plan should take 10 minutes or less. Secondly, our dose accuracy is very high. This is partly because we do not have small off-axis fields with small numbers of monitor units. This means that you should need only one QA measurement for each patient.
  • Less wear and tear on you LINAC. With IMRT, people often encounter problems with the multileaf collimators due to all of the additional wear and tear. DAO produces such efficient plans that this is reduced dramatically. This means that you should have fewer LINAC problems.

What forms of dose calculations are supported?

For IMRT planning we offer a full 3D collapsed cone convolution superposition dose calculation algorithms whose accuracies has been validated on clinical data. For 3D conformal planning we offer, Fast Photon - which is a TMR and OCR based calculation as well as Clarkson scatter algorithm. Starting version 4.4 we also offer the convolution family of dose calculations for forward planning. These calculations take into account boluses, wedges, blocks and other accessories while performing full 3D dose calculations.

What is Direct Aperture Optimization?

Direct Aperture Optimization is an IMRT planning technique where all of the constraints of the multileaf collimator are included in the optimization. It allows the user to specify how many aperture shapes he or she is willing to deliver from each beam angle. The key feature of Direct Aperture Optimization is that it produces highly efficient IMRT plans.

What is the single most important feature of Prowess?

Simplicity and ease of use is the most important feature of Prowess. With its familiar Windows interface and simple file systems, Prowess reduces the learning curve dramatically and gets you started with the planning process in no time. In spite of its simple look, Prowess has all the features necessary for planning in the clinic.

What kind of seeds/templates do you support for Brachytherapy?

We support all vendors of Seed and Line Sources. The user can obtain the source data from the vendors and input them using our Seed Editor modules. We also support templates from all major manufacturers. Users also have the option to create custom templates using the template editor.

What kind of treatments does Prowess support?

Prowess supports, electron and photon beam planning as well as Brachytherapy planning. Prowess has an innovative algorithm for Intensity Modulated Radiation Therapy (IMRT) planning. We also support mixed photon and electron beam plans as well as mixed source plans for Brachytherapy.

What optimization technique do you use? Conjugate gradient?

The optimization algorithm is simulated annealing.

What type of pencil beam dose calculation do you use?

We use a convolution/superposition algorithm to compute our pencil beam dose distributions.

Who sets up the machines for commissioning?Who sets up the machines for commissioning?

Prowess performs the machine commissioning/tuning for the users. If the user can provide the measured data (PDD/profiles etc), Prowess will enter the data for them, tune the machines so that the calculations match the measured data and create the requires machine for them. The users can then verify and validate the machines before using this in the clinic.

Why simulated annealing? Isn't that slow?

No, Our implementation of simulated annealing can come up with good plans in a matter of minutes. Simulated annealing provides the flexibility for us to include the MLC constraints in the optimization. Simulated annealing also has the advantage that it can avoid local minima in the objective function, and it converges to the globally optimal solution. Through extensive research we have set the parameters for the simulated annealing such that the treatment plans could be optimized in just a few minutes.