Techniques for guaranteeing that intensity patterns are properly positioned relative to a patient’s anatomy have not been developed for rotational IMRT. This paper points out the importance of at least assuring that the "slit" opening covers the treatment volume. Using the NOMOS MIMiC collimator, it is possible, and sometimes necessary, to position the treatment unit isocenter so that the collimator opening does not provide full target coverage as the gantry rotates. This report demonstrates that missing the target compromises the dose distribution by forcing normalization to relatively low isodose lines and increasing the dose to critical healthy structures. A simple check of the NOMOS output can detect potential problems. Caution is advised when the beam elements at the field edge are used in the modulation process. As a more complete check, a double-exposure film technique using the standard jaws is recommended to document field coverage. The technique uses both an anterior and one lateral film to show the extent of the collimator opening. The field size for these films is set to agree with the length of the beam slit (approximately 20 cm) in one direction and the slit width in the other direction. A double-exposure over-flash is used to show this slit relative to the patient’s anatomy. Simulation of these fields is also recommended. Since CT scans must be available for any IMRT treatment, the use of CT-simulation simplifies this part of the process.

Verification of field placement for rotational IMRT

The NOMOS MIMiC intensity modulating device is a tertiary collimator and clearance can be a problem. Installation on a Philips SL 75/5 allows only 29 cm from the front face of the collimator to the treatment unit isocenter. This distance sometimes forces placement of the mechanical isocenter near the surface of the tabletop to avoid couch collision as the gantry rotates. However, the number of leaves limits the length of the slit opening for this device. For the SL 75/5 unit, this length is 18 cm. If a lesion plus margin extends more than 9 cm from the position of the isocenter, it is not be possible to deliver radiation to some parts of the target for one or more portions of the arc. This compromises the optimization and results in reduced dose homogeneity relative to the situation where complete coverage is possible. There are no warnings within the planning software or its documentation to caution the dosimetrist that a sub-optimal plan has been generated. Since verification of IMRT treatments is not standard, this problem can go undetected. Using film dosimetry to verify the calculated treatment plan against measurement will not demonstrate the problem because the two poor plans will agree. For this reason, simple checks are suggested here to detect situations where complete coverage by the collimator is not achieved.

The problem is illustrated for a patient where it was not possible to position the isocenter so that the collimator opening covered the anterior portion of the lesion when the gantry was in the lateral position. Dose-volume histograms (DVH) for the plan where collision was a problem are compared to DVH’s obtained when the isocenter was placed near the center of the target. Due to the possibility of collision, the second plan is achievable with our equipment. The figure above compares the two treatment plans. Both plans cover 90% of the GTV with the boost dose of 24Gy. Notice that the dose inhomogeneity for the plan in the upper part of the figure is much greater than the inhomogeneity for the plan where the field opening fully covers the target. Also notice that the dose for all critical structures is increased for the plan where the field misses the target.

A number of checks can be performed to guarantee target coverage. A simple check is to observe the intensity patterns printed by the NOMOS CORVUS software. If dose delivery extends to the very edge of the collimator opening, some shielding of the target should be suspected. Port filming of the available field opening is also recommended. The filming technique developed here uses a double exposure without the NOMOS collimator in place. The first exposure shows the patient’s anatomy with extensive margin, and the second reduces the collimator size to correspond to the MIMiC opening. A single lateral field film and an anterior film are used to demonstrate acceptable coverage. If the printed intensity patterns show the possibility of missing the target at angles other than 0 and 90 degrees, additional angles should be included in the filming. The figure below shows an example of a lateral port film. Notice that the field opening does not extend to the anterior portion of the patient where there is disease. At our institution, the treatment couch was modified to allow shifting of the isocenter to a greater distance above the tabletop to provide a larger range for positioning the irradiated volume.

It should be pointed out that some clipping of the target may be acceptable in that dose delivered form other directions can possibly make up for projections where the volume is not irradiated. However, it is always prudent to generate a plan with full coverage to determine if the optimization is seriously compromised.

Abstracts Submitted to 1999 Annual AAPM Meeting | Medical Physics Home