Room: Exhibit Hall
Purpose: We have found that most of the time The Eclipse EMC predicts MUâ€™s within 5% of the hand calculation. Occasionally, the MUâ€™s may be off by as much as 10% or more. The purpose of this work was to make realistic phantoms to investigate the cause of the discrepancy of the EMC and how to correct for it.
Methods: Custom wax phantoms with anatomical shape were constructed with embedded chambers (parallel plate is preferred for non-water, TG70). These phantoms were CT scanned with and without heterogeneity. Heterogeneity was introduced by either using Play-dohÂ® or CT simulation skin markers/wire material which the physician typically places on breast scars. Electron plans were made using the Varian EMC algorithm with enface and 45Â° gantry rotations at extended SSD (our clinical SSD= 110cm).
Results: Preliminary results show that wax phantoms are easy to construct and can be used to determine which parameters affect the EMC MU prediction. Clinically, the EMC MUâ€™s occasionally exceed our 5% tolerance from the hand calculations, especially with any heterogeneity (bone, wire, BB, etc.) near the CAX within the effective range. To resolve this difference, we deliberately move the table and adjust the cutout opening to cover the same target so that the CAX is away from any wire by at least 1cm. Even though the wires are contoured and subtracted from the body, there may be blooming effects that can only be seen with proper window and leveling. Then we run and save 3 identical plans and take the one with MUâ€™s that matches the hand calculation closest.
Conclusion: If using wires, it is recommended to CT scan the patient twice: with and without wires. Shifting and moving the block may yield improved MUâ€™s. More work is required to determine how to correct for non-flat shapes and heterogeneity.