Program Information
Efficient Method of Metal Artifact Correction for Breast Tissue Expander
J Park1*, A Morrow1 , A McVicker2 , A Altman3 , M Deveau4 , S Kim1 , (1) Division of Radiation Physics, Department of Radiation Oncology, Baylor Scott and White Health, Temple, TX, (2) Department of Radiation Oncology, Baylor Scott and White Health, College Station, TX, (3) Department of Surgery, Baylor Scott and White Health, Temple, TX, (4) Department of Veterinary Radiation Oncology, Texas A&M University, College Station, TX
Presentations
SU-I-GPD-T-412 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose: To correct metal artifacts in the breast CT image caused by the temporary breast implant used for breast cancer patient undergone mastectomy
Methods: The breast implant consists of the silicon bag and the valve with titanium shell and neodymium magnet, which causes significant metal artifacts. The exact shape and dimensions of the components were determined with diagnostic X-ray and CT simulator with metal artifact correction algorithm. Virtual structure model with same dimension and electron density as a real breast implant was created in as a DICOM CT image and RT structure format. Those DICOM files are made on-the-fly per each patient in combination of Eclipse ESAPI script, so the planner just need to import and register it to the planning CT image. Treatment plans with manual artifact correction and model based correction were calculated with Eclipse AAA and AXB algorithm, and dosimetric characteristics were compared.
Results: The planning CT image had an exaggerated shape of artifacts compared to actual implant. With model based correction, metallic components of breast implant and breast tissue around the implant were contoured accurately, and correct HU numbers were assigned. The AAA calculation did not show much differences of dose distribution between manual and model based correction, while the AXB calculation showed hot spots of ~120% around the magnet for the latter. There was also more homogeneous dose distribution of ~105% in the breast in comparison with the cold spots of ~90% inside the artifacts for the manual correction.
Conclusion: New correction method provided accurate and fast way to correct metal artifacts compared to manual correction. The dose distribution after model based correction showed higher hot spots around the magnet and less dose degradation in the breast compared to the manual correction.
Contact Email: