Room: Exhibit Hall | Forum 5
Purpose: Surface imaging is commonly used for daily positioning of breast patients and its accuracy is typically validated by weekly portal imaging. However, portal imaging adds additional time to the treatment and delivers extra dose to the patient. It may be desirable to position the patient by surface imaging only if reasonable accuracy can be achieved in clinical practice. With retrospective patient data, this study aims to quantify the frequency and magnitude of portal imaging shifts for patients positioned with daily surface imaging.
Methods: For 149 breast patients (87 deep inspiration breath hold (DIBH) and 62 non-DIBH patients), positioning residuals were extracted from the surface imaging records for all treatment fractions with weekly portal images acquired. Shifts based on weekly portal imaging were retrieved from the patient record. The action level for shifts was 3 mm for non-DIBH and 5 mm for DIBH The residual errors by surface imaging and shifts based on portal imaging were compared. The rate of failure, when the portal imaging indicated a shift greater than the action level but surface imaging did not, was calculated. Statistical distributions of residuals among different sub-groups (+/- portal imaging indicated shift, DIBH vs non-DIBH, etc.) were also compared.
Results: The rate of failure is 4.3% and 2.5% for DIBH and non-DIBH patients respectively, and the difference is not statistically significant (P>0.05). Differences in the distribution of residuals by surface imaging is not statistically significant for patients +/- DIBH but is statistically significant (P<0.05) for patients with/without portal imaging shifts.
Conclusion: Initial results indicate that the positioning of breast patients by surface imaging only has a low failure rate and support a prospective study for further investigation. Patient characteristics and dosimetric effects will be investigated for patients with shifts based on portal imaging but not indicated by surface imaging.
Not Applicable / None Entered.