Adam Cunha1, Timmy Siauw2, I-Chow Hsu2, Jean Pouliot2. 1. Radiation Oncology, University of California, San Francisco, CA 94115. Electronic address: adam.cunha@ucsf.edu. 2. Radiation Oncology, University of California, San Francisco, CA 94115.
Abstract
PURPOSE: To present the algorithm of a modification to the inverse planning simulated annealing (IPSA) optimization engine that allows for restriction of the intracatheter dwell time variance. METHODS AND MATERIALS: IPSA was modified to allow user control of dwell time variance within each catheter through a single parameter, the dwell time deviation constraint (DTDC). The minimum DTDC value (DTDC = 0) does not impose any restriction on dwell time variance, and the maximum value (DTDC = 1) restricts all dwell times within each catheter to take on the same value. The final optimization penalty function value was evaluated as a function of DTDC. RESULTS: The algorithm proposed fully preserves the inverse planning nature of the IPSA algorithm along with the penalty-based dose optimization workflow. Increasing DTDC creates less variance in dwell time between dwell positions in each catheter and may be used to induce a more smooth change in dwell time with dwell position in each catheter. Nonzero DTDC values always increased the optimization penalty function value. CONCLUSIONS: The DTDC was developed as an extension to IPSA to allow restriction of the difference in dwell time between adjacent dwell positions. This results in less variation between neighboring dwell positions which can be clinically desirable. However, the impact of this restriction needs to be considered for its clinical relevance on a case-by-case basis because considerable degradation in dose-volume histogram metrics can result for large DTDC values.
PURPOSE: To present the algorithm of a modification to the inverse planning simulated annealing (IPSA) optimization engine that allows for restriction of the intracatheter dwell time variance. METHODS AND MATERIALS: IPSA was modified to allow user control of dwell time variance within each catheter through a single parameter, the dwell time deviation constraint (DTDC). The minimum DTDC value (DTDC = 0) does not impose any restriction on dwell time variance, and the maximum value (DTDC = 1) restricts all dwell times within each catheter to take on the same value. The final optimization penalty function value was evaluated as a function of DTDC. RESULTS: The algorithm proposed fully preserves the inverse planning nature of the IPSA algorithm along with the penalty-based dose optimization workflow. Increasing DTDC creates less variance in dwell time between dwell positions in each catheter and may be used to induce a more smooth change in dwell time with dwell position in each catheter. Nonzero DTDC values always increased the optimization penalty function value. CONCLUSIONS: The DTDC was developed as an extension to IPSA to allow restriction of the difference in dwell time between adjacent dwell positions. This results in less variation between neighboring dwell positions which can be clinically desirable. However, the impact of this restriction needs to be considered for its clinical relevance on a case-by-case basis because considerable degradation in dose-volume histogram metrics can result for large DTDC values.