2008-12-07
Complaint-adaptive power density optimization as a tool for HTP-guided steering in deep hyperthermia treatment of pelvic tumors
Publication
Publication
Physics in Medicine and Biology , Volume 53 - Issue 23 p. 6799- 6820
For an efficient clinical use of HTP (hyperthermia treatment planning), optimization methods are needed. In this study, a complaint-adaptive PD (power density) optimization as a tool for HTP-guided steering in deep hyperthermia of pelvic tumors is developed and tested. PD distribution in patients is predicted using FE-models. Two goal functions, Opt1 and Opt2, are applied to optimize PD distributions. Optimization consists of three steps: initial optimization, adaptive optimization after a first complaint and increasing the weight of a region after recurring complaints. Opt1 initially considers only target PD whereas Opt2 also takes into account hot spots. After patient complaints though, both limit PD in a region. Opt1 and Opt2 are evaluated in a phantom test, using patient models and during hyperthermia treatment. The phantom test and a sensitivity study in ten patient models, show that HTP-guided steering is most effective in peripheral complaint regions. Clinical evaluation in two groups of five patients shows that time between complaints is longer using Opt2 (p = 0.007). However, this does not lead to significantly different temperatures (T50s of 40.3 (Opt1) versus 40.1°C (Opt2) (p = 0.898)). HTP-guided steering is feasible in terms of PD reduction in complaint regions and in time consumption. Opt2 is preferable in future use, because of better complaint reduction and control.
Additional Metadata | |
---|---|
doi.org/10.1088/0031-9155/53/23/010, hdl.handle.net/1765/60036 | |
Physics in Medicine and Biology | |
Organisation | Department of Radiation Oncology |
Canters, R., Franckena, M., van der Zee, J., & van Rhoon, G. (2008). Complaint-adaptive power density optimization as a tool for HTP-guided steering in deep hyperthermia treatment of pelvic tumors. Physics in Medicine and Biology, 53(23), 6799–6820. doi:10.1088/0031-9155/53/23/010 |