A GA-based optimization method on target temperature control in RF-capacitive hyperthermia

X Zhu,BK Wan,XM Cheng,LX Zhang,SY Lin,W Wang
2000-01-01
Abstract:RF-capacitive Hyperthermia is a very effective tumor therapy when applied combinedly with X-ray and Chemical ones,especially for some shallow-seated and deep-seated tumor tissues insensitive to the latter two. As for the lower blood perfusion rate and heat sensitivity of tumor tissues,they can be fatally killed when heated in a temperature ranging from 43 degreesC to 45 degreesC for a lasting time. Howerver, the therapy plan is strenuous to design and usually depends heavily on the experiences of the medical personnel because of the rather hard controllability of SAR distribution of RF-capative electro-magnetic fields. To overcome such a problem, an inverse optimization method, applied in Hyperthermia to search the ideal heating physical configurations in the expected temperature distribution of heat-field, was proposed. First, in order to realize the target temperature distribution of tumor tissues and avoid the over-heating of normal tissues, an objective function and some related weight coefficients were introduced. Later, based on the 2-D finite element method solution of Laplace and Pennes' bio-heat transfer equations,the genetic algorithm recursively modified the heating conditions to minimize the objective function,that is,the difference between the calculated temperature distribution and the expected one. The heating configurations are defined as the sizes, voltages, and locations of the two electrods; and they were discreted and coded to binary strings or chromosomes (so called in genetic algorithms) to represent solutions. A satisfying case simulation was demonstrated in 900s 2-Capacitance RF-Capcitive Hyperthermia of a simplified non-even human model obtained from an X-CT human slice containing both shallow-seated and deep-seated tumors. The spatial mean and deviation of the temperatures were 43.0+/-1.9 degreesC for shallow-tumors and 42.9+/-1.9 degreesC for deep-seated tumors and 38.2+/-1.4 degreesC for normal tissues. This method of target heat-field temperature control is believed to have an important practical value in the designs of hyperthermia devices and clinical treatment plans.
What problem does this paper attempt to address?