Discriminant analysis on the treatment results of interstitial radium tongue implants.

Abstract

Discriminant analysis was carried out for 48 tongue cancer patients who were treated with radium single-plane implantation. The 48 patients were grouped into 32 successfully cured without complications, five successfully cured with complications, six successfully cured but requiring additional boost therapy and five with local recurrence. To evaluate the relation between the dose distribution and the local treatment results, the analysis was based on a volume-dose relationship. The functions introduced by this discriminant analysis were linear, and the parameters used were modal dose, average dose and shape factors of histograms. Each group of treatment results had a correction rate of greater than 80%, except for the successfully cured group with ulcers. The discriminant functions were useful as an index to obtain a final clinical treatment result at the early time of implantation, and these functions could be used as a criterion for the optimal treatment of tongue carcinoma. We were also able to recognize the limitation of the actual arrangement of sources in the single-plane implant.