Synthetic image modalities generated from matched CT and MRI data: A new approach for using MRI in brachytherapy

Abstract

Target definition and dosimetric evaluation of brachytherapy procedures is crucial in developing proper technique and has prognostic implications. Accurate definition of tumour volume and organs at risk is essential for treatment planing. CT accurately localise sources, but often poorly delineates tumour volume. Magnetic resonance imaging (MRI) has introduced several imaging benefits such as improved soft tissue definition and unrestricted multiplanar and volumetric imaging. However, MRI has not yet seriously challenged CT for brachytherapy planing in most sites. One main reasons for this is the poor imaging of the applicator dummy markers required for detecting the source dwell positions. Therefore a new system generating synthetic image modalities from matched and fused CT and MRI-data has been developed. The system presented consists of five modules serving for viewing, matching, segmentation, fusion and validation. With the presented new system manual or automatic image registration and fusion of CT and MR-images can be done, enabling the use of the imaging advantages of MRI and CT in brachytherapy planing. Tumour structures or source dwell positions could be measured in any plane in one imaging modality and then copied into the other. With synthetic image-based brachytherapy planing a optimisation of target volume covering and dose distribution to treatment volume and structures at risk could be achieved. These preliminary data show that image registration and fusion is feasible for brachytherapy planing. The procedure integrating synthetic images from CT and MRI into brachytherapy planing has the advantage of precise target volume definition and of better determination of dose within this target taking into account the critical dose at structures at risk. Synthetic images may lead to an improved tumour control and reduced side effects by brachytherapy.