Semi-automatic tumour segmentation by selective navigation in a three-parameter volume, obtained by voxel-wise kinetic modelling of C-11-acetate
2010 (English)In: Radiation Protection Dosimetry, ISSN 0144-8420, E-ISSN 1742-3406, Vol. 139, no 1-3, 214-218 p.Article in journal (Refereed) Published
Positron emission tomography (PET) is increasingly used for delineation of tumour tissue in, for example, radiotherapy treatment planning. The most common method used is to outline volumes with a certain per cent uptake over background in a static image. However, PET data can also be collected dynamically and analysed by kinetic models, which potentially represent the underlying biology better. In the present study, a three-parameter kinetic model was used for voxel-wise evaluation of C-11-acetate data of head/neck tumours. These parameters which represent the tumour blood volume, the uptake rate and the clearance rate of the tissue were derived for each voxel using a linear regression method and used for segmentation of active tumour tissue. This feasibility study shows that it is possible to segment images based on derived model parameters. There is, however, room for improvements concerning the PET data acquisition, noise reduction and the kinetic modelling. In conclusion, this early study indicates a strong potential of the method even though no 'true' tumour volume was available for validation.
Place, publisher, year, edition, pages
2010. Vol. 139, no 1-3, 214-218 p.
Medical and Health Sciences Radiology, Nuclear Medicine and Medical Imaging
IdentifiersURN: urn:nbn:se:uu:diva-136430DOI: 10.1093/rpd/ncq075ISI: 000277738200038PubMedID: 20200103OAI: oai:DiVA.org:uu-136430DiVA: diva2:376927