A Graph-Based Method for PET Image Segmentation in Radiotherapy Planning: A Pilot Study

Abstract

Abstract. Target volume delineation of Positron Emission Tomography (PET) images in radiation treatment planning is challenging because of the low spatial resolution and high noise level in PET data. The aim of this work is the development of an accurate and fast method for semi-automatic segmentation of metabolic regions on PET images. For this purpose, an algorithm for the biological tumor volume delineation based on random walks on graphs has been used. Validation was first performed on phantoms containing spheres and irregular inserts of different and known volumes, then tumors from a patient with head and neck cancer were segmented to discuss the clinical applicability of this algorithm. Experimental results show that the segmentation algorithm is accurate and fast and meets the physician requirements in a radiotherapy environment.

Gamma Knife treatment planning: MR brain tumor segmentation and volume measurement based on unsupervised Fuzzy C-Means clustering

“…Therefore, further studies are planned to use a multimodal approach to the target volume contouring, which integrates MRI with MET‐PET ([ 11 C]‐labeled Methionine ‐ Positron Emission Tomography) imaging (Levivier et al, ). In this way, MR images will highlight the morphology of the lesion volume (GTV), while PET images will provide information on metabolically active areas (Biological Target Volume, BTV; Stefano et al, ).…”

“…Therefore, further studies are planned to use a multimodal approach to the target volume contouring, which integrates MRI with MET-PET ([ 11 C]-labeled Methionine -Positron Emission Tomography) imaging (Levivier et al, 2002). In this way, MR images will highlight the morphology of the lesion volume (GTV), while PET images will provide information on metabolically active areas (Biological Target Volume, BTV; Stefano et al, 2013). The proposed approach for lesion segmentation from brain MRI, properly integrated with PET segmentation, would provide a comprehensive support for the accurate identification of the region to be treated with the high precision radiation therapy approach.…”

Correlation between histological grade and positron emission tomography parameters in cervical carcinoma

“…Volumetric and semi-quantitative parameters were extracted using a software package that has been developed and implemented ‘ ad hoc ’ by our group to assess treatment response in oncological patients: PET parameters were measured on the lesions that had previously been identified as suspicious by visual analysis using our automatic segmentation algorithm based on random walks on graphs ( 37 ). The nuclear medicine physician interactively selected the PET region with uptake higher than the background, excluding any physiological uptake findings, by clicking on the image.…”

A fully automatic method for biological target volume segmentation of brain metastases

“…Despite this issue, the RW algorithm is capable of performing the segmentation. For this reason, starting from our previous study (Stefano et al, ), parameters in (1) have been modulated to incorporate metabolic information in the original RW algorithm: g i and g j have been replaced with the standardized uptake value (SUV) in the voxels i and j , where …”

“…So, RW is an efficient and accurate method in low contrast images characterized by noise and weak edges such as PET images (Boellaard et al, 2004;Zaidi and El Naqa, 2010): the behavior of the RW algorithm in the presence of these issues distinguishes it from other approaches (Grady, 2006). Even if the study proposed by Rzeszutek et al (2009) affirms that a drawback to RW algorithm is that it has difficulty producing clean segmentations in the presence of noise, in previous PET studies (Bagci et al, 2011;Stefano et al, 2013;Onoma et al, 2014) the RW method showed accurate results in the delineation of metabolic targets. For these reasons, we propose a fully automatic and operator independent method for the BTV delineation of brain metastases for Gamma Knife treatments based on an extension of the RW algorithm.…”