MAP-MRF segmentation of lung tumours in PET/CT images

Automatic lung tumor segmentation with leaks removal in follow-up CT studies

“…Other methods rely on the PET scan, in which automatic tumor detection, e.g., by thresholding, is much simpler. For PET/CT scans, Gribben et al [14] propose to use the PET scan for tumor detection, followed by unsupervised Maximum A Posterior Markov Random Field on the registered CT scan values. Kanakatte et al [16] also use the PET scan for tumor detection, but combine thresholding and components analysis to produce the final segmentation.…”

3D automatic lung segmentation in low-dose CT

“…The contrast between lungs and the surrounding healthy tissue in CT scans is evident and this fact explains why a significant number of segmentation approaches incorporate thresholding 2 nd Portuguese Meeting in Bioengineering, February 2012 Portuguese chapter of IEEE EMBS Rectory of the University of Coimbra operations such as [4]- [6], among many others. In addition to thresholding approaches, other types of different segmentation methods can be defined [7], such as: region growing approaches [3] which extract regions according to predefined criteria, classifiers [8], clustering approaches, [9] Markov random field models [10], artificial neural networks [11], [12], deformable models [13] and atlas guided approaches [14]. In this paper we propose a method for the delineation of the lungs based on an automatic threshold technique.…”

Lung tumor delineation in PET-CT images using a downhill region growing and a Gaussian mixture model

“…These challenges have led various investigators [3][4][5][6][7] to improve the accuracy of tumor segmentation. One approach used the anatomical information to extract the lung region to improve lung tumor segmentation in PET [3][4] by focusing only on the lung region.…”

“…Gribben et al [6] recently outlined the role of the widely known segmentation technique, maximum a posterioriMarkov random field (MAP-MRF), in PET-CT lung tumor segmentation. This technique outperformed the thresholding approach but the implementation assumed that the tumors in PET-CT images have been isolated within a box.…”