Automatic segmentation of pulmonary nodules on CT images by use of NCI lung image database consortium

Segmentation of lung lesions on CT scans using watershed, active contours, and Markov random field

“…Tachibana and Kido developed a method based on threshold, watershed, and distance transformation. 32 They trained and tested their algorithm with the first LIDC dataset, and achieved a mean overlap ratio of 51%. Wang et al proposed a method based on spiral scanning and dynamic programming.…”

“…A voxel was considered to be inside the lesion if its probability was 50% or greater (with at least half of radiologists' consensus). 14,16,32 The diameter of the lesions ranged from 4.0 to 33.6 mm, 14 and volumes ranged from 32 to 19874 mm 3 (median 127 mm 3 ) based on the 50% or greater probability map. 3.…”

Segmentation of pulmonary nodules of various densities with morphological approaches and convexity models

“…In Wang et al (2007), the authors reported the average s O of 66%, which is comparable to our study. In Tachibana and Kido (2006), the average s O was reported as 51%. In Way et al (2006), the authors reported average s O with various Gs generated by applying different thresholds to the probability map.…”

“…LIDC1 data set was also used in Wang et al (2007), Tachibana and Kido (2006), Way et al (2006) for performance evaluation. In Wang et al (2007), the authors reported the average s O of 66%, which is comparable to our study.…”

Automatic detection of GGO regions on CT images in LIDC dataset based on statistical features

“…However, most of them focuses on solid lung nodules only and few works deal with the detection of GGO [5][6][7].…”