Semi-automatic Volumetric Measurement of Lung Cancer Using Multi-detector CT

Comparison of three software systems for semi-automatic volumetry of pulmonary nodules on baseline and follow-up CT examinations

“…background voxels (17). Since attenuation information is not sufficient to distinguish nodule boundaries from attached vessels, most semi-automated methods incorporate morphologic operators (18).…”

Current Readings: Radiologic Interpretation of the Part-Solid Nodule: Clinical Relevance and Novel Technologies

“…Nearly all of the prior studies pertaining to the size of subsolid nodules entail primarily linear manual measurements and, more recently, manually obtained 2D area and 3D volume techniques. 26,54,55 As an example, de Hoop et al report greater measurement agreement with nodule mass than for nodule volume, both deriving from manual segmentation of the entire nodule on all CT sections. 27 For subsolid nodules, 3D volumetric measurements, in contrast to single dimensions, are promising as methods to evaluate amorphously shaped subsolid nodules.…”

“…Therefore, computer-assisted techniques directed toward "segmenting" or separating nodules from the surrounding lung were initially developed for measuring small solid nodules and more recently applied to subsolid lesions as a way to facilitate measurement and reduce variability. [54][55][56][57] Methods are currently "semiautomated," in which the interpreter indicates a nodule with subsequent quantitative analysis by the computer program, and for subsolid nodules, such tools are primarily investigational. Technical factors such as varying lung inspiratory volume, which influences the contrast between the nodule and lung parenchyma; CT image thickness; and reconstruction algorithm affect measurement, whether manual or automated.…”

Exploring new quantitative CT image features to improve assessment of lung cancer prognosis

“…First, as shown in Figure 1(a), applying an initial region growing algorithm, the segmented tumor has a leakage to outside lung area (as shown in Figure 1(b)). To segment this Juxtapleural tumor, our CAD scheme applies a modified convex hull function based algorithm, which was initially introduced in the previous work done by Kuhgnik et al [20], to stop leakage of the segmented tumor area to the outside of normal lung structure and smooth the tumor boundary contour. Specifically, based on an anatomical fact that lungs are mostly convex, this algorithm is an efficient and adaptive tumor segmentation method, which has good capability of removing the thoracic lesions from the chest wall (as shown in example of Figure 1(c)).…”