Andrea Massafra

Automatic Lung Segmentation in CT Images with Accurate Handling of the Hilar Region

A fully automated and three-dimensional (3D) segmentation method for the identification of the pulmonary parenchyma in thorax X-ray computed tomography (CT) datasets is proposed. It is meant to be used as pre-processing step in the computer-assisted detection (CAD) system for malignant lung nodule detection that is being developed by the Medical Applications in a Grid Infrastructure Connection (MAGIC-5) Project. In this new approach the segmentation of the external airways (trachea and bronchi), is obtained by 3D region growing with wavefront simulation and suitable stop conditions, thus allowing an accurate handling of the hilar region, notoriously difficult to be segmented. Particular attention was also devoted to checking and solving the problem of the apparent ‘fusion’ between the lungs, caused by partial-volume effects, while 3D morphology operations ensure the accurate inclusion of all the nodules (internal, pleural, and vascular) in the segmented volume. The new algorithm was initially developed and tested on a dataset of 130 CT scans from the Italung-CT trial, and was then applied to the ANODE09-competition images (55 scans) and to the LIDC database (84 scans), giving very satisfactory results. In particular, the lung contour was adequately located in 96% of the CT scans, with incorrect segmentation of the external airways in the remaining cases. Segmentation metrics were calculated that quantitatively express the consistency between automatic and manual segmentations: the mean overlap degree of the segmentation masks is 0.96 ± 0.02, and the mean and the maximum distance between the mask borders (averaged on the whole dataset) are 0.74 ± 0.05 and 4.5 ± 1.5, respectively, which confirms that the automatic segmentations quite correctly reproduce the borders traced by the radiologist. Moreover, no tissue containing internal and pleural nodules was removed in the segmentation process, so that this method proved to be fit for the use in the framework of a CAD system. Finally, in the comparison with a two-dimensional segmentation procedure, inter-slice smoothness was calculated, showing that the masks created by the 3D algorithm are significantly smoother than those calculated by the 2D-only procedure.

Improving the channeler ant model for lung CT analysis

The Channeler Ant Model (CAM) is an algorithm based on virtual ant colonies, conceived for the segmentation of complex structures with different shapes and intensity in a 3D environment. It exploits the natural capabilities of virtual ant colonies to modify the environment and communicate with each other by pheromone deposition. When applied to lung CTs, the CAM can be turned into a Computer Aided Detection (CAD) method for the identification of pulmonary nodules and the support to radiologists in the identification of early-stage pathological objects. The CAM has been validated with the segmentation of 3D artificial objects and it has already been successfully applied to the lung nodules detection in Computed Tomography images within the ANODE09 challenge. The model improvements for the segmentation of nodules attached to the pleura and to the vessel tree are discussed, as well as a method to enhance the detection of low-intensity nodules. The results on five datasets annotated with different criteria show that the analytical modules (i.e. up to the filtering stage) provide a sensitivity in the 80 - 90% range with a number of FP/scan of the order of 20. The classification module, although not yet optimised, keeps the sensitivity in the 70 - 85% range at about 10 FP/scan, in spite of the fact that the annotation criteria for the training and the validation samples are different.

Lung uniformization for juxta-pleural nodule detection

We propose a method for automatic lung juxta-pleural nodule detection in thorax CT images, to be used as a Computer Assisted Detection (CAD) tool by radiologists. It is based on the calculation and automatic analysis of local curvature on the lung surface as extracted from high-resolution CT scans, and exploits uniformization to a sphere (e.g. through conformal mapping) to allow a global view of the lung surface, with marking of high curvature regions which can be suspected of being pleural nodules. Schematically, the tool works as follows. First, lung binary masks are extracted from the image by 3D segmentation of the CT scan. On these masks, pleural nodules appear as small surface concavities of the mask surface. After patching the entrance of vessels into the parenchyma in the hilus pulmonis, the lung frontier Σ is a smooth genus-0 surface. This surface is triangulated and is then uniformized to a sphere Σ′. In this parameterization a suitable function Ψ of the mean and the Gaussian curvatures can be calculated over Σ. Function Ψ is displayed as a colour variation onto both Σ and Σ′, so marking regions that represent high-curvature concavities. A threshold on Ψ is then applied and regions of interest (ROIs), containing little concavities with a low radius of curvature (such as pleural nodules), are detected. ROIs are then examined and classified; techniques such as spherical wavelets are available on the sphere, which will be used to distinguish between false and true positives, helping in diagnosing pleural nodules.

Integrated model for the analysis of two-dimensional electrophore is gel image

Proteomics is the science that studies the proteome, that is the proteic expression of the genome. Cell proteome is extremely complex, and is composed of several thousand proteins. Twodimensional polyacrylamide gel electrophoresis (2DPAGE) is widely used as a standard method to separate and display proteins in a tissue or compound with a theoretical resolution of 104 proteins simultaneously. This technique combines the resolution power of isoelectrofocalization (IEF), which distinguishes proteins by their isoelectric point (pI), with SDSPAGE (sodium dodecyl sulphate PAGE), in which proteins are separated according to their weight and molecular size. Our group is developing software algorithms for the automatic analysis of images obtained by 2DPAGE gel optical scanning; our aim is the reduction of human intervention in the analysis process, currently quite slow, operatordependent, and prone to errors. In our approach, image noise is first reduced, in order to limit false positives and protein missing. Proteins appear as dark spots on a light background, so the next step is local minima search. Then, the watershed transform is applied, which partitions the gel image into basins: each basin contains a single (recognized) minimum, but can possibly include more than one protein spot if lessdeep minima are masked by the main one. At this point, we perform a registration between the work image and an atlas image (already analyzed by a biologist), and map the atlas spot positions to the work image. Each basin is then used as a region of interest (ROI) in which the shape of the spot (or spots) is fit to a model through a χ2-minimization procedure. The coordinates of the transformed spots are used as the fit initialization parameters.