The liver is considered as one of the complicated organs in human body. It has close proximity to the neighboring organs in abdomen with numerous anatomical variations. It is difficult to find out the severity of disease connected to the liver unless the scanned image is subjected to segmentation process. The difficulty level also varies with the diseases that affect the liver. Any of the disease alters its density, homogeneity, color and texture. Liver image segmentation is necessary to identify the complexity and severity of the disease and it remains as an open challenge to researchers. Among all liver segmentation algorithms, threshold segmentation is fastest, simplest and numerically less complex. The accuracy of threshold-based segmentation lies in the selection of threshold values which separates foreground and background. This paper proposes a novel multi-threshold liver segmentation model based on “Slope Difference Distribution” (SDD) of image histogram. It consists of three stages. In the first stage, the noise in Computed Tomography (CT) scan image is reduced using a median filter. In the second stage, automatic threshold values are obtained from SDD of image histogram. These threshold values separate the liver image accurately from abdominal CT scan image. In the third stage, seed points are selected automatically which grow outwardly using the Fast Marching Method (FMM) discovering liver border in CT scan image. The proposed model is tested on 55 CT scan images and it is providing satisfactory results.
Computer-aided detection and diagnosis (CAD) of lung-related diseases will be helpful for early detection. Lung parenchyma segmentation is considered as a prerequisite for most of CAD systems. The available traditional methods for lung parenchyma segmentation are not accurate because the nodules that adhere to the lung pleura are recognized as fat. This paper proposes an automated lung parenchyma segmentation for accurate detection of lung nodules, mainly juxtapleural nodules. The proposed method includes the bidirectional chain code to improve the segmentation, and the support vector machine classifier is used to avoid false inclusion of regions. The proposed method is verified on various datasets for robustness of the algorithm. This automated method provides an accuracy of 97% in segmentation compared to ground truth results obtained by experts, which drastically reduces the complexity and intervention of a radiologist.
The traditional segmentation methods available for pulmonary parenchyma are not accurate because most of the methods exclude nodules or tumors adhering to the lung pleural wall as fat. In this paper, several techniques are exhaustively used in different phases, including two-dimensional (2D) optimal threshold selection and 2D reconstruction for lung parenchyma segmentation. Then, lung parenchyma boundaries are repaired using improved chain code and Bresenham pixel interconnection. The proposed method of segmentation and repairing is fully automated. Here, 21 thoracic computer tomography slices having juxtapleural nodules and 115 lung parenchyma scans are used to verify the robustness and accuracy of the proposed method. Results are compared with the most cited active contour methods. Empirical results show that the proposed fully automated method for segmenting lung parenchyma is more accurate. The proposed method is 100% sensitive to the inclusion of nodules/tumors adhering to the lung pleural wall, the juxtapleural nodule segmentation is >98%, and the lung parenchyma segmentation accuracy is >96%.
Computer-aided diagnosis of lung segmentation is the fundamental requirement to diagnose lung diseases. In this paper, a two-dimensional (2D) Otsu algorithm by Darwinian particle swarm optimization (DPSO) and fractional-order Darwinian particle swarm optimization (FODPSO) is proposed to segment the pulmonary parenchyma from the lung image obtained through computed tomography (CT) scans. The proposed method extracts pulmonary parenchyma from multi-sliced CT. This is a preprocessing step to identify pulmonary diseases such as emphysema, tumor, and lung cancer. Image segmentation plays a significant role in automated pulmonary disease diagnosis. In traditional 2D Otsu, exhaustive search plays an important role in image segmentation. However, the main disadvantage of the 2D Otsu method is its complex computation and processing time. In this paper, the 2D Otsu method optimized by DPSO and FODPSO is developed to reduce complex computations and time. The efficient segmentation is very important in object classification and detection. The particle swarm optimization (PSO) method is widely used to speed up the computation and maintain the same efficiency. In the proposed algorithm, the limitation of PSO of getting trapped in local optimum solutions is overcome. The segmentation technique is assessed and equated with the traditional 2D Otsu method. The test results demonstrate that the proposed strategy gives better results. The algorithm is tested on the Lung Image Database Consortium image collections.
The authors presented a novel computerised scheme to segment pulmonary nodules using freehand sketch. Here, freehand sketching is considered to identify the location of a nodule and it serves as a natural way and also it provides inferring adaptive information i.e. size, density, texture and mass centre etc. The proposed scheme includes two phases. In first phase using the freehand sketch analysis the multi seed points to select RoI. In the second phase nodule volumetric extraction is done using geometric modelling and implicit surface reconstruction for volumetric analysis. Spherical bins are used for ray triangle intersections and then local implicit surface fitting and blending method for surface reconstruction and depiction. The performance of the proposed scheme is assessed by accuracy and consistency using 112 CT examinations from LIDC. The IoU and ASD were used to assess the discrepancy between proposed method and inter observer agreement in the proposed approach. In estimating the reproducibility, the discrepancy in proposed scheme and the manual contouring by the expert is observed to be on an average of 0.13 ± 0.07 mm and 3.04 ± 1.7 mm respectively. The experiment shows that, the proposed scheme performs reasonably well and demonstrate merit of freehand sketch.