Detection of pulmonary nodules in CT images based on fuzzy integrated active contour model and hybrid parametric mixture model

Automatic Liver Segmentation in CT Images with Enhanced GAN and Mask Region-Based CNN Architectures

Liver image segmentation has been increasingly employed for key medical purposes, including liver functional assessment, disease diagnosis, and treatment. In this work, we introduce a liver image segmentation method based on generative adversarial networks (GANs) and mask region-based convolutional neural networks (Mask R-CNN). Firstly, since most resulting images have noisy features, we further explored the combination of Mask R-CNN and GANs in order to enhance the pixel-wise classification. Secondly, k-means clustering was used to lock the image aspect ratio, in order to get more essential anchors which can help boost the segmentation performance. Finally, we proposed a GAN Mask R-CNN algorithm which achieved superior performance in comparison with the conventional Mask R-CNN, Mask-CNN, and k-means algorithms in terms of the Dice similarity coefficient (DSC) and the MICCAI metrics. The proposed algorithm also achieved superior performance in comparison with ten state-of-the-art algorithms in terms of six Boolean indicators. We hope that our work can be effectively used to optimize the segmentation and classification of liver anomalies.

Negative CT Contrast Agents for the Diagnosis of Malignant Osteosarcoma

The current positive computed tomography (CT) contrast agents (PCTCAs) including clinical iodides, present high CT density value (CT-DV). However, they are incapable for the accurate diagnosis of some diseases with high CT-DV, such as osteosarcoma. Because bones and PCTCAs around osteosarcoma generate similar X-ray attenuations. Here, an innovative strategy of negative CT contrast agents (NCTCAs) to reduce the CT-DV of osteosarcoma is proposed, contributing to accurate detection of osteosarcoma. Hollow mesoporous silica nanoparticles, loading ammonia borane molecules and further modified by polyethylene glycol, are synthesized as NCTCAs for the diagnosis of osteosarcoma. The nanocomposites can produce H2 in situ at osteosarcoma areas by responding to the acidic microenvironment of osteosarcoma, resulting in nearly 20 times reduction of CT density in osteosarcoma. This helps form large CT density contrast between bones and osteosarcoma, and successfully achieves accurate diagnosis of osteosarcoma. Meanwhile, The NCTCAs strategy greatly expands the scope of CT application, and provides profound implications for the precise clinical diagnosis, treatment, and prognosis of diseases.

A novel approach to CAD system for the detection of lung nodules in CT images

Detection of pulmonary nodule plays a significant role in the diagnosis of lung cancer in early stage that improves the chances of survival of an individual. In this paper, a computer aided nodule detection method is proposed for the segmentation and detection of challenging nodules like juxtavascular and juxtapleural nodules. Lungs are segmented from computed tomography (CT) images using intensity thresholding; brief analysis of CT image histogram is done to select a suitable threshold value for better segmentation results. Simple morphological closing is used to include juxtapleural nodules in segmented lung regions. K-means clustering is applied for the initial detection and segmentation of potential nodules; shape specific morphological opening is implemented to refine segmentation outcomes. These segmented potential nodules are then divided into six groups on the basis of their thickness and percentage connectivity with lung walls. Grouping not only helped in improving system's efficiency but also reduced computational time, otherwise consumed in calculating and analyzing unnecessary features for all nodules. Different sets of 2D and 3D features are extracted from nodules in each group to eliminate false positives. Small size nodules are differentiated from false positives (FPs) on the basis of their salient features; sensitivity of the system for small nodules is 83.33%. SVM classifier is used for the classification of large nodules, for which the sensitivity of the proposed system is 93.8% applying 10-fold cross-validation. Receiver Operating Characteristic (ROC) curve is used for the analysis of CAD system. Overall sensitivity of the system is 91.65% with 3.19 FPs per case, and accuracy is 96.22%. The system took 3.8 seconds to analyze each image.

Segmentation of pulmonary nodules using adaptive local region energy with probability density function-based similarity distance and multi-features clustering

BACKGROUND

Pulmonary nodules in computerized tomography (CT) images are potential manifestations of lung cancer. Segmentation of potential nodule objects is the first necessary and crucial step in computer-aided detection system of pulmonary nodules. The segmentation of various types of nodules, especially for ground-glass opacity (GGO) nodules and juxta-vascular nodules, present various challenges. The nodule with GGO characteristic possesses typical intensity inhomogeneity and weak edges, which is difficult to define the boundary; the juxta-vascular nodule is connected to a vessel, and they have very similar intensities. Traditional segmentation methods may result in the problems of boundary leakage and a small volume over-segmentation. This paper deals with the above mentioned problems.

METHODS

A novel segmentation method for pulmonary nodules is proposed, which uses an adaptive local region energy model with probability density function (PDF)-based similarity distance and multi-features dynamic clustering refinement method. Our approach has several novel aspects: (1) in the proposed adaptive local region energy model, the local domain for local energy model is selected adaptively based on k-nearest-neighbour (KNN) estimate method, and measurable distances between probability density functions of multi-dimension features with high class separability are used to build the cost function. (2) A multi-features dynamic clustering method is used for the segmentation refinement of juxta-vascular nodules, which is based on the nodule segmentation using active contour model (ACM) with adaptive local region energy and vessel segmentation using flow direction feature (FDF)-based region growing method. (3) it handles various types of nodules under a united framework.

RESULTS

The proposed method has been validated on a clinical dataset of 113 chest CT scans that contain 157 nodules determined by a ground truth reading process, and evaluating the algorithm on the provided data leads to an average Tanimoto/Jaccard error of 0.17, 0.20 and 0.24 for GGO, juxta-vascular and GGO juxta-vascular nodules, respectively.

CONCLUSIONS

Experimental results show desirable performances of the proposed method. The proposed segmentation method outperforms the traditional methods.

Nasopharyngeal carcinoma segmentation via HMRF-EM with maximum entropy

This paper presents a novel automatic nasopharyngeal carcinoma segmentation approach used in magnetic resonance images. Adaptive calculation of the nasopharyngeal region location is first performed. The contour of the tumor is determined through distance regularized level set evolution with the initial contour obtained by the nearest neighbor graph model. To further refine the segmentation, a hidden Markov random field model with maximum entropy (HMRF-EM) is introduced to model the spatial information with prior knowledge. The proposed method is tested on magnetic resonance images of 26 nasopharyngeal carcinoma patients, and achieves good results.

Classification of pulmonary nodules by using hybrid features

Early detection of pulmonary nodules is extremely important for the diagnosis and treatment of lung cancer. In this study, a new classification approach for pulmonary nodules from CT imagery is presented by using hybrid features. Four different methods are introduced for the proposed system. The overall detection performance is evaluated using various classifiers. The results are compared to similar techniques in the literature by using standard measures. The proposed approach with the hybrid features results in 90.7% classification accuracy (89.6% sensitivity and 87.5% specificity).