Abstract:Accurate segmentation of lungs in pathological thoracic computed tomography (CT) scans plays an important role in pulmonary disease diagnosis. However, it is still a challenging task due to the variability of pathological lung appearances and shapes. In this paper, we proposed a novel segmentation algorithm based on random forest (RF), deep convolutional network, and multi-scale superpixels for segmenting pathological lungs from thoracic CT images accurately. A pathological thoracic CT image is first segmented based on multi-scale superpixels, and deep features, texture, and intensity features extracted from superpixels are taken as inputs of a group of RF classifiers. With the fusion of classification results of RFs by a fractional-order gray correlation approach, we capture an initial segmentation of pathological lungs. We finally utilize a divide-and-conquer strategy to deal with segmentation refinement combining contour correction of left lungs and region repairing of right lungs. Our algorithm is tested on a group of thoracic CT images affected with interstitial lung diseases. Experiments show that our algorithm can achieve a high segmentation accuracy with an average DSC of 96.45% and PPV of 95.07%. Compared with several existing lung segmentation methods, our algorithm exhibits a robust performance on pathological lung segmentation. Our algorithm can be employed reliably for lung field segmentation of pathologic thoracic CT images with a high accuracy, which is helpful to assist radiologists to detect the presence of pulmonary diseases and quantify its shape and size in regular clinical practices.

A New Lung Segmentation Algorithm for Pathological CT Images

Pathological Lung Segmentation for CT Images Based on Snake Model

Interactive Lung Segmentation Algorithm for CT Chest Images Based on Live-Wire Model and Snake Model

Research on the Lung Parenchyma Segmentation Based on Chest CT

Segmentation of Lung Parenchyma in Chest CT Image Based on T-Snakes

Imporved Interactive Lung Segmentation Algorithm for CT Chest Images

Automatic Lung Segmentation Based on Image Decomposition and Wavelet Transform

A method of lung segmentation based on low dose CT images

An Automatic Lung Field Segmentation Algorithm Based on Improved Snake Model in X-ray Chest Radiograph

Global Optimal Hybrid Geometric Active Contour for Automated Lung Segmentation on CT Images

Many is Better Than One: an Integration of Multiple Simple Strategies for Accurate Lung Segmentation in CT Images

Extracting Lungs from CT Images Via Deep Convolutional Neural Network Based Segmentation and Two-Pass Contour Refinement

Automatic Pathological Lung Segmentation in Low-Dose CT Image Using Eigenspace Sparse Shape Composition

Lung Segmentation From Ct With Severe Pathologies Using Anatomical Constraints

Automatic Lung Segmentation in Chest CT Image Using Morphology

3-D Segmentation Of Lung Nodules In Ct Images Based On Improved Level Set Method

Automated Lung Segmentation for Chest CT Images Used for Computer Aided Diagnostics

Pathological Lung Segmentation Based on Random Forest Combined with Deep Model and Multi-Scale Superpixels

A Fast Automatic Method Of Lung Segmentation In Ct Images Using Mathematical Morphology

Lung Segmentation Based on Random Forest and Multi-Scale Edge Detection

A Dynamic Adaptive 3D Voxel-growing Segmentation Algorithm for Pulmonary CT Images