Texture-based identification and characterization of interstitial pneumonia patterns in lung multidetector CT

Pattern classification of interstitial lung diseases from computed tomography images using a ResNet-based network with a split-transform-merge strategy and split attention

In patients with interstitial lung disease (ILD), accurate pattern assessment from their computed tomography (CT) images could help track lung abnormalities and evaluate treatment efficacy. Based on excellent image classification performance, convolutional neural networks (CNNs) have been massively investigated for classifying and labeling pathological patterns in the CT images of ILD patients. However, previous studies rarely considered the three-dimensional (3D) structure of the pathological patterns of ILD and used two-dimensional network input. In addition, ResNet-based networks such as SE-ResNet and ResNeXt with high classification performance have not been used for pattern classification of ILD. This study proposed a SE-ResNeXt-SA-18 for classifying pathological patterns of ILD. The SE-ResNeXt-SA-18 integrated the multipath design of the ResNeXt and the feature weighting of the squeeze-and-excitation network with split attention. The classification performance of the SE-ResNeXt-SA-18 was compared with the ResNet-18 and SE-ResNeXt-18. The influence of the input patch size on classification performance was also evaluated. Results show that the classification accuracy was increased with the increase of the patch size. With a 32 × 32 × 16 input, the SE-ResNeXt-SA-18 presented the highest performance with average accuracy, sensitivity, and specificity of 0.991, 0.979, and 0.994. High-weight regions in the class activation maps of the SE-ResNeXt-SA-18 also matched the specific pattern features. In comparison, the performance of the SE-ResNeXt-SA-18 is superior to the previously reported CNNs in classifying the ILD patterns. We concluded that the SE-ResNeXt-SA-18 could help track or monitor the progress of ILD through accuracy pattern classification.

Semi-Supervised Segmentation of Interstitial Lung Disease Patterns from CT Images via Self-Training with Selective Re-Training

Accurate segmentation of interstitial lung disease (ILD) patterns from computed tomography (CT) images is an essential prerequisite to treatment and follow-up. However, it is highly time-consuming for radiologists to pixel-by-pixel segment ILD patterns from CT scans with hundreds of slices. Consequently, it is hard to obtain large amounts of well-annotated data, which poses a huge challenge for data-driven deep learning-based methods. To alleviate this problem, we propose an end-to-end semi-supervised learning framework for the segmentation of ILD patterns (ESSegILD) from CT images via self-training with selective re-training. The proposed ESSegILD model is trained using a large CT dataset with slice-wise sparse annotations, i.e., only labeling a few slices in each CT volume with ILD patterns. Specifically, we adopt a popular semi-supervised framework, i.e., Mean-Teacher, that consists of a teacher model and a student model and uses consistency regularization to encourage consistent outputs from the two models under different perturbations. Furthermore, we propose introducing the latest self-training technique with a selective re-training strategy to select reliable pseudo-labels generated by the teacher model, which are used to expand training samples to promote the student model during iterative training. By leveraging consistency regularization and self-training with selective re-training, our proposed ESSegILD can effectively utilize unlabeled data from a partially annotated dataset to progressively improve the segmentation performance. Experiments are conducted on a dataset of 67 pneumonia patients with incomplete annotations containing over 11,000 CT images with eight different lung patterns of ILDs, with the results indicating that our proposed method is superior to the state-of-the-art methods.

Interstitial lung disease detection using template matching combined sparse coding and blended multi class support vector machine

Interstitial lung disease (ILD), representing a collection of disorders, is considered to be the deadliest one, which increases the mortality rate of humans. In this paper, an automated scheme for detection and classification of ILD patterns is presented, which eliminates low inter-class feature variation and high intra-class feature variation in patterns, caused by translation and illumination effects. A novel and efficient feature extraction method named Template-Matching Combined Sparse Coding (TMCSC) is proposed, which extracts features invariant to translation and illumination effects, from defined regions of interest (ROI) within lung parenchyma. The translated image patch is compared with all possible templates of the image using template matching process. The corresponding sparse matrix for the set of translated image patches and their nearest template is obtained by minimizing the objective function of the similarity matrix of translated image patch and the template. A novel Blended-Multi Class Support Vector Machine (B-MCSVM) is designed for tackling high-intra class feature variation problems, which provides improved classification accuracy. Region of interests (ROIs) of five lung tissue patterns (healthy, emphysema, ground glass, micronodule, and fibrosis) selected from an internal multimedia database that contains high-resolution computed tomography (HRCT) image series are identified and utilized in this work. Performance of the proposed scheme outperforms most of the state-of-art multi-class classification algorithms.

How AI Can Help in the Diagnostic Dilemma of Pulmonary Nodules

Simple Summary

Pulmonary nodules are considered a sign of bronchogenic carcinoma, detecting them early will reduce their progression and can save lives. Lung cancer is the second most common type of cancer in both men and women. This manuscript discusses the current applications of artificial intelligence (AI) in lung segmentation as well as pulmonary nodule segmentation and classification using computed tomography (CT) scans, published in the last two decades, in addition to the limitations and future prospects in the field of AI.

Abstract

Pulmonary nodules are the precursors of bronchogenic carcinoma, its early detection facilitates early treatment which save a lot of lives. Unfortunately, pulmonary nodule detection and classification are liable to subjective variations with high rate of missing small cancerous lesions which opens the way for implementation of artificial intelligence (AI) and computer aided diagnosis (CAD) systems. The field of deep learning and neural networks is expanding every day with new models designed to overcome diagnostic problems and provide more applicable and simply used models. We aim in this review to briefly discuss the current applications of AI in lung segmentation, pulmonary nodule detection and classification.

Special Convolutional Neural Network for Identification and Positioning of Interstitial Lung Disease Patterns in Computed Tomography Images

In this paper, automated detection of interstitial lung disease patterns in high resolution computed tomography images is achieved by developing a faster region-based convolutional network based detector with GoogLeNet as a backbone. GoogLeNet is simplified by removing few inception models and used as the backbone of the detector network. The proposed framework is developed to detect several interstitial lung disease patterns without doing lung field segmentation. The proposed method is able to detect the five most prevalent interstitial lung disease patterns: fibrosis, emphysema, consolidation, micronodules and ground-glass opacity, as well as normal. Five-fold cross-validation has been used to avoid bias and reduce over-fitting. The proposed framework performance is measured in terms of F-score on the publicly available MedGIFT database. It outperforms state-of-the-art techniques. The detection is performed at slice level and could be used for screening and differential diagnosis of interstitial lung disease patterns using high resolution computed tomography images.

Pulmonary Functional Imaging: Part 2-State-of-the-Art Clinical Applications and Opportunities for Improved Patient Care

Pulmonary functional imaging may be defined as the regional quantification of lung function by using primarily CT, MRI, and nuclear medicine techniques. The distribution of pulmonary physiologic parameters, including ventilation, perfusion, gas exchange, and biomechanics, can be noninvasively mapped and measured throughout the lungs. This information is not accessible by using conventional pulmonary function tests, which measure total lung function without viewing the regional distribution. The latter is important because of the heterogeneous distribution of virtually all lung disorders. Moreover, techniques such as hyperpolarized xenon 129 and helium 3 MRI can probe lung physiologic structure and microstructure at the level of the alveolar-air and alveolar-red blood cell interface, which is well beyond the spatial resolution of other clinical methods. The opportunities, challenges, and current stage of clinical deployment of pulmonary functional imaging are reviewed, including applications to chronic obstructive pulmonary disease, asthma, interstitial lung disease, pulmonary embolism, and pulmonary hypertension. Among the challenges to the deployment of pulmonary functional imaging in routine clinical practice are the need for further validation, establishment of normal values, standardization of imaging acquisition and analysis, and evidence of patient outcomes benefit. When these challenges are addressed, it is anticipated that pulmonary functional imaging will have an expanding role in the evaluation and management of patients with lung disease.

Three-Dimensional Multiscale Fuzzy Entropy: Validation and Application to Idiopathic Pulmonary Fibrosis

Idiopathic Pulmonary Fibrosis (IPF) is a chronic, severe, and progressive lung disease with short life expectancy. Based on information theory and entropy measurement, a three-dimensional multiscale fuzzy entropy (MFE _3D) algorithm is proposed to identify IPF patients from their computed tomography (CT) volumetric data. First, the validation of the algorithm was performed by analyzing several volumetric synthetic noises (white, blue, brown, and pink), MIX(p) processes-based volumes, and texture-based volumes. The entropy values obtained by MFE _3D were consistent with the values obtained using the one, and two-dimensional versions, validating its use in biomedical data. Hence, MFE _3D was applied to CT scans to identify the existence of IPF within two different groups, one of healthy subjects (26) and another of IPF patients (26). Statistical differences were found (p < 0.05) between the entropy values of each group in 5 scale factors out of 10. These results demonstrate that MFE _3D could be an interesting metric to identify IPF in CT scans.

Deep learning for screening of interstitial lung disease patterns in high-resolution CT images

AIM

To develop a screening tool for the detection of interstitial lung disease (ILD) patterns using a deep-learning method.

MATERIALS AND METHODS

A fully convolutional network was used for semantic segmentation of several ILD patterns. Improved segmentation of ILD patterns was achieved using multi-scale feature extraction. Dilated convolution was used to maintain the resolution of feature maps and to enlarge the receptive field. The proposed method was evaluated on a publicly available ILD database (MedGIFT) and a private clinical research database. Several metrics, such as success rate, sensitivity, and false positives per section were used for quantitative evaluation of the proposed method.

RESULTS

Sections with fibrosis and emphysema were detected with a similar success rate and sensitivity for both databases but the performance of detection was lower for consolidation compared to fibrosis and emphysema.

CONCLUSION

Automatic identification of ILD patterns in a high-resolution computed tomography (CT) image was implemented using a deep-learning framework. Creation of a pre-trained model with natural images and subsequent transfer learning using a particular database gives acceptable results.

Pulmonary Textures Classification via a Multi-Scale Attention Network

Precise classification of pulmonary textures is crucial to develop a computer aided diagnosis (CAD) system of diffuse lung diseases (DLDs). Although deep learning techniques have been applied to this task, the classification performance is not satisfied for clinical requirements, since commonly-used deep networks built by stacking convolutional blocks are not able to learn discriminative feature representation to distinguish complex pulmonary textures. For addressing this problem, we design a multi-scale attention network (MSAN) architecture comprised by several stacked residual attention modules followed by a multi-scale fusion module. Our deep network can not only exploit powerful information on different scales but also automatically select optimal features for more discriminative feature representation. Besides, we develop visualization techniques to make the proposed deep model transparent for humans. The proposed method is evaluated by using a large dataset. Experimental results show that our method has achieved the average classification accuracy of 94.78% and the average f-value of 0.9475 in the classification of 7 categories of pulmonary textures. Besides, visualization results intuitively explain the working behavior of the deep network. The proposed method has achieved the state-of-the-art performance to classify pulmonary textures on high resolution CT images.

Comparing effectiveness of prognostic tests in idiopathic pulmonary fibrosis

Introduction: Idiopathic pulmonary fibrosis (IPF) is a debilitating and progressive fibrotic interstitial lung disease often resulting in death over several years. Prediction of disease course or survival remains of keen interest for clinicians and patients though a commonly used test or tool remain elusive. Areas covered: We undertook a comprehensive review of the published literature highlighting prognostic indicators and predictors of survival in IPF. Baseline and longitudinal clinical, functional, histopathologic, and radiologic findings have been extensively studied as prognostic predictors, both individually and in composite models. Recent approaches include automated quantifiable radiologic scoring, circulating biomarkers, and genetic polymorphisms or abnormalities. This review highlights individual and composite predictors and their relative utility in clinical practice and research studies. Expert opinion: There is a growing body of knowledge highlighting readily available individual and composite predictors of outcome, though none have come to the forefront for common clinical use. Recent advances include quantitative imaging analysis, circulating serologic markers, and genetic testing, which may be more standardized and less prone to lead-time bias or related complications and comorbidities.

Experimental and quantitative imaging techniques in interstitial lung disease

Interstitial lung diseases (ILDs) are a heterogeneous group of conditions, with a wide and complex variety of imaging features. Difficulty in monitoring, treating and exploring novel therapies for these conditions is in part due to the lack of robust, readily available biomarkers. Radiological studies are vital in the assessment and follow-up of ILD, but currently CT analysis in clinical practice is qualitative and therefore somewhat subjective. In this article, we report on the role of novel and quantitative imaging techniques across a range of imaging modalities in ILD and consider how they may be applied in the assessment and understanding of ILD. We critically appraised evidence found from searches of Ovid online, PubMed and the TRIP database for novel and quantitative imaging studies in ILD. Recent studies have explored the capability of texture-based lung parenchymal analysis in accurately quantifying several ILD features. Newer techniques are helping to overcome the challenges inherent to such approaches, in particular distinguishing peripheral reticulation of lung parenchyma from pleura and accurately identifying the complex density patterns that accompany honeycombing. Robust and validated texture-based analysis may remove the subjectivity that is inherent to qualitative reporting and allow greater objective measurements of change over time. In addition to lung parenchymal feature quantification, pulmonary vessel volume analysis on CT has demonstrated prognostic value in two retrospective analyses and may be a sign of vascular changes in ILD which, to date, have been difficult to quantify in the absence of overt pulmonary hypertension. Novel applications of existing imaging techniques, such as hyperpolarised gas MRI and positron emission tomography (PET), show promise in combining structural and functional information. Although structural imaging of lung tissue is inherently challenging in terms of conventional proton MRI techniques, inroads are being made with ultrashort echo time, and dynamic contrast-enhanced MRI may be used for lung perfusion assessment. In addition, inhaled hyperpolarised ¹²⁹Xenon gas MRI may provide multifunctional imaging metrics, including assessment of ventilation, intra-acinar gas diffusion and alveolar-capillary diffusion. PET has demonstrated high standard uptake values (SUVs) of ¹⁸F-fluorodeoxyglucose in fibrosed lung tissue, challenging the assumption that these are ‘burned out’ and metabolically inactive regions. Regions that appear structurally normal also appear to have higher SUV, warranting further exploration with future longitudinal studies to assess if this precedes future regions of macroscopic structural change. Given the subtleties involved in diagnosing, assessing and predicting future deterioration in many forms of ILD, multimodal quantitative lung structure-function imaging may provide the means of identifying novel, sensitive and clinically applicable imaging markers of disease. Such imaging metrics may provide mechanistic and phenotypic information that can help direct appropriate personalised therapy, can be used to predict outcomes and could potentially be more sensitive and specific than global pulmonary function testing. Quantitative assessment may objectively assess subtle change in character or extent of disease that can assist in efficacy of antifibrotic therapy or detecting early changes of potentially pneumotoxic drugs involved in early intervention studies.

Segmentation of lung parenchyma in CT images using CNN trained with the clustering algorithm generated dataset

Background

Lung segmentation constitutes a critical procedure for any clinical-decision supporting system aimed to improve the early diagnosis and treatment of lung diseases. Abnormal lungs mainly include lung parenchyma with commonalities on CT images across subjects, diseases and CT scanners, and lung lesions presenting various appearances. Segmentation of lung parenchyma can help locate and analyze the neighboring lesions, but is not well studied in the framework of machine learning.

Methods

We proposed to segment lung parenchyma using a convolutional neural network (CNN) model. To reduce the workload of manually preparing the dataset for training the CNN, one clustering algorithm based method is proposed firstly. Specifically, after splitting CT slices into image patches, the k-means clustering algorithm with two categories is performed twice using the mean and minimum intensity of image patch, respectively. A cross-shaped verification, a volume intersection, a connected component analysis and a patch expansion are followed to generate final dataset. Secondly, we design a CNN architecture consisting of only one convolutional layer with six kernels, followed by one maximum pooling layer and two fully connected layers. Using the generated dataset, a variety of CNN models are trained and optimized, and their performances are evaluated by eightfold cross-validation. A separate validation experiment is further conducted using a dataset of 201 subjects (4.62 billion patches) with lung cancer or chronic obstructive pulmonary disease, scanned by CT or PET/CT. The segmentation results by our method are compared with those yielded by manual segmentation and some available methods.

Results

A total of 121,728 patches are generated to train and validate the CNN models. After the parameter optimization, our CNN model achieves an average F-score of 0.9917 and an area of curve up to 0.9991 for classification of lung parenchyma and non-lung-parenchyma. The obtain model can segment the lung parenchyma accurately for 201 subjects with heterogeneous lung diseases and CT scanners. The overlap ratio between the manual segmentation and the one by our method reaches 0.96.

Conclusions

The results demonstrated that the proposed clustering algorithm based method can generate the training dataset for CNN models. The obtained CNN model can segment lung parenchyma with very satisfactory performance and have the potential to locate and analyze lung lesions.

Quantitative morphometric analysis of adult teleost fish by X-ray computed tomography

Vertebrate models provide indispensable paradigms to study development and disease. Their analysis requires a quantitative morphometric study of the body, organs and tissues. This is often impeded by pigmentation and sample size. X-ray micro-computed tomography (micro-CT) allows high-resolution volumetric tissue analysis, largely independent of sample size and transparency to visual light. Importantly, micro-CT data are inherently quantitative. We report a complete pipeline of high-throughput 3D data acquisition and image analysis, including tissue preparation and contrast enhancement for micro-CT imaging down to cellular resolution, automated data processing and organ or tissue segmentation that is applicable to comparative 3D morphometrics of small vertebrates. Applied to medaka fish, we first create an annotated anatomical atlas of the entire body, including inner organs as a quantitative morphological description of an adult individual. This atlas serves as a reference model for comparative studies. Using isogenic medaka strains we show that comparative 3D morphometrics of individuals permits identification of quantitative strain-specific traits. Thus, our pipeline enables high resolution morphological analysis as a basis for genotype-phenotype association studies of complex genetic traits in vertebrates.

CT Texture Analysis: Definitions, Applications, Biologic Correlates, and Challenges

This review discusses potential oncologic and nononcologic applications of CT texture analysis ( CTTA CT texture analysis ), an emerging area of "radiomics" that extracts, analyzes, and interprets quantitative imaging features. CTTA CT texture analysis allows objective assessment of lesion and organ heterogeneity beyond what is possible with subjective visual interpretation and may reflect information about the tissue microenvironment. CTTA CT texture analysis has shown promise in lesion characterization, such as differentiating benign from malignant or more biologically aggressive lesions. Pretreatment CT texture features are associated with histopathologic correlates such as tumor grade, tumor cellular processes such as hypoxia or angiogenesis, and genetic features such as KRAS or epidermal growth factor receptor (EGFR) mutation status. In addition, and likely as a result, these CT texture features have been linked to prognosis and clinical outcomes in some tumor types. CTTA CT texture analysis has also been used to assess response to therapy, with decreases in tumor heterogeneity generally associated with pathologic response and improved outcomes. A variety of nononcologic applications of CTTA CT texture analysis are emerging, particularly quantifying fibrosis in the liver and lung. Although CTTA CT texture analysis seems to be a promising imaging biomarker, there is marked variability in methods, parameters reported, and strength of associations with biologic correlates. Before CTTA CT texture analysis can be considered for widespread clinical implementation, standardization of tumor segmentation and measurement techniques, image filtration and postprocessing techniques, and methods for mathematically handling multiple tumors and time points is needed, in addition to identification of key texture parameters among hundreds of potential candidates, continued investigation and external validation of histopathologic correlates, and structured reporting of findings. ^©RSNA, 2017.

Differences in Texture Analysis Parameters Between Active Alveolitis and Lung Fibrosis in Chest CT of Patients with Systemic Sclerosis: A Feasibility Study

RATIONALE AND OBJECTIVES

This study aimed to determine the diagnostic aid of computed tomography (CT) features for the differentiation of active alveolitis and fibrosis using a CT texture analysis (CTTA) prototype and CT densitometry in patients with systemic sclerosis (SSc) using ancillary high-resolution computed tomography (HRCT) features and their longitudinal course as standard of reference.

MATERIALS AND METHODS

We retrospectively analyzed thin-slice noncontrast chest CT image data of 43 patients with SSc (18 men, mean age 51.55 ± 15.52 years; range 23-71 years). All of them had repeated noncontrast enhanced HRCT of the lung. Classification into active alveolitis or fibrosis was done on HRCT based on classical HRCT findings (active alveolitis [19; 44.2%] and fibrosis [24; 55.8%]) and their course at midterm. Results were compared to pulmonary functional tests and were followed up by CT. Ground glass opacity was considered suggestive of alveolitis, whereas coarse reticulation with parenchymal distortion, traction bronchiectasis, and honeycombing were assigned to fibrosis.

RESULTS

Statistically significant differences in CTTA were found for first-order textural features (mean intensity, average, deviation, skewness) and second-order statistics (entropy of co-occurrence matrix, mean number of nonuniformity (NGLDM), entropy of NGLDM, entropy of heterogeneity, intensity, and average). Cut-off value for the prediction of fibrosis at baseline was significant for entropy of intensity (P value < .001) and for mean deviation (P value < .001), and for prediction of alveolitis was significant for uniformity of intensity (P value < .001) and for NGLDM (P value < .001). At pulmonary functional tests, forced expiratory volume in 1 second and single-breath diffusion capacity for carbon monoxide were significantly lower in fibrosis than in alveolitis 2.03 ± 0.78 vs. 2.61 ± 0.83, P < .016 and 4.51 ± 1.61 vs. 6.04 ± 1.75, P < .009, respectively. Differences in CT densitometry between alveolitis and fibrosis were not significant.

CONCLUSIONS

CTTA parameters are significantly different in active alveolitis vs. fibrosis in patients with SSc and may be helpful for differentiation of these two entities.

Multisource Transfer Learning With Convolutional Neural Networks for Lung Pattern Analysis

Early diagnosis of interstitial lung diseases is crucial for their treatment, but even experienced physicians find it difficult, as their clinical manifestations are similar. In order to assist with the diagnosis, computer-aided diagnosis systems have been developed. These commonly rely on a fixed scale classifier that scans CT images, recognizes textural lung patterns, and generates a map of pathologies. In a previous study, we proposed a method for classifying lung tissue patterns using a deep convolutional neural network (CNN), with an architecture designed for the specific problem. In this study, we present an improved method for training the proposed network by transferring knowledge from the similar domain of general texture classification. Six publicly available texture databases are used to pretrain networks with the proposed architecture, which are then fine-tuned on the lung tissue data. The resulting CNNs are combined in an ensemble and their fused knowledge is compressed back to a network with the original architecture. The proposed approach resulted in an absolute increase of about 2% in the performance of the proposed CNN. The results demonstrate the potential of transfer learning in the field of medical image analysis, indicate the textural nature of the problem and show that the method used for training a network can be as important as designing its architecture.

Advances in Imaging and Automated Quantification of Pulmonary Diseases in Non-neoplastic Diseases

Histological examination has always been the gold standard for the detection and quantification of lung remodeling. However, this method has some limitations regarding the invasiveness of tissue acquisition. Quantitative imaging methods enable the acquisition of valuable information on lung structure and function without the removal of tissue from the body; thus, they are useful for disease identification and follow-up. This article reviews the various quantitative imaging modalities used currently for the non-invasive study of chronic obstructive pulmonary disease, asthma, and interstitial lung diseases. Some promising computer-aided diagnosis methods are also described.

Lung Pattern Classification for Interstitial Lung Diseases Using a Deep Convolutional Neural Network

Automated tissue characterization is one of the most crucial components of a computer aided diagnosis (CAD) system for interstitial lung diseases (ILDs). Although much research has been conducted in this field, the problem remains challenging. Deep learning techniques have recently achieved impressive results in a variety of computer vision problems, raising expectations that they might be applied in other domains, such as medical image analysis. In this paper, we propose and evaluate a convolutional neural network (CNN), designed for the classification of ILD patterns. The proposed network consists of 5 convolutional layers with 2 × 2 kernels and LeakyReLU activations, followed by average pooling with size equal to the size of the final feature maps and three dense layers. The last dense layer has 7 outputs, equivalent to the classes considered: healthy, ground glass opacity (GGO), micronodules, consolidation, reticulation, honeycombing and a combination of GGO/reticulation. To train and evaluate the CNN, we used a dataset of 14696 image patches, derived by 120 CT scans from different scanners and hospitals. To the best of our knowledge, this is the first deep CNN designed for the specific problem. A comparative analysis proved the effectiveness of the proposed CNN against previous methods in a challenging dataset. The classification performance ( ~ 85.5%) demonstrated the potential of CNNs in analyzing lung patterns. Future work includes, extending the CNN to three-dimensional data provided by CT volume scans and integrating the proposed method into a CAD system that aims to provide differential diagnosis for ILDs as a supportive tool for radiologists.

Segmentation and Image Analysis of Abnormal Lungs at CT: Current Approaches, Challenges, and Future Trends

The computer-based process of identifying the boundaries of lung from surrounding thoracic tissue on computed tomographic (CT) images, which is called segmentation, is a vital first step in radiologic pulmonary image analysis. Many algorithms and software platforms provide image segmentation routines for quantification of lung abnormalities; however, nearly all of the current image segmentation approaches apply well only if the lungs exhibit minimal or no pathologic conditions. When moderate to high amounts of disease or abnormalities with a challenging shape or appearance exist in the lungs, computer-aided detection systems may be highly likely to fail to depict those abnormal regions because of inaccurate segmentation methods. In particular, abnormalities such as pleural effusions, consolidations, and masses often cause inaccurate lung segmentation, which greatly limits the use of image processing methods in clinical and research contexts. In this review, a critical summary of the current methods for lung segmentation on CT images is provided, with special emphasis on the accuracy and performance of the methods in cases with abnormalities and cases with exemplary pathologic findings. The currently available segmentation methods can be divided into five major classes: (a) thresholding-based, (b) region-based, (c) shape-based, (d) neighboring anatomy-guided, and (e) machine learning-based methods. The feasibility of each class and its shortcomings are explained and illustrated with the most common lung abnormalities observed on CT images. In an overview, practical applications and evolving technologies combining the presented approaches for the practicing radiologist are detailed.

Radiomics: a new application from established techniques

The increasing use of biomarkers in cancer have led to the concept of personalized medicine for patients. Personalized medicine provides better diagnosis and treatment options available to clinicians. Radiological imaging techniques provide an opportunity to deliver unique data on different types of tissue. However, obtaining useful information from all radiological data is challenging in the era of "big data". Recent advances in computational power and the use of genomics have generated a new area of research termed Radiomics. Radiomics is defined as the high throughput extraction of quantitative imaging features or texture (radiomics) from imaging to decode tissue pathology and creating a high dimensional data set for feature extraction. Radiomic features provide information about the gray-scale patterns, inter-pixel relationships. In addition, shape and spectral properties can be extracted within the same regions of interest on radiological images. Moreover, these features can be further used to develop computational models using advanced machine learning algorithms that may serve as a tool for personalized diagnosis and treatment guidance.

Automated classification of usual interstitial pneumonia using regional volumetric texture analysis in high-resolution computed tomography

OBJECTIVES

We propose a novel computational approach for the automated classification of classic versus atypical usual interstitial pneumonia (UIP).

MATERIALS AND METHODS

Thirty-three patients with UIP were enrolled in this study. They were classified as classic versus atypical UIP by a consensus of 2 thoracic radiologists with more than 15 years of experience using the American Thoracic Society evidence-based guidelines for computed tomography diagnosis of UIP. Two cardiothoracic fellows with 1 year of subspecialty training provided independent readings. The system is based on regional characterization of the morphological tissue properties of lung using volumetric texture analysis of multiple-detector computed tomography images. A simple digital atlas with 36 lung subregions is used to locate texture properties, from which the responses of multidirectional Riesz wavelets are obtained. Machine learning is used to aggregate and to map the regional texture attributes to a simple score that can be used to stratify patients with UIP into classic and atypical subtypes.

RESULTS

We compared the predictions on the basis of regional volumetric texture analysis with the ground truth established by expert consensus. The area under the receiver operating characteristic curve of the proposed score was estimated to be 0.81 using a leave-one-patient-out cross-validation, with high specificity for classic UIP. The performance of our automated method was found to be similar to that of the 2 fellows and to the agreement between experienced chest radiologists reported in the literature. However, the errors of our method and the fellows occurred on different cases, which suggests that combining human and computerized evaluations may be synergistic.

CONCLUSIONS

Our results are encouraging and suggest that an automated system may be useful in routine clinical practice as a diagnostic aid for identifying patients with complex lung disease such as classic UIP, obviating the need for invasive surgical lung biopsy and its associated risks.

Multicenter study of quantitative computed tomography analysis using a computer-aided three-dimensional system in patients with idiopathic pulmonary fibrosis

PURPOSE

To evaluate the feasibility of automated quantitative analysis with a three-dimensional (3D) computer-aided system (i.e., Gaussian histogram normalized correlation, GHNC) of computed tomography (CT) images from different scanners.

MATERIALS AND METHODS

Each institution's review board approved the research protocol. Informed patient consent was not required. The participants in this multicenter prospective study were 80 patients (65 men, 15 women) with idiopathic pulmonary fibrosis. Their mean age was 70.6 years. Computed tomography (CT) images were obtained by four different scanners set at different exposures. We measured the extent of fibrosis using GHNC, and used Pearson's correlation analysis, Bland-Altman plots, and kappa analysis to directly compare the GHNC results with manual scoring by radiologists. Multiple linear regression analysis was performed to determine the association between the CT data and forced vital capacity (FVC).

RESULTS

For each scanner, the extent of fibrosis as determined by GHNC was significantly correlated with the radiologists' score. In multivariate analysis, the extent of fibrosis as determined by GHNC was significantly correlated with FVC (p < 0.001). There was no significant difference between the results obtained using different CT scanners.

CONCLUSION

Gaussian histogram normalized correlation was feasible, irrespective of the type of CT scanner used.

Automated 3D ιnterstitial lung disease εxtent quantification: performance evaluation and correlation to PFTs

In this study, the performance of a recently proposed computer-aided diagnosis (CAD) scheme in detection and 3D quantification of reticular and ground glass pattern extent in chest computed tomography of interstitial lung disease (ILD) patients is evaluated. CAD scheme performance was evaluated on a dataset of 37 volumetric chest scans, considering five representative axial anatomical levels per scan. CAD scheme reliability analysis was performed by estimating agreement (intraclass correlation coefficient, ICC) of automatically derived ILD pattern extent to semi-quantitative disease extent assessment in terms of 29-point rating scale provided by two expert radiologists. Receiver operating characteristic (ROC) analysis was employed to assess CAD scheme accuracy in ILD pattern detection in terms of area under ROC curve (A z ). Correlation of reticular and ground glass volumetric pattern extent to pulmonary function tests (PFTs) was also investigated. CAD scheme reliability was substantial for ILD extent (ICC = 0.809) and distinct reticular pattern extent (0.806) and moderate for distinct ground glass pattern extent (0.543), performing within inter-observer agreement. CAD scheme demonstrated high accuracy in detecting total ILD (A z  = 0.950 ± 0.018), while accuracy in detecting distinct reticular and ground glass patterns was 0.920 ± 0.023 and 0.883 ± 0.024, respectively. Moderate and statistically significant negative correlation was found between reticular volumetric pattern extent and diffusing capacity, forced expiratory volume in 1 s, forced vital capacity, and total lung capacity (R = -0.581, -0.513, -0.494, and -0.446, respectively), similar to correlations found between radiologists' semi-quantitative ratings with PFTs. CAD-based quantification of disease extent is in agreement with radiologists' semi-quantitative assessment and correlates to specific PFTs, suggesting a potential imaging biomarker for ILD staging and management.

Early prediction of tumor recurrence based on CT texture changes after stereotactic ablative radiotherapy (SABR) for lung cancer

PURPOSE

Benign computed tomography (CT) changes due to radiation induced lung injury (RILI) are common following stereotactic ablative radiotherapy (SABR) and can be difficult to differentiate from tumor recurrence. The authors measured the ability of CT image texture analysis, compared to more traditional measures of response, to predict eventual cancer recurrence based on CT images acquired within 5 months of treatment.

METHODS

A total of 24 lesions from 22 patients treated with SABR were selected for this study: 13 with moderate to severe benign RILI, and 11 with recurrence. Three-dimensional (3D) consolidative and ground-glass opacity (GGO) changes were manually delineated on all follow-up CT scans. Two size measures of the consolidation regions (longest axial diameter and 3D volume) and nine appearance features of the GGO were calculated: 2 first-order features [mean density and standard deviation of density (first-order texture)], and 7 second-order texture features [energy, entropy, correlation, inverse difference moment (IDM), inertia, cluster shade, and cluster prominence]. For comparison, the corresponding response evaluation criteria in solid tumors measures were also taken for the consolidation regions. Prediction accuracy was determined using the area under the receiver operating characteristic curve (AUC) and two-fold cross validation (CV).

RESULTS

For this analysis, 46 diagnostic CT scans scheduled for approximately 3 and 6 months post-treatment were binned based on their recorded scan dates into 2-5 month and 5-8 month follow-up time ranges. At 2-5 months post-treatment, first-order texture, energy, and entropy provided AUCs of 0.79-0.81 using a linear classifier. On two-fold CV, first-order texture yielded 73% accuracy versus 76%-77% with the second-order features. The size measures of the consolidative region, longest axial diameter and 3D volume, gave two-fold CV accuracies of 60% and 57%, and AUCs of 0.72 and 0.65, respectively.

CONCLUSIONS

Texture measures of the GGO appearance following SABR demonstrated the ability to predict recurrence in individual patients within 5 months of SABR treatment. Appearance changes were also shown to be more accurately predictive of recurrence, as compared to size measures within the same time period. With further validation, these results could form the substrate for a clinically useful computer-aided diagnosis tool which could provide earlier salvage of patients with recurrence.

New techniques for assessing response after hypofractionated radiotherapy for lung cancer

Hypofractionated radiotherapy (HFRT) is an effective and increasingly-used treatment for early stage non-small cell lung cancer (NSCLC). Stereotactic ablative radiotherapy (SABR) is a form of HFRT and delivers biologically effective doses (BEDs) in excess of 100 Gy10 in 3-8 fractions. Excellent long-term outcomes have been reported; however, response assessment following SABR is complicated as radiation induced lung injury can appear similar to a recurring tumor on CT. Current approaches to scoring treatment responses include Response Evaluation Criteria in Solid Tumors (RECIST) and positron emission tomography (PET), both of which appear to have a limited role in detecting recurrences following SABR. Novel approaches to assess response are required, but new techniques should be easily standardized across centers, cost effective, with sensitivity and specificity that improves on current CT and PET approaches. This review examines potential novel approaches, focusing on the emerging field of quantitative image feature analysis, to distinguish recurrence from fibrosis after SABR.

Three-dimensional solid texture analysis in biomedical imaging: review and opportunities

Three-dimensional computerized characterization of biomedical solid textures is key to large-scale and high-throughput screening of imaging data. Such data increasingly become available in the clinical and research environments with an ever increasing spatial resolution. In this text we exhaustively analyze the state-of-the-art in 3-D biomedical texture analysis to identify the specific needs of the application domains and extract promising trends in image processing algorithms. The geometrical properties of biomedical textures are studied both in their natural space and on digitized lattices. It is found that most of the tissue types have strong multi-scale directional properties, that are well captured by imaging protocols with high resolutions and spherical spatial transfer functions. The information modeled by the various image processing techniques is analyzed and visualized by displaying their 3-D texture primitives. We demonstrate that non-convolutional approaches are expected to provide best results when the size of structures are inferior to five voxels. For larger structures, it is shown that only multi-scale directional convolutional approaches that are non-separable allow for an unbiased modeling of 3-D biomedical textures. With the increase of high-resolution isotropic imaging protocols in clinical routine and research, these models are expected to best leverage the wealth of 3-D biomedical texture analysis in the future. Future research directions and opportunities are proposed to efficiently model personalized image-based phenotypes of normal biomedical tissue and its alterations. The integration of the clinical and genomic context is expected to better explain the intra class variation of healthy biomedical textures. Using texture synthesis, this provides the exciting opportunity to simulate and visualize texture atlases of normal ageing process and disease progression for enhanced treatment planning and clinical care management.

Feature-based image patch approximation for lung tissue classification

In this paper, we propose a new classification method for five categories of lung tissues in high-resolution computed tomography (HRCT) images, with feature-based image patch approximation. We design two new feature descriptors for higher feature descriptiveness, namely the rotation-invariant Gabor-local binary patterns (RGLBP) texture descriptor and multi-coordinate histogram of oriented gradients (MCHOG) gradient descriptor. Together with intensity features, each image patch is then labeled based on its feature approximation from reference image patches. And a new patch-adaptive sparse approximation (PASA) method is designed with the following main components: minimum discrepancy criteria for sparse-based classification, patch-specific adaptation for discriminative approximation, and feature-space weighting for distance computation. The patch-wise labelings are then accumulated as probabilistic estimations for region-level classification. The proposed method is evaluated on a publicly available ILD database, showing encouraging performance improvements over the state-of-the-arts.

Texture-based medical image compression

Image processing is one of the most researched areas these days due to the flooding of the internet with an overload of images. The noble medicine industry is not left untouched. It has also suffered with an excess of patient record storage and maintenance. With the advent of automation of the industries in the world, the medicine industry has sought to change and provide a more portable feel to it, leading to the fields of telemedicine and such. Our algorithm comes in handy in such scenarios where large amount of data needs to be transmitted over the network for perusal by another consultant. We aim for a visual quality approach in our algorithm rather than pixel-wise fidelity. We utilize parameters of edges and textures as the basic parameters in our compression algorithm.

Automatic detection and quantification of tree-in-bud (TIB) opacities from CT scans

This study presents a novel computer-assisted detection (CAD) system for automatically detecting and precisely quantifying abnormal nodular branching opacities in chest computed tomography (CT), termed tree-in-bud (TIB) opacities by radiology literature. The developed CAD system in this study is based on 1) fast localization of candidate imaging patterns using local scale information of the images, and 2) Möbius invariant feature extraction method based on learned local shape and texture properties of TIB patterns. For fast localization of candidate imaging patterns, we use ball-scale filtering and, based on the observation of the pattern of interest, a suitable scale selection is used to retain only small size patterns. Once candidate abnormality patterns are identified, we extract proposed shape features from regions where at least one candidate pattern occupies. The comparative evaluation of the proposed method with commonly used CAD methods is presented with a dataset of 60 chest CTs (laboratory confirmed 39 viral bronchiolitis human parainfluenza CTs and 21 normal chest CTs). The quantitative results are presented as the area under the receiver operator characteristics curves and a computer score (volume affected by TIB) provided as an output of the CAD system. In addition, a visual grading scheme is applied to the patient data by three well-trained radiologists. Interobserver and observer-computer agreements are obtained by the relevant statistical methods over different lung zones. Experimental results demonstrate that the proposed CAD system can achieve high detection rates with an overall accuracy of 90.96%. Moreover, correlations of observer-observer (R(2)=0.8848, and observer-CAD agreements (R(2)=0.824, validate the feasibility of the use of the proposed CAD system in detecting and quantifying TIB patterns.