A Combination of Shape and Texture Features for Classification of Pulmonary Nodules in Lung CT Images

Dhara, Ashis Kumar; Mukhopadhyay, Sudipta; Dutta, Anirvan; Garg, Mandeep; Khandelwal, Niranjan

doi:10.1007/s10278-015-9857-6

Cited by 172 publications

(111 citation statements)

References 27 publications

Supporting

Mentioning

101

Contrasting

Unclassified

Order By: Relevance

“…Despite the fact that the GLCM-based texture analysis may be dated as method, it is still relevant to scientific literature and recent works have used texture features as image descriptors, using bidimensional or tridimensional analysis [38][39][40]. Texture analysis over a pulmonary nodule slice was performed in a previous work [41].…”

Section: D Texture Analysismentioning

confidence: 99%

Cloud-Based NoSQL Open Database of Pulmonary Nodules for Computer-Aided Lung Cancer Diagnosis and Reproducible Research

2016

View full text Add to dashboard Cite

Lung cancer is the leading cause of cancer-related deaths in the world, and its main manifestation is pulmonary nodules. Detection and classification of pulmonary nodules are challenging tasks that must be done by qualified specialists, but image interpretation errors make those tasks difficult. In order to aid radiologists on those hard tasks, it is important to integrate the computer-based tools with the lesion detection, pathology diagnosis, and image interpretation processes. However, computer-aided diagnosis research faces the problem of not having enough shared medical reference data for the development, testing, and evaluation of computational methods for diagnosis. In order to minimize this problem, this paper presents a public nonrelational document-oriented cloud-based database of pulmonary nodules characterized by 3D texture attributes, identified by experienced radiologists and classified in nine different subjective characteristics by the same specialists. Our goal with the development of this database is to improve computer-aided lung cancer diagnosis and pulmonary nodule detection and classification research

show abstract

Section: D Texture Analysismentioning

confidence: 99%

Cloud-Based NoSQL Open Database of Pulmonary Nodules for Computer-Aided Lung Cancer Diagnosis and Reproducible Research

2016

View full text Add to dashboard Cite

show abstract

“…Therefore, many attempts have been made to develop computer-aided diagnosis (CAD) systems for automatic discrimination. [3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] Conventional CAD systems first use classical image processing techniques, such as morphologic operators, 3-5 region growing, 6 energy optimization, 7,8 and statistical learning, 9,10 to segment a region of interest (ROI) that includes the nodule. Then, handcrafted features are extracted from the ROI, which are then fed to a classifier for nodule classification.…”

Section: Introductionmentioning

confidence: 99%

Feature‐shared adaptive‐boost deep learning for invasiveness classification of pulmonary subsolid nodules in CT images

Wang

Chen

et al. 2020

Medical Physics

View full text Add to dashboard Cite

Purpose: In clinical practice, invasiveness is an important reference indicator for differentiating the malignant degree of subsolid pulmonary nodules. These nodules can be classified as atypical adenomatous hyperplasia (AAH), adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), or invasive adenocarcinoma (IAC). The automatic determination of a nodule's invasiveness based on chest CT scans can guide treatment planning. However, it is challenging, owing to the insufficiency of training data and their interclass similarity and intraclass variation. To address these challenges, we propose a two-stage deep learning strategy for this task: prior-feature learning followed by adaptive-boost deep learning. Methods: The adaptive-boost deep learning is proposed to train a strong classifier for invasiveness classification of subsolid nodules in chest CT images, using multiple 3D convolutional neural network (CNN)-based weak classifiers. Because ensembles of multiple deep 3D CNN models have a huge number of parameters and require large computing resources along with more training and testing time, the prior-feature learning is proposed to reduce the computations by sharing the CNN layers between all weak classifiers. Using this strategy, all weak classifiers can be integrated into a single network. Results: Tenfold cross validation of binary classification was conducted on a total of 1357 nodules, including 765 noninvasive (AAH and AIS) and 592 invasive nodules (MIA and IAC). Ablation experimental results indicated that the proposed binary classifier achieved an accuracy of 73:4% AE 1:4 with an AUC of 81.3% AE 2:2. These results are superior compared to those achieved by three experienced chest imaging specialists who achieved an accuracy of 69:1% , 69:3% , and 67:9% , respectively. About 200 additional nodules were also collected. These nodules covered 50 cases for each category (AAH, AIS, MIA, and IAC, respectively). Both binary and multiple classifications were performed on these data and the results demonstrated that the proposed method definitely achieves better performance than the performance achieved by nonensemble deep learning methods. Conclusions: It can be concluded that the proposed adaptive-boost deep learning can significantly improve the performance of invasiveness classification of pulmonary subsolid nodules in CT images, while the prior-feature learning can significantly reduce the total size of deep models. The promising results on clinical data show that the trained models can be used as an effective lung cancer screening tool in hospitals. Moreover, the proposed strategy can be easily extended to other similar classification tasks in 3D medical images.

show abstract

“…Although a large number of studies have been conducted to develop CADx schemes of lung nodules, most of these CADx schemes were built and evaluated by using lung nodules with only suspicious assessment rating scores provided by the radiologists, such as using the Lung Image Database Consortium and Image Database Resource Initiative (LIDC/IDRI) dataset . Thus, these CADx schemes face a common issue that their training or testing nodules lack the biopsy‐confirmed gold standards …”

Section: Introductionmentioning

confidence: 99%

“…23 Thus, these CADx schemes face a common issue that their training or testing nodules lack the biopsy-confirmed gold standards. 24,25 In order to overcome the above limitations, the motivation or hypothesis of this study is that (a) CADx scheme should be trained and tested using the lung nodules with biopsy or pathology-confirmed results, and (b) the QI features and serum biomarkers may contain complementary information to classify between malignant and benign lung nodules. To test our study hypothesis, we built two lung cancer CADx schemes by using QI features and five serum biomarkers of biopsy-confirmed pulmonary nodules, respectively, and then investigated the feasibility of applying an information-fusion method to further improve the performance of CADx scheme.…”

Section: Introductionmentioning

confidence: 99%

Fusion of quantitative imaging features and serum biomarkers to improve performance of computer‐aided diagnosis scheme for lung cancer: A preliminary study

et al. 2018

View full text Add to dashboard Cite

Objectives To develop and test a new multifeature‐based computer‐aided diagnosis (CADx) scheme of lung cancer by fusing quantitative imaging (QI) features and serum biomarkers to improve CADx performance in classifying between malignant and benign pulmonary nodules. Methods First, a dataset involving 173 patients was retrospectively assembled, which includes computed tomography (CT) images and five serum biomarkers extracted from blood samples. Second, a CADx scheme using a four‐step–based semiautomatic segmentation method was applied to segment the targeted lung nodules, and compute 78 QI features from each segmented nodule from CT images. Third, two support vector machine (SVM) classifiers were built using QI features and serum biomarkers, respectively. SVM classifiers were trained and tested using the overall dataset with a Relief feature selection method, a synthetic minority oversampling technique and a leave‐one‐case‐out validation method. Finally, to further improve CADx performance, an information‐fusion method was used to combine the prediction scores generated by two SVM classifiers. Results Areas under receiver operating characteristic curves (AUC) generated by QI feature and serum biomarker‐based SVMs were 0.81 ± 0.03 and 0.69 ± 0.05, respectively. Using an optimal weighted fusion method to combine prediction scores generated by two SVMs, AUC value significantly increased to 0.85 ± 0.03 (P < 0.05). Conclusions This study demonstrates (a) higher CADx performance by using QI features than using the serum biomarkers and (b) feasibility of further improving CADx performance by fusion of QI features and serum biomarkers, which indicates that QI features and serum biomarkers contain the complementary classification information.

show abstract

A Combination of Shape and Texture Features for Classification of Pulmonary Nodules in Lung CT Images

Cited by 172 publications

References 27 publications

Cloud-Based NoSQL Open Database of Pulmonary Nodules for Computer-Aided Lung Cancer Diagnosis and Reproducible Research

Cloud-Based NoSQL Open Database of Pulmonary Nodules for Computer-Aided Lung Cancer Diagnosis and Reproducible Research

Feature‐shared adaptive‐boost deep learning for invasiveness classification of pulmonary subsolid nodules in CT images

Fusion of quantitative imaging features and serum biomarkers to improve performance of computer‐aided diagnosis scheme for lung cancer: A preliminary study

Contact Info

Product

Resources

About