Differential geometry-based techniques for characterization of boundary roughness of pulmonary nodules in CT images

Pretreatment Tumor Volume and Tumor Sphericity as Prognostic Factors in Patients with Oral Cavity Squamous Cell Carcinoma: A Prospective Clinical Study in 95 Patients

The prognostic impact of tumor volume and tumor sphericity was analyzed in 95 patients affected by oral cancer. The pre-operative computed tomography (CT) scans were used to segment the tumor mass with threshold tools, obtaining the corresponding volume and sphericity. Events of recurrence and tumor-related death were detected for each patient. The mean follow-up time was 31 months. A p-value of 0.05 was adopted. Mean tumor volume resulted higher in patients with recurrence or tumor-related death at the Student's t-test (respectively, 19.8 cm3 vs. 11.1 cm3, p = 0.03; 23.3 cm3 vs. 11.7 cm3, p = 0.02). Mean tumor sphericity was higher in disease-free patients (0.65 vs. 0.59, p = 0.04). Recurrence-free survival and disease-specific survival were greater for patients with a tumor volume inferior to the cut-off values of 21.1 cm3 (72 vs. 21 months, p < 0.01) and 22.4 cm3 (85 vs. 32 months, p < 0.01). Recurrence-free survival and disease-specific survival were higher for patients with a tumor sphericity superior to the cut-off value of 0.57 (respectively, 49 vs. 33 months, p < 0.01; 56 vs. 51 months, p = 0.01). To conclude, tumor volume and sphericity, three-dimensional parameters, could add useful information for better stratification of prognosis in oral cancer.

CIRDataset: A large-scale Dataset for Clinically-Interpretable lung nodule Radiomics and malignancy prediction

Spiculations/lobulations, sharp/curved spikes on the surface of lung nodules, are good predictors of lung cancer malignancy and hence, are routinely assessed and reported by radiologists as part of the standardized Lung-RADS clinical scoring criteria. Given the 3D geometry of the nodule and 2D slice-by-slice assessment by radiologists, manual spiculation/lobulation annotation is a tedious task and thus no public datasets exist to date for probing the importance of these clinically-reported features in the SOTA malignancy prediction algorithms. As part of this paper, we release a large-scale Clinically-Interpretable Radiomics Dataset, CIRDataset, containing 956 radiologist QA/QC'ed spiculation/lobulation annotations on segmented lung nodules from two public datasets, LIDC-IDRI (N=883) and LUNGx (N=73). We also present an end-to-end deep learning model based on multi-class Voxel2Mesh extension to segment nodules (while preserving spikes), classify spikes (sharp/spiculation and curved/lobulation), and perform malignancy prediction. Previous methods have performed malignancy prediction for LIDC and LUNGx datasets but without robust attribution to any clinically reported/actionable features (due to known hyperparameter sensitivity issues with general attribution schemes). With the release of this comprehensively-annotated CIRDataset and end-to-end deep learning baseline, we hope that malignancy prediction methods can validate their explanations, benchmark against our baseline, and provide clinically-actionable insights. Dataset, code, pretrained models, and docker containers are available at https://github.com/nadeemlab/CIR.

Reproducible and Interpretable Spiculation Quantification for Lung Cancer Screening

Spiculations are important predictors of lung cancer malignancy, which are spikes on the surface of the pulmonary nodules. In this study, we proposed an interpretable and parameter-free technique to quantify the spiculation using area distortion metric obtained by the conformal (angle-preserving) spherical parameterization. We exploit the insight that for an angle-preserved spherical mapping of a given nodule, the corresponding negative area distortion precisely characterizes the spiculations on that nodule. We introduced novel spiculation scores based on the area distortion metric and spiculation measures. We also semi-automatically segment lung nodule (for reproducibility) as well as vessel and wall attachment to differentiate the real spiculations from lobulation and attachment. A simple pathological malignancy prediction model is also introduced. We used the publicly-available LIDC-IDRI dataset pathologists (strong-label) and radiologists (weak-label) ratings to train and test radiomics models containing this feature, and then externally validate the models. We achieved AUC = 0.80 and 0.76, respectively, with the models trained on the 811 weakly-labeled LIDC datasets and tested on the 72 strongly-labeled LIDC and 73 LUNGx datasets; the previous best model for LUNGx had AUC = 0.68. The number-of-spiculations feature was found to be highly correlated (Spearman's rank correlation coefficient ρ=0.44) with the radiologists' spiculation score. We developed a reproducible and interpretable, parameter-free technique for quantifying spiculations on nodules. The spiculation quantification measures was then applied to the radiomics framework for pathological malignancy prediction with reproducible semi-automatic segmentation of nodule. Using our interpretable features (size, attachment, spiculation, lobulation), we were able to achieve higher performance than previous models. In the future, we will exhaustively test our model for lung cancer screening in the clinic.

Spiculation Sign Recognition in a Pulmonary Nodule Based on Spiking Neural P Systems

The spiculation sign is one of the main signs to distinguish benign and malignant pulmonary nodules. In order to effectively extract the image feature of a pulmonary nodule for the spiculation sign distinguishment, a new spiculation sign recognition model is proposed based on the doctors' diagnosis process of pulmonary nodules. A maximum density projection model is established to fuse the local three-dimensional information into the two-dimensional image. The complete boundary of a pulmonary nodule is extracted by the improved Snake model, which can take full advantage of the parallel calculation of the Spike Neural P Systems to build a new neural network structure. In this paper, our experiments show that the proposed algorithm can accurately extract the boundary of a pulmonary nodule and effectively improve the recognition rate of the spiculation sign.

The Robustness of the Gray Level Co-Occurrence Matrices and X-Ray Computed Tomography Method for the Quantification of 3D Mineral Texture

Mineral textural quantification methods have become critical in both geosciences and mineral processing as mineral texture is a critical factor contributing to ore variability. However, the lack of objective mineral texture classification has made quantification difficult. The aim of this study is therefore to investigate the robustness of applying the gray level co-occurrence matrices (GLCM) to 3-dimensional (3D) gray scale images measured by X-ray computed tomography (XCT) for the quantification of mineral texture in 3D. The data quality of the GLCM outputs like statistics, heat maps and histograms in response to changes in XCT conditions such as artefacts, resolution, and calibration was tested. The response of the GLCM outputs with respect to different mineral texture types with anisotropic features and inter-sample variability was also explored. The methodology included testing core sizes of 26, 19, 14, and 6 mm diameter. Calibration was tested using copper and tungsten wires. The study demonstrated the versatility of the method for different sample types. Inter-sample calibration and optimal scanning conditions (quality and integrity) were also demonstrated, and a basic link between the 3D GLCM statistical descriptors with the mineral texture features of rocks was established. The 3D mineral texture method can potentially bypass the XCT segmentation process for direct automation of 3D mineral texture information.

CT-based radiomics for prediction of histologic subtype and metastatic disease in primary malignant lung neoplasms

PURPOSE

As some of the most important factors for treatment decision of lung cancer (which is the deadliest neoplasm) are staging and histology, this work aimed to associate quantitative contrast-enhanced computed tomography (CT) features from malignant lung tumors with distant and nodal metastases (according to clinical TNM staging) and histopathology (according to biopsy and surgical resection) using radiomics assessment.

METHODS

A local cohort of 85 patients were retrospectively (2010-2017) analyzed after approval by the institutional research review board. CT images acquired with the same protocol were semiautomatically segmented by a volumetric segmentation method. Tumors were characterized by quantitative CT features of shape, first-order, second-order, and higher-order textures. Statistical and machine learning analyses assessed the features individually and combined with clinical data.

RESULTS

Univariate and multivariate analyses identified 40, 2003, and 45 quantitative features associated with distant metastasis, nodal metastasis, and histopathology (adenocarcinoma and squamous cell carcinoma), respectively. A machine learning model yielded the highest areas under the receiver operating characteristic curves of 0.92, 0.84, and 0.88 to predict the same previous patterns.

CONCLUSION

Several radiomic features (including wavelet energies, information measures of correlation and maximum probability from co-occurrence matrix, busyness from neighborhood intensity-difference matrix, directionalities from Tamura's texture, and fractal dimension estimation) significantly associated with distant metastasis, nodal metastasis, and histology were discovered in this work, presenting great potential as imaging biomarkers for pathological diagnosis and target therapy decision.

Content-Based Image Retrieval System for Pulmonary Nodules Using Optimal Feature Sets and Class Membership-Based Retrieval

Lung cancer manifests itself in the form of lung nodules, the diagnosis of which is essential to plan the treatment. Automated retrieval of nodule cases will assist the budding radiologists in self-learning and differential diagnosis. This paper presents a content-based image retrieval (CBIR) system for lung nodules using optimal feature sets and learning to enhance the performance of retrieval. The classifiers with more features suffer from the curse of dimensionality. Like classification schemes, we found that the optimal feature set selected using the minimal-redundancy-maximal-relevance (mRMR) feature selection technique improves the precision performance of simple distance-based retrieval (SDR). The performance of the classifier is always superior to SDR, which leans researchers towards conventional classifier-based retrieval (CCBR). While CCBR improves the average precision and provides 100% precision for correct classification, it fails for misclassification leading to zero retrieval precision. The class membership-based retrieval (CMR) is found to bridge this gap for texture-based retrieval. Here, CMR is proposed for nodule retrieval using shape-, margin-, and texture-based features. It is found again that optimal feature set is important for the classifier used in CMR as well as for the feature set used for retrieval, which may lead to different feature sets. The proposed system is evaluated using two independent databases from two continents: a public database LIDC/IDRI and a private database PGIMER-IITKGP, using three distance metrics, i.e., Canberra, City block, and Euclidean. The proposed CMR-based retrieval system with optimal feature sets performs better than CCBR and SDR with optimal features in terms of average precision. Apart from average precision and standard deviation of precision, the fraction of queries with zero precision retrieval is also measured.

Identification of Nonaggressive Pulmonary Nodules Using an Optimized Scoring System

Supplemental Digital Content is available in the text.

Purpose:

The purpose of this study was to define the optimal scoring method for identifying benign intrapulmonary lymph nodes.

Materials and Methods:

Subjects for this study were selected from the COPDGene study, a large multicenter longitudinal observational cohort study. A retrospective case-control analysis was performed using identified nodules on a subset of 377 patients who demonstrated 765 pulmonary nodules on their baseline computed tomography (CT) study. Nodule characteristics of 636 benign nodules (which resolved or showed <20% growth rate at 5 y follow-up) were compared with 51 nodules that occurred in the same lobe as a reported malignancy. Two radiologists scored each pulmonary nodule on the basis of intrapulmonary lymph node characteristics. A simple scoring strategy weighing all characteristics equally was compared with an optimized scoring strategy that weighed characteristics on the basis of their relative importance in identifying benign pulmonary nodules.

Results:

A total of 479 of 636 benign pulmonary nodules had the majority of lymph node characteristics, whereas only 1 subpleural nodule with the majority of lymph node characteristics appeared to be malignant. Only 279 of 479 (58%) of benign pulmonary nodules with the majority of lymph node characteristics were intrafissural or subpleural. The optimized scoring strategy showed improved performance compared with the simple scoring strategy with average area under the curve of 0.80 versus 0.55. Optimized cutoff scores showed negative likelihood values for both readers of <0.2. A simulation showed a potential reduction in CT utilization of up to 36% for Fleischner criteria and up to 5% for LUNG-RADS.

Conclusions:

Nodules with the majority of lymph node characteristics, regardless of location, are likely benign, and weighing certain lymph node characteristics greater than others can improve overall performance. Given the potential to reduce CT utilization, lymph node characteristics should be considered when recommending appropriate follow-up.

Integrating quantitative morphological and intratumoural textural characteristics in FDG-PET for the prediction of prognosis in pharynx squamous cell carcinoma patients

AIM

To assess potential prognostic factors in pharynx squamous cell carcinoma (SCC) patients by quantitative morphological and intratumoural characteristics obtained by 2-[¹⁸F]-fluoro-2-deoxy-d-glucose positron-emission tomography/computed tomography (FDG-PET/CT).

MATERIALS AND METHODS

The cases of 54 patients with pharynx SCC who underwent chemoradiation therapy were analysed retrospectively. Using their FDG-PET data, the quantitative morphological and intratumoural characteristics of 14 parameters were calculated. The progression-free survival (PFS) and overall survival (OS) information was obtained from patient medical records. Univariate and multivariate analyses were performed to assess the 14 quantitative parameters as well as the T-stage, N-stage, and tumour location data for their relation to PFS and OS. When an independent predictor was suggested in the multivariate analysis, the parameter was further assessed using the Kaplan-Meier method.

RESULTS

In the assessment of PFS, the univariate and multivariate analyses indicated the following as independent predictors: the texture parameter of homogeneity and the morphological parameter of sphericity. In the Kaplan-Meier analysis, the PFS rate was significantly improved in the patients who had both a higher value of homogeneity (p=0.01) and a higher value of sphericity (p=0.002). With the combined use of homogeneity and sphericity, the patients with different PFS rates could be divided more clearly.

CONCLUSION

The quantitative parameters of homogeneity and sphericity obtained by FDG-PET can be useful for the prediction of the PFS of pharynx SCC patients, especially when used in combination.

Semi-quantitative analysis of pre-treatment morphological and intratumoral characteristics using 18F-fluorodeoxyglucose positron-emission tomography as predictors of treatment outcome in nasal and paranasal squamous cell carcinoma

Background

To investigate the utility of quantitative morphological and intratumoral characteristics obtained by ¹⁸F-fluorodeoxyglucose positron-emission tomography/computed tomography (FDG-PET/CT) for the prediction of treatment outcome in patients with nasal or paranasal cavity squamous cell carcinoma (SCC).

Methods

Twenty-four patients with nasal or paranasal cavity SCC who received curative non-surgical therapy (a combination of super-selective arterial cisplatin infusion and radiotherapy) were retrospectively analyzed. From pre-treatment FDG-PET data, a total of 13 parameters of quantitative morphological characteristics (tumor volume, surface area and sphericity), intratumoral characteristics (the maximum and mean standard uptake value, three intratumoral histogram and four textural parameters) and total lesion glycolysis (TLG) were respectively calculated. Information regarding the treatment outcome was determined from the histological diagnosis or clinical follow-up. Each of the 13 quantitative parameters as well as T- and N-stage was assessed for its relation to treatment outcome of local control or failure.

Results

In univariate analysis, significant differences in surface area and sphericity between the local control and failure groups were observed. The receiver operating characteristic (ROC) curve analysis showed that sphericity had the highest accuracy of 0.88. In the multivariate analysis, sphericity was revealed as an independent predictor of the local control or failure.

Conclusions

The quantitative parameters of sphericity are useful to predict the treatment outcome in patients with nasal or paranasal SCC.

Content-Based Image Retrieval System for Pulmonary Nodules: Assisting Radiologists in Self-Learning and Diagnosis of Lung Cancer

Visual information of similar nodules could assist the budding radiologists in self-learning. This paper presents a content-based image retrieval (CBIR) system for pulmonary nodules, observed in lung CT images. The reported CBIR systems of pulmonary nodules cannot be put into practice as radiologists need to draw the boundary of nodules during query formation and feature database creation. In the proposed retrieval system, the pulmonary nodules are segmented using a semi-automated technique, which requires a seed point on the nodule from the end-user. The involvement of radiologists in feature database creation is also reduced, as only a seed point is expected from radiologists instead of manual delineation of the boundary of the nodules. The performance of the retrieval system depends on the accuracy of the segmentation technique. Several 3D features are explored to improve the performance of the proposed retrieval system. A set of relevant shape and texture features are considered for efficient representation of the nodules in the feature space. The proposed CBIR system is evaluated for three configurations such as configuration-1 (composite rank of malignancy "1","2" as benign and "4","5" as malignant), configuration-2 (composite rank of malignancy "1","2", "3" as benign and "4","5" as malignant), and configuration-3 (composite rank of malignancy "1","2" as benign and "3","4","5" as malignant). Considering top 5 retrieved nodules and Euclidean distance metric, the precision achieved by the proposed method for configuration-1, configuration-2, and configuration-3 are 82.14, 75.91, and 74.27 %, respectively. The performance of the proposed CBIR system is close to the most recent technique, which is dependent on radiologists for manual segmentation of nodules. A computer-aided diagnosis (CAD) system is also developed based on CBIR paradigm. Performance of the proposed CBIR-based CAD system is close to performance of the CAD system using support vector machine.

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

Classification of malignant and benign pulmonary nodules is important for further treatment plan. The present work focuses on the classification of benign and malignant pulmonary nodules using support vector machine. The pulmonary nodules are segmented using a semi-automated technique, which requires only a seed point from the end user. Several shape-based, margin-based, and texture-based features are computed to represent the pulmonary nodules. A set of relevant features is determined for the efficient representation of nodules in the feature space. The proposed classification scheme is validated on a data set of 891 nodules of Lung Image Database Consortium and Image Database Resource Initiative public database. The proposed classification scheme is evaluated for three configurations such as configuration 1 (composite rank of malignancy "1" and "2" as benign and "4" and "5" as malignant), configuration 2 (composite rank of malignancy "1","2", and "3" as benign and "4" and "5" as malignant), and configuration 3 (composite rank of malignancy "1" and "2" as benign and "3","4" and "5" as malignant). The performance of the classification is evaluated in terms of area (A z) under the receiver operating characteristic curve. The A z achieved by the proposed method for configuration-1, configuration-2, and configuration-3 are 0.9505, 0.8822, and 0.8488, respectively. The proposed method outperforms the most recent technique, which depends on the manual segmentation of pulmonary nodules by a trained radiologist.

Advances in intelligent diagnosis methods for pulmonary ground-glass opacity nodules

Pulmonary nodule is one of the important lesions of lung cancer, mainly divided into two categories of solid nodules and ground glass nodules. The improvement of diagnosis of lung cancer has significant clinical significance, which could be realized by machine learning techniques. At present, there have been a lot of researches focusing on solid nodules. But the research on ground glass nodules started late, and lacked research results. This paper summarizes the research progress of the method of intelligent diagnosis for pulmonary nodules since 2014. It is described in details from four aspects: nodular signs, data analysis methods, prediction models and system evaluation. This paper aims to provide the research material for researchers of the clinical diagnosis and intelligent analysis of lung cancer, and further improve the precision of pulmonary ground glass nodule diagnosis.

3D skeletonization feature based computer-aided detection system for pulmonary nodules in CT datasets

Pulmonary nodule detection has a significant impact on early diagnosis of lung cancer. To effectively detect pulmonary nodules from interferential vessels in chest CT datasets, this paper proposes a novel 3D skeletonization feature, named as voxels remove rate. Based on this feature, a computer-aided detection system is constructed to validate its performance. The system mainly consists of five stages. Firstly, the lung tissues are segmented by a global optimal active contour model, which can extract all structures (including juxta-pleural nodules) in the lung region. Secondly, thresholding, 3D binary morphological operations, and 3D connected components labeling are utilized to extract candidates of pulmonary nodules. Thirdly, combining the voxels remove rate with other nine existing 3D features (including gray features and shape features), the extracted candidates are characterized. Then, prior anatomical knowledge is utilized for preliminary screening of numerous invalid nodule candidates. Finally, false positives are reduced by support vector machine. Our system is evaluated on early stage lung cancer subjects obtained from the publicly available LIDC-IDRI database. The result shows the proposed 3D skeletonization feature is a useful indicator that efficiently differentiates lung nodules from the other suspicious structures. The computer-aided detection system based on this feature can detect various types of nodules, including solitary, juxta-pleural and juxta-vascular nodules.

Computerized margin and texture analyses for differentiating bacterial pneumonia and invasive mucinous adenocarcinoma presenting as consolidation

Radiologists have used margin characteristics based on routine visual analysis; however, the attenuation changes at the margin of the lesion on CT images have not been quantitatively assessed. We established a CT-based margin analysis method by comparing a target lesion with normal lung attenuation, drawing a slope to represent the attenuation changes. This approach was applied to patients with invasive mucinous adenocarcinoma (n = 40) or bacterial pneumonia (n = 30). Correlations among multiple regions of interest (ROIs) were obtained using intraclass correlation coefficient (ICC) values. CT visual assessment, margin and texture parameters were compared for differentiating the two disease entities. The attenuation and margin parameters in multiple ROIs showed excellent ICC values. Attenuation slopes obtained at the margins revealed a difference between invasive mucinous adenocarcinoma and pneumonia (P<0.001), and mucinous adenocarcinoma produced a sharply declining attenuation slope. On multivariable logistic regression analysis, pneumonia had an ill-defined margin (odds ratio (OR), 4.84; 95% confidence interval (CI), 1.26–18.52; P = 0.02), ground-glass opacity (OR, 8.55; 95% CI, 2.09–34.95; P = 0.003), and gradually declining attenuation at the margin (OR, 12.63; 95% CI, 2.77–57.51, P = 0.001). CT-based margin analysis method has a potential to act as an imaging parameter for differentiating invasive mucinous adenocarcinoma and bacterial pneumonia.