Classification of Small Lesions in Breast MRI: Evaluating The Role of Dynamically Extracted Texture Features Through Feature Selection

3D DCE-MRI Radiomic Analysis for Malignant Lesion Prediction in Breast Cancer Patients

RATIONALE AND OBJECTIVES

To develop and validate a radiomic model, with radiomic features extracted from breast Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) from a 1.5T scanner, for predicting the malignancy of masses with enhancement. Images were acquired using an 8-channel breast coil in the axial plane. The rationale behind this study is to show the feasibility of a radiomics-powered model that could be integrated into the clinical practice by exploiting only standard-of-care DCE-MRI with the goal of reducing the required image pre-processing (ie, normalization and quantitative imaging map generation).

MATERIALS AND METHODS

107 radiomic features were extracted from a manually annotated dataset of 111 patients, which was split into discovery and test sets. A feature calibration and pre-processing step was performed to find only robust non-redundant features. An in-depth discovery analysis was performed to define a predictive model: for this purpose, a Support Vector Machine (SVM) was trained in a nested 5-fold cross-validation scheme, by exploiting several unsupervised feature selection methods. The predictive model performance was evaluated in terms of Area Under the Receiver Operating Characteristic (AUROC), specificity, sensitivity, PPV and NPV. The test was performed on unseen held-out data.

RESULTS

The model combining Unsupervised Discriminative Feature Selection (UDFS) and SVMs on average achieved the best performance on the blinded test set: AUROC = 0.725±0.091, sensitivity = 0.709±0.176, specificity = 0.741±0.114, PPV = 0.72±0.093, and NPV = 0.75±0.114.

CONCLUSION

In this study, we built a radiomic predictive model based on breast DCE-MRI, using only the strongest enhancement phase, with promising results in terms of accuracy and specificity in the differentiation of malignant from benign breast lesions.

Characterization of breast tumors using machine learning based upon multiparametric magnetic resonance imaging features

Magnetic resonance imaging (MRI) is playing an important role in the classification of breast tumors. MRI can be used to obtain multiparametric (mp) information, such as structural, hemodynamic, and physiological information. Quantitative analysis of mp-MRI data has shown potential in improving the accuracy of breast tumor classification. In general, a large set of quantitative and texture features can be generated depending upon the type of methodology used. A suitable combination of selected quantitative and texture features can further improve the accuracy of tumor classification. Machine learning (ML) classifiers based upon features derived from MRI data have shown potential in tumor classification. There is a need for further research studies on selecting an appropriate combination of features and evaluating the performance of different ML classifiers for accurate classification of breast tumors. The objective of the current study was to develop and optimize an ML framework based upon mp-MRI features for the characterization of breast tumors (malignant vs. benign and low- vs. high-grade). This study included the breast mp-MRI data of 60 female patients with histopathology results. A total of 128 features were extracted from the mp-MRI tumor data followed by features selection. Five ML classifiers were evaluated for tumor classification using 10-fold crossvalidation with 10 repetitions. The support vector machine (SVM) classifier based on optimum features selected using a wrapper method with an adaptive boosting (AdaBoost) technique provided the highest sensitivity (0.96 ± 0.03), specificity (0.92 ± 0.09), and accuracy (94% ± 2.91%) in the classification of malignant versus benign tumors. This method also provided the highest sensitivity (0.94 ± 0.07), specificity (0.80 ± 0.05), and accuracy (90% ± 5.48%) in the classification of low- versus high-grade tumors. These findings suggest that the SVM classifier outperformed other ML methods in the binary classification of breast tumors.

Association between breast cancer's prognostic factors and 3D textural features of non-contrast-enhanced T1 weighted breast MRI

OBJECTIVES

The aim of this exploratory study was to evaluate whether three-dimensional texture analysis (3D-TA) features of non-contrast-enhanced T₁ weighted MRI associate with traditional prognostic factors and disease-free survival (DFS) of breast cancer.

METHODS

3D-T₁ weighted images from 78 patients with 81 malignant histopathologically verified breast lesions were retrospectively analysed using standard-size volumes of interest. Grey-level co-occurrence matrix (GLCM)-based features were selected for statistical analysis. In statistics the Mann-Whitney U and the Kruskal-Wallis tests, the Cox proportional hazards model and the Kaplan-Meier method were used.

RESULTS

Tumours with higher histological grade were significantly associated with higher contrast (1 voxel: p = 0.033, 2 voxels: p = 0.036). All the entropy parameters showed significant correlation with tumour grade (p = 0.015-0.050) but there were no statistically significant associations between other TA parameters and tumour grade. The Nottingham Prognostic Index (NPI) was correlated with contrast and sum entropy parameters. A higher sum variance TA parameter was a significant predictor of shorter DFS.

CONCLUSION

Texture parameters, assessed by 3D-TA from non-enhanced T₁ weighted images, indicate tumour heterogeneity but have limited independent prognostic value. However, they are associated with tumour grade, NPI, and DFS. These parameters could be used as an adjunct to contrast-enhanced TA parameters.

ADVANCES IN KNOWLEDGE

3D-TA of non-contrast enhanced T₁ weighted breast MRI associates with tumour grade, NPI, and DFS. The use of non-contrast 3D-TA parameters in adjunct with contrast-enhanced 3D-TA parameters warrants further research.

A Pipelined Tracer-Aware Approach for Lesion Segmentation in Breast DCE-MRI

The recent spread of Deep Learning (DL) in medical imaging is pushing researchers to explore its suitability for lesion segmentation in Dynamic Contrast-Enhanced Magnetic-Resonance Imaging (DCE-MRI), a complementary imaging procedure increasingly used in breast-cancer analysis. Despite some promising proposed solutions, we argue that a "naive" use of DL may have limited effectiveness as the presence of a contrast agent results in the acquisition of multimodal 4D images requiring thorough processing before training a DL model. We thus propose a pipelined approach where each stage is intended to deal with or to leverage a peculiar characteristic of breast DCE-MRI data: the use of a breast-masking pre-processing to remove non-breast tissues; the use of Three-Time-Points (3TP) slices to effectively highlight contrast agent time course; the application of a motion-correction technique to deal with patient involuntary movements; the leverage of a modified U-Net architecture tailored on the problem; and the introduction of a new "Eras/Epochs" training strategy to handle the unbalanced dataset while performing a strong data augmentation. We compared our pipelined solution against some literature works. The results show that our approach outperforms the competitors by a large margin (+9.13% over our previous solution) while also showing a higher generalization ability.

Vascularity and Dynamic Contrast-Enhanced Breast Magnetic Resonance Imaging

Angiogenesis is a key step in the initiation and progression of an invasive breast cancer. High microvessel density by morphological characterization predicts metastasis and poor survival in women with invasive breast cancers. However, morphologic characterization is subject to variability and only can evaluate a limited portion of an invasive breast cancer. Consequently, breast Magnetic Resonance Imaging (MRI) is currently being evaluated to assess vascularity. Recently, through the new field of radiomics, dynamic contrast enhanced (DCE)-MRI is being used to evaluate vascular density, vascular morphology, and detection of aggressive breast cancer biology. While DCE-MRI is a highly sensitive tool, there are specific features that limit computational evaluation of blood vessels. These include (1) DCE-MRI evaluates gadolinium contrast and does not directly evaluate biology, (2) the resolution of DCE-MRI is insufficient for imaging small blood vessels, and (3) DCE-MRI images are very difficult to co-register. Here we review computational approaches for detection and analysis of blood vessels in DCE-MRI images and present some of the strategies we have developed for co-registry of DCE-MRI images and early detection of vascularization.

Diagnostic accuracy of MRI textural analysis in the classification of breast tumors

OBJECTIVE

To investigate whether textural analysis (TA) of MRI heterogeneity may play a role in the clinical assessment and classification of breast tumors.

MATERIALS AND METHODS

For this retrospective study, patients with breast masses ≥1 cm on contrast-enhanced MRI were obtained in 69 women (mean age: 51 years; range 21-78 years) with 77 masses (38 benign, 39 malignant) from 2006 to 2018. The selected single slice sagittal peak post-contrast T1-weighted image was analyzed with commercially available TA software [TexRAD Ltd., UK]. Eight histogram TA parameters were evaluated at various spatial scaling factors (SSF) including mean pixel intensity, standard deviation of the pixel histogram (SD), entropy, mean of the positive pixels (MPP), skewness, kurtosis, sigma, and Tx_sigma. Additional statistical tests were used to determine their predictiveness.

RESULTS

Entropy showed a significant difference between benign and malignant tumors at all textural scales (p < 0.0001) and kurtosis was significant at SSF = 0-5 (p = 0.0026-0.0241). The single best predictor was entropy at SSF = 4 with AUC = 0.80, giving a sensitivity of 95% and specificity of 53%. An AUC of 0.91 was found using a model combining entropy with sigma, which yielded better performance with a sensitivity of 92% and specificity of 79%.

CONCLUSION

TA of breast masses has the potential to assist radiologists in categorizing tumors as benign or malignant on MRI. Measurements of entropy, kurtosis, and entropy combined with sigma may provide the best predictability.

Decision support system for breast lesions via dynamic contrast enhanced magnetic resonance imaging

The presented study aims to design a computer-aided detection and diagnosis system for breast dynamic contrast enhanced magnetic resonance imaging. In the proposed system, the segmentation task is performed in two stages. The first stage is called breast region segmentation in which adaptive noise filtering, local adaptive thresholding, connected component analysis, integral of horizontal projection, and breast region of interest detection algorithms are applied to the breast images consecutively. The second stage of segmentation is breast lesion detection that consists of 32-class Otsu thresholding and Markov random field techniques. Histogram, gray level co-occurrence matrix and neighboring gray tone difference matrix based feature extraction, Fisher score based feature selection and, tenfold and leave-one-out cross-validation steps are carried out after segmentation to increase the reliability of the designed system while decreasing the computational time. Finally, support vector machines, k- nearest neighbor, and artificial neural network classifiers are performed to separate the breast lesions as benign and malignant. The average accuracy, sensitivity, specificity, and positive predictive values of each classifier are calculated and the best results are compared with the existing similar studies. According to the achieved results, the proposed decision support system for breast lesion segmentation distinguishes the breast lesions with 86%, 100%, 67%, and 85% accuracy, sensitivity, specificity, and positive predictive values, respectively. These results show that the proposed system can be used to support the radiologists during a breast cancer diagnosis.

Breast dynamic contrast-enhanced-magnetic resonance imaging and radiomics: State of art

Breast cancer represents the most common malignancy in women, being one of the most frequent cause of cancer-related mortality. Ultrasound, mammography, and magnetic resonance imaging (MRI) play a pivotal role in the diagnosis of breast lesions, with different levels of accuracy. Particularly, dynamic contrast-enhanced MRI has shown high diagnostic value in detecting multifocal, multicentric, or contralateral breast cancers. Radiomics is emerging as a promising tool for quantitative tumor evaluation, allowing the extraction of additional quantitative data from radiological imaging acquired with different modalities. Radiomics analysis may provide novel information through the quantification of lesions heterogeneity, that may be relevant in clinical practice for the characterization of breast lesions, prediction of tumor response to systemic therapies and evaluation of prognosis in patients with breast cancers. Several published studies have explored the value of radiomics with good-to-excellent diagnostic and prognostic performances for the evaluation of breast lesions. Particularly, the integrations of radiomics data with other clinical and histopathological parameters have demonstrated to improve the prediction of tumor aggressiveness with high accuracy and provided precise models that will help to guide clinical decisions and patients management. The purpose of this article in to describe the current application of radiomics in breast dynamic contrast-enhanced MRI.

Performance evaluation of texture analysis based on kinetic parametric maps from breast DCE-MRI in classifying benign from malignant lesions

BACKGROUND AND OBJECTIVES

To investigate the performance of texture analysis based on enhancement kinetic parametric maps derived from breast dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) in discriminating benign from malignant tumors.

METHODS

A total of 192 cases confirmed by histopathology were retrospectively selected from our Picture Archiving and Communication System, including 93 benign and 99 malignant tumors. Lesion areas were delineated semi-automatically, and six kinetic parametric maps reflecting the perfusion information were generated, namely the maximum slope of increase (MSI), slope of signal intensity (SI_slope ), initial percentage of peak enhancement (E_initial ), percentage of peak enhancement (E_peak ), early signal enhancement ratio (ESER), and second enhancement percentage (SEP) maps. A total of 286 texture features were extracted from those quantitative parametric maps. The Student t test or Mann-Whitney U test was used to select features that were statistically significantly different between the benign and malignant groups. A support vector machine was employed with a leave-one-out cross-validation method to establish the classification model. Classification performance was evaluated according to the receiver operating characteristic (ROC) theory.

RESULTS

The areas under ROC curves (AUCs) indicating the diagnostic performance were 0.925 for MSI, 0.854 for SI_slope , 0.756 for E_initial , 0.923 for E_peak , 0.871 for ESER and 0.881 for SEP. Significant differences in AUCs were found between E_initial vs MSI, E_initial vs E_peak and E_initial vs SEP (P < .05). There were no significant differences in other pairwise comparisons.

CONCLUSION

Texture analysis of the kinetic parametric maps derived from breast DCE-MRI can contribute to the discrimination between malignant and benign lesions. It can be considered as a supplementary tool for breast diagnosis.

Precision diagnostics based on machine learning-derived imaging signatures

The complexity of modern multi-parametric MRI has increasingly challenged conventional interpretations of such images. Machine learning has emerged as a powerful approach to integrating diverse and complex imaging data into signatures of diagnostic and predictive value. It has also allowed us to progress from group comparisons to imaging biomarkers that offer value on an individual basis. We review several directions of research around this topic, emphasizing the use of machine learning in personalized predictions of clinical outcome, in breaking down broad umbrella diagnostic categories into more detailed and precise subtypes, and in non-invasively estimating cancer molecular characteristics. These methods and studies contribute to the field of precision medicine, by introducing more specific diagnostic and predictive biomarkers of clinical outcome, therefore pointing to better matching of treatments to patients.

Radiomics Analysis on Contrast-Enhanced Spectral Mammography Images for Breast Cancer Diagnosis: A Pilot Study

Contrast-enhanced spectral mammography is one of the latest diagnostic tool for breast care; therefore, the literature is poor in radiomics image analysis useful to drive the development of automatic diagnostic support systems. In this work, we propose a preliminary exploratory analysis to evaluate the impact of different sets of textural features in the discrimination of benign and malignant breast lesions. The analysis is performed on 55 ROIs extracted from 51 patients referred to Istituto Tumori “Giovanni Paolo II” of Bari (Italy) from the breast cancer screening phase between March 2017 and June 2018. We extracted feature sets by calculating statistical measures on original ROIs, gradiented images, Haar decompositions of the same original ROIs, and on gray-level co-occurrence matrices of the each sub-ROI obtained by Haar transform. First, we evaluated the overall impact of each feature set on the diagnosis through a principal component analysis by training a support vector machine classifier. Then, in order to identify a sub-set for each set of features with higher diagnostic power, we developed a feature importance analysis by means of wrapper and embedded methods. Finally, we trained an SVM classifier on each sub-set of previously selected features to compare their classification performances with respect to those of the overall set. We found a sub-set of significant features extracted from the original ROIs with a diagnostic accuracy greater than 80 % . The features extracted from each sub-ROI decomposed by two levels of Haar transform were predictive only when they were all used without any selection, reaching the best mean accuracy of about 80 % . Moreover, most of the significant features calculated by HAAR decompositions and their GLCMs were extracted from recombined CESM images. Our pilot study suggested that textural features could provide complementary information about the characterization of breast lesions. In particular, we found a sub-set of significant features extracted from the original ROIs, gradiented ROI images, and GLCMs calculated from each sub-ROI previously decomposed by the Haar transform.

Characterization of Sub-1 cm Breast Lesions Using Radiomics Analysis

BACKGROUND

Small breast lesions are difficult to visually categorize due to the inherent lack of morphological and kinetic detail.

PURPOSE

To assess the efficacy of radiomics analysis in discriminating small benign and malignant lesions utilizing model free parameter maps.

STUDY TYPE

Retrospective, single center.

POPULATION

In all, 149 patients, with a total of 165 lesions scored as BI-RADS 4 or 5 on MRI, with an enhancing volume of <0.52 cm³ .

FIELD STRENGTH/SEQUENCE

Higher spatial resolution T₁ -weighted dynamic contrast-enhanced imaging with a temporal resolution of ~90 seconds performed at 3.0T.

ASSESSMENT

Parameter maps reflecting initial enhancement, overall enhancement, area under the enhancement curve, and washout were generated. Heterogeneity measures based on first-order statistics, gray level co-occurrence matrices, run length matrices, size zone matrices, and neighborhood gray tone difference matrices were calculated. Data were split into a training dataset (~75% of cases) and a test dataset (~25% of cases).

STATISTICAL TESTS

Comparison of medians was assessed using the nonparametric Mann-Whitney U-test. The Spearman rank correlation coefficient was utilized to determine significant correlations between individual features. Finally, a support vector machine was employed to build multiparametric predictive models.

RESULTS

Univariate analysis revealed significant differences between benign and malignant lesions for 58/133 calculated features (P < 0.05). Support vector machine analysis resulted in areas under the curve (AUCs) ranging from 0.75-0.81. High negative (>89%) and positive predictive values (>83%) were found for all models.

DATA CONCLUSION

Radiomics analysis of small contrast-enhancing breast lesions is of value. Texture features calculated from later timepoints on the enhancement curve appear to offer limited additional value when compared with features determined from initial enhancement for this patient cohort.

LEVEL OF EVIDENCE

4 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;50:1468-1477.

Artificial Intelligence for Breast MRI in 2008-2018: A Systematic Mapping Review

OBJECTIVE

The purpose of this study is to review literature from the past decade on applications of artificial intelligence (AI) to breast MRI.

MATERIALS AND METHODS

In June 2018, a systematic search of the literature was performed to identify articles on the use of AI in breast MRI. For each article identified, the surname of the first author, year of publication, journal of publication, Web of Science Core Collection journal category, country of affiliation of the first author, study design, dataset, study aim(s), AI methods used, and, when available, diagnostic performance were recorded.

RESULTS

Sixty-seven studies, 58 (87%) of which had a retrospective design, were analyzed. When journal categories were considered, 36% of articles were identified as being included in the radiology and imaging journal category. Contrast-enhanced sequences were used for most AI applications (n = 50; 75%) and, on occasion, were combined with other MRI sequences (n = 8; 12%). Four main clinical aims were addressed: breast lesion classification (n = 36; 54%), image processing (n = 14; 21%), prognostic imaging (n = 9; 13%), and response to neoadjuvant therapy (n = 8; 12%). Artificial neural networks, support vector machines, and clustering were the most frequently used algorithms, accounting for 66%. The performance achieved and the most frequently used techniques were then analyzed according to specific clinical aims. Supervised learning algorithms were primarily used for lesion characterization, with the AUC value from ROC analysis ranging from 0.74 to 0.98 (median, 0.87) and with that from prognostic imaging ranging from 0.62 to 0.88 (median, 0.80), whereas unsupervised learning was mainly used for image processing purposes.

CONCLUSION

Interest in the application of advanced AI methods to breast MRI is growing worldwide. Although this growth is encouraging, the current performance of AI applications in breast MRI means that such applications are still far from being incorporated into clinical practice.

Role of texture analysis in breast MRI as a cancer biomarker: A review

Breast cancer is a known heterogeneous disease. Current clinically utilized histopathologic biomarkers may undersample tumor heterogeneity, resulting in higher rates of misdiagnosis for breast cancer. MRI can provide a whole-tumor sampling of disease burden and is widely utilized in clinical care. Texture analysis can provide a localized description of breast cancer, with particular emphasis on quantifying breast lesion heterogeneity. The object of this review is to provide an overview of texture analysis applications towards breast cancer diagnosis, prognosis, and treatment response evaluation and review the role of image-based texture features as noninvasive prognostic and predictive biomarkers. Level of Evidence: 5 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;49:927-938.

Density estimation of grey-level co-occurrence matrices for image texture analysis

The Haralick texture features are common in the image analysis literature, partly because of their simplicity and because their values can be interpreted. It was recently observed that the Haralick texture features are very sensitive to the size of the GLCM that was used to compute them, which led to a new formulation that is invariant to the GLCM size. However, these new features still depend on the sample size used to compute the GLCM, i.e. the size of the input image region-of-interest (ROI). The purpose of this work was to investigate the performance of density estimation methods for approximating the GLCM and subsequently the corresponding invariant features. Three density estimation methods were evaluated, namely a piece-wise constant distribution, the Parzen-windows method, and the Gaussian mixture model. The methods were evaluated on 29 different image textures and 20 invariant Haralick texture features as well as a wide range of different ROI sizes. The results indicate that there are two types of features: those that have a clear minimum error for a particular GLCM size for each ROI size, and those whose error decreases monotonically with increased GLCM size. For the first type of features, the Gaussian mixture model gave the smallest errors, and in particular for small ROI sizes (less than about [Formula: see text]). In conclusion, the Gaussian mixture model is the preferred method for the first type of features (in particular for small ROIs). For the second type of features, simply using a large GLCM size is preferred.

Radiomics strategy for glioma grading using texture features from multiparametric MRI

BACKGROUND

Accurate glioma grading plays an important role in the clinical management of patients and is also the basis of molecular stratification nowadays.

PURPOSE/HYPOTHESIS

To verify the superiority of radiomics features extracted from multiparametric MRI to glioma grading and evaluate the grading potential of different MRI sequences or parametric maps.

STUDY TYPE

Retrospective; radiomics.

POPULATION

A total of 153 patients including 42, 33, and 78 patients with Grades II, III, and IV gliomas, respectively.

FIELD STRENGTH/SEQUENCE

3.0T MRI/T₁ -weighted images before and after contrast-enhanced, T₂ -weighted, multi-b-value diffusion-weighted and 3D arterial spin labeling images.

ASSESSMENT

After multiparametric MRI preprocessing, high-throughput features were derived from patients' volumes of interests (VOIs). The support vector machine-based recursive feature elimination was adopted to find the optimal features for low-grade glioma (LGG) vs. high-grade glioma (HGG), and Grade III vs. IV glioma classification tasks. Then support vector machine (SVM) classifiers were established using the optimal features. The accuracy and area under the curve (AUC) was used to assess the grading efficiency.

STATISTICAL TESTS

Student's t-test or a chi-square test were applied on different clinical characteristics to confirm whether intergroup significant differences exist.

RESULTS

Patients' ages between LGG and HGG groups were significantly different (P < 0.01). For each patient, 420 texture and 90 histogram parameters were derived from 10 VOIs of multiparametric MRI. SVM models were established using 30 and 28 optimal features for classifying LGGs from HGGs and grades III from IV, respectively. The accuracies/AUCs were 96.8%/0.987 for classifying LGGs from HGGs, and 98.1%/0.992 for classifying grades III from IV, which were more promising than using histogram parameters or using the single sequence MRI.

DATA CONCLUSION

Texture features were more effective for noninvasively grading gliomas than histogram parameters. The combined application of multiparametric MRI provided a higher grading efficiency. The proposed radiomic strategy could facilitate clinical decision-making for patients with varied glioma grades.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;48:1518-1528.

Identifying HIV Associated Neurocognitive Disorder Using Large-Scale Granger Causality Analysis on Resting-State Functional MRI

We investigate the applicability of large-scale Granger Causality (lsGC) for extracting a measure of multivariate information flow between pairs of regional brain activities from resting-state functional MRI (fMRI) and test the effectiveness of these measures for predicting a disease state. Such pairwise multivariate measures of interaction provide high-dimensional representations of connectivity profiles for each subject and are used in a machine learning task to distinguish between healthy controls and individuals presenting with symptoms of HIV Associated Neurocognitive Disorder (HAND). Cognitive impairment in several domains can occur as a result of HIV infection of the central nervous system. The current paradigm for assessing such impairment is through neuropsychological testing. With fMRI data analysis, we aim at non-invasively capturing differences in brain connectivity patterns between healthy subjects and subjects presenting with symptoms of HAND. To classify the extracted interaction patterns among brain regions, we use a prototype-based learning algorithm called Generalized Matrix Learning Vector Quantization (GMLVQ). Our approach to characterize connectivity using lsGC followed by GMLVQ for subsequent classification yields good prediction results with an accuracy of 87% and an area under the ROC curve (AUC) of up to 0.90. We obtain a statistically significant improvement (p<0.01) over a conventional Granger causality approach (accuracy = 0.76, AUC = 0.74). High accuracy and AUC values using our multivariate method to connectivity analysis suggests that our approach is able to better capture changes in interaction patterns between different brain regions when compared to conventional Granger causality analysis known from the literature.

Investigating Changes in Resting-State Connectivity from Functional MRI Data in Patients with HIV Associated Neurocognitive Disorder Using MCA and Machine Learning

Infection of the brain by the Human Immunodeficiency Virus (HIV) causes irreversible damage to the synaptic connections resulting in cognitive impairment. Patients with HIV infection, showing signs of impairment in multiple cognitive domains, as assessed by neuropsychological testing, are said to exhibit symptoms of HIV Associated Neurocognitive Disorder (HAND). In this study, we use resting-state functional MRI (fMRI) data to distinguish between healthy subjects and subjects with symptoms of HAND. To this end, we first establish a measure of interaction between pairs of regional time-series by quantifying their non-linear functional connectivity using Mutual Connectivity Analysis (MCA). Subsequently, we use a classifier to distinguish patterns of interaction between healthy and diseased individuals. Our results, quantified as the mean Area under the ROC curve (AUC) over 75 iterations, indicate that, using fMRI data, we can discriminate between the two cohorts well (AUC > 0.8). Specifically, we find that MCA (mean AUC = 0.89) based connectivity features perform significantly better (p < 0.05) when compared to cross-correlation (mean AUC = 0.82) at the classification task. A higher AUC using our approach suggests that such a nonlinear approach is better able to capture connectivity changes between brain regions and has potential for the development of novel neuro-imaging biomarkers.

Classification of micro-CT images using 3D characterization of bone canal patterns in human osteogenesis imperfecta

Few studies have analyzed the microstructural properties of bone in cases of Osteogenenis Imperfecta (OI), or 'brittle bone disease'. Current approaches mainly focus on bone mineral density measurements as an indirect indicator of bone strength and quality. It has been shown that bone strength would depend not only on composition but also structural organization. This study aims to characterize 3D structure of the cortical bone in high-resolution micro CT images. A total of 40 bone fragments from 28 subjects (13 with OI and 15 healthy controls) were imaged using micro tomography using a synchrotron light source (SRμCT). Minkowski functionals - volume, surface, curvature, and Euler characteristics - describing the topological organization of the bone were computed from the images. The features were used in a machine learning task to classify between healthy and OI bone. The best classification performance (mean AUC - 0.96) was achieved with a combined 4-dimensional feature of all Minkowski functionals. Individually, the best feature performance was seen using curvature (mean AUC - 0.85), which characterizes the edges within a binary object. These results show that quantitative analysis of cortical bone microstructure, in a computer-aided diagnostics framework, can be used to distinguish between healthy and OI bone with high accuracy.

Using Large-Scale Granger Causality to Study Changes in Brain Network Properties in the Clinically Isolated Syndrome (CIS) Stage of Multiple Sclerosis

Clinically Isolated Syndrome (CIS) is often considered to be the first neurological episode associated with Multiple sclerosis (MS). At an early stage the inflammatory demyelination occurring in the CNS can manifest as a change in neuronal metabolism, with multiple asymptomatic white matter lesions detected in clinical MRI. Such damage may induce topological changes of brain networks, which can be captured by advanced functional MRI (fMRI) analysis techniques. We test this hypothesis by capturing the effective relationships of 90 brain regions, defined in the Automated Anatomic Labeling (AAL) atlas, using a large-scale Granger Causality (lsGC) framework. The resulting networks are then characterized using graph-theoretic measures that quantify various network topology properties at a global as well as at a local level. We study for differences in these properties in network graphs obtained for 18 subjects (10 male and 8 female, 9 with CIS and 9 healthy controls). Global network properties captured trending differences with modularity and clustering coefficient (p<0.1). Additionally, local network properties, such as local efficiency and the strength of connections, captured statistically significant (p<0.01) differences in some regions of the inferior frontal and parietal lobe. We conclude that multivariate analysis of fMRI time-series can reveal interesting information about changes occurring in the brain in early stages of MS.

Mutual Connectivity Analysis (MCA) Using Generalized Radial Basis Function Neural Networks for Nonlinear Functional Connectivity Network Recovery in Resting-State Functional MRI

We investigate the applicability of a computational framework, called mutual connectivity analysis (MCA), for directed functional connectivity analysis in both synthetic and resting-state functional MRI data. This framework comprises of first evaluating non-linear cross-predictability between every pair of time series prior to recovering the underlying network structure using community detection algorithms. We obtain the non-linear cross-prediction score between time series using Generalized Radial Basis Functions (GRBF) neural networks. These cross-prediction scores characterize the underlying functionally connected networks within the resting brain, which can be extracted using non-metric clustering approaches, such as the Louvain method. We first test our approach on synthetic models with known directional influence and network structure. Our method is able to capture the directional relationships between time series (with an area under the ROC curve = 0.92 ± 0.037) as well as the underlying network structure (Rand index = 0.87 ± 0.063) with high accuracy. Furthermore, we test this method for network recovery on resting-state fMRI data, where results are compared to the motor cortex network recovered from a motor stimulation sequence, resulting in a strong agreement between the two (Dice coefficient = 0.45). We conclude that our MCA approach is effective in analyzing non-linear directed functional connectivity and in revealing underlying functional network structure in complex systems.

Detecting Altered connectivity patterns in HIV associated neurocognitive impairment using Mutual Connectivity Analysis

The use of functional Magnetic Resonance Imaging (fMRI) has provided interesting insights into our understanding of the brain. In clinical setups these scans have been used to detect and study changes in the brain network properties in various neurological disorders. A large percentage of subjects infected with HIV present cognitive deficits, which are known as HIV associated neurocognitive disorder (HAND). In this study we propose to use our novel technique named Mutual Connectivity Analysis (MCA) to detect differences in brain networks in subjects with and without HIV infection. Resting state functional MRI scans acquired from 10 subjects (5 HIV+ and 5 HIV-) were subject to standard pre-processing routines. Subsequently, the average time-series for each brain region of the Automated Anatomic Labeling (AAL) atlas are extracted and used with the MCA framework to obtain a graph characterizing the interactions between them. The network graphs obtained for different subjects are then compared using Network-Based Statistics (NBS), which is an approach to detect differences between graphs edges while controlling for the family-wise error rate when mass univariate testing is performed. Applying this approach on the graphs obtained yields a single network encompassing 42 nodes and 65 edges, which is significantly different between the two subject groups. Specifically connections to the regions in and around the basal ganglia are significantly decreased. Also some nodes corresponding to the posterior cingulate cortex are affected. These results are inline with our current understanding of pathophysiological mechanisms of HIV associated neurocognitive disease (HAND) and other HIV based fMRI connectivity studies. Hence, we illustrate the applicability of our novel approach with network-based statistics in a clinical case-control study to detect differences connectivity patterns.

Investigating Changes in Brain Network Properties in HIV-Associated Neurocognitive Disease (HAND) using Mutual Connectivity Analysis (MCA)

About 50% of subjects infected with HIV present deficits in cognitive domains, which are known collectively as HIV associated neurocognitive disorder (HAND). The underlying synaptodendritic damage can be captured using resting state functional MRI, as has been demonstrated by a few earlier studies. Such damage may induce topological changes of brain connectivity networks. We test this hypothesis by capturing the functional interdependence of 90 brain network nodes using a Mutual Connectivity Analysis (MCA) framework with non-linear time series modeling based on Generalized Radial Basis function (GRBF) neural networks. The network nodes are selected based on the regions defined in the Automated Anatomic Labeling (AAL) atlas. Each node is represented by the average time series of the voxels of that region. The resulting networks are then characterized using graph-theoretic measures that quantify various network topology properties at a global as well as at a local level. We tested for differences in these properties in network graphs obtained for 10 subjects (6 male and 4 female, 5 HIV+ and 5 HIV-). Global network properties captured some differences between these subject cohorts, though significant differences were seen only with the clustering coefficient measure. Local network properties, such as local efficiency and the degree of connections, captured significant differences in regions of the frontal lobe, precentral and cingulate cortex amongst a few others. These results suggest that our method can be used to effectively capture differences occurring in brain network connectivity properties revealed by resting-state functional MRI in neurological disease states, such as HAND.

Large-Scale Granger Causality Analysis on Resting-State Functional MRI

We demonstrate an approach to measure the information flow between each pair of time series in resting-state functional MRI (fMRI) data of the human brain and subsequently recover its underlying network structure. By integrating dimensionality reduction into predictive time series modeling, large-scale Granger Causality (lsGC) analysis method can reveal directed information flow suggestive of causal influence at an individual voxel level, unlike other multivariate approaches. This method quantifies the influence each voxel time series has on every other voxel time series in a multivariate sense and hence contains information about the underlying dynamics of the whole system, which can be used to reveal functionally connected networks within the brain. To identify such networks, we perform non-metric network clustering, such as accomplished by the Louvain method. We demonstrate the effectiveness of our approach to recover the motor and visual cortex from resting state human brain fMRI data and compare it with the network recovered from a visuomotor stimulation experiment, where the similarity is measured by the Dice Coefficient (DC). The best DC obtained was 0.59 implying a strong agreement between the two networks. In addition, we thoroughly study the effect of dimensionality reduction in lsGC analysis on network recovery. We conclude that our approach is capable of detecting causal influence between time series in a multivariate sense, which can be used to segment functionally connected networks in the resting-state fMRI.

Assessing vertebral fracture risk on volumetric quantitative computed tomography by geometric characterization of trabecular bone structure

The current clinical standard for measuring Bone Mineral Density (BMD) is dual X-ray absorptiometry, however more recently BMD derived from volumetric quantitative computed tomography has been shown to demonstrate a high association with spinal fracture susceptibility. In this study, we propose a method of fracture risk assessment using structural properties of trabecular bone in spinal vertebrae. Experimental data was acquired via axial multi-detector CT (MDCT) from 12 spinal vertebrae specimens using a whole-body 256-row CT scanner with a dedicated calibration phantom. Common image processing methods were used to annotate the trabecular compartment in the vertebral slices creating a circular region of interest (ROI) that excluded cortical bone for each slice. The pixels inside the ROI were converted to values indicative of BMD. High dimensional geometrical features were derived using the scaling index method (SIM) at different radii and scaling factors (SF). The mean BMD values within the ROI were then extracted and used in conjunction with a support vector machine to predict the failure load of the specimens. Prediction performance was measured using the root-mean-square error (RMSE) metric and determined that SIM combined with mean BMD features (RMSE = 0.82 ± 0.37) outperformed MDCT-measured mean BMD (RMSE = 1.11 ± 0.33) (p < 10^-4). These results demonstrate that biomechanical strength prediction in vertebrae can be significantly improved through the use of SIM-derived texture features from trabecular bone.

Automatic Contrast Enhancement of Brain MR Images Using Hierarchical Correlation Histogram Analysis

Parkinson’s disease is a progressive neurodegenerative disorder that has a higher probability of occurrence in middle-aged and older adults than in the young. With the use of a computer-aided diagnosis (CAD) system, abnormal cell regions can be identified, and this identification can help medical personnel to evaluate the chance of disease. This study proposes a hierarchical correlation histogram analysis based on the grayscale distribution degree of pixel intensity by constructing a correlation histogram, that can improves the adaptive contrast enhancement for specific objects. The proposed method produces significant results during contrast enhancement preprocessing and facilitates subsequent CAD processes, thereby reducing recognition time and improving accuracy. The experimental results show that the proposed method is superior to existing methods by using two estimation image quantitative methods of PSNR and average gradient values. Furthermore, the edge information pertaining to specific cells can effectively increase the accuracy of the results.

Automatic classification of prostate cancer Gleason scores from multiparametric magnetic resonance images

Noninvasive, radiological image-based detection and stratification of Gleason patterns can impact clinical outcomes, treatment selection, and the determination of disease status at diagnosis without subjecting patients to surgical biopsies. We present machine learning-based automatic classification of prostate cancer aggressiveness by combining apparent diffusion coefficient (ADC) and T2-weighted (T2-w) MRI-based texture features. Our approach achieved reasonably accurate classification of Gleason scores (GS) 6(3 + 3) vs. ≥7 and 7(3 + 4) vs. 7(4 + 3) despite the presence of highly unbalanced samples by using two different sample augmentation techniques followed by feature selection-based classification. Our method distinguished between GS 6(3 + 3) and ≥7 cancers with 93% accuracy for cancers occurring in both peripheral (PZ) and transition (TZ) zones and 92% for cancers occurring in the PZ alone. Our approach distinguished the GS 7(3 + 4) from GS 7(4 + 3) with 92% accuracy for cancers occurring in both the PZ and TZ and with 93% for cancers occurring in the PZ alone. In comparison, a classifier using only the ADC mean achieved a top accuracy of 58% for distinguishing GS 6(3 + 3) vs. GS ≥7 for cancers occurring in PZ and TZ and 63% for cancers occurring in PZ alone. The same classifier achieved an accuracy of 59% for distinguishing GS 7(3 + 4) from GS 7(4 + 3) occurring in the PZ and TZ and 60% for cancers occurring in PZ alone. Separate analysis of the cancers occurring in TZ alone was not performed owing to the limited number of samples. Our results suggest that texture features derived from ADC and T2-w MRI together with sample augmentation can help to obtain reasonably accurate classification of Gleason patterns.

Volumetric Characterization of Human Patellar Cartilage Matrix on Phase Contrast X-Ray Computed Tomography

Phase contrast X-ray computed tomography (PCI-CT) has recently emerged as a novel imaging technique that allows visualization of cartilage soft tissue, subsequent examination of chondrocyte patterns, and their correlation to osteoarthritis. Previous studies have shown that 2D texture features are effective at distinguishing between healthy and osteoarthritic regions of interest annotated in the radial zone of cartilage matrix on PCI-CT images. In this study, we further extend the texture analysis to 3D and investigate the ability of volumetric texture features at characterizing chondrocyte patterns in the cartilage matrix for purposes of classification. Here, we extracted volumetric texture features derived from Minkowski Functionals and gray-level co-occurrence matrices (GLCM) from 496 volumes of interest (VOI) annotated on PCI-CT images of human patellar cartilage specimens. The extracted features were then used in a machine-learning task involving support vector regression to classify ROIs as healthy or osteoarthritic. Classification performance was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC). The best classification performance was observed with GLCM features correlation (AUC = 0.83 ± 0.06) and homogeneity (AUC = 0.82 ± 0.07), which significantly outperformed all Minkowski Functionals (p < 0.05). These results suggest that such quantitative analysis of chondrocyte patterns in human patellar cartilage matrix involving GLCM-derived statistical features can distinguish between healthy and osteoarthritic tissue with high accuracy.

Investigating the use of mutual information and non-metric clustering for functional connectivity analysis on resting-state functional MRI

Functional MRI (fMRI) is currently used to investigate structural and functional connectivity in human brain networks. To this end, previous studies have proposed computational methods that involve assumptions that can induce information loss, such as assumed linear coupling of the fMRI signals or requiring dimension reduction. This study presents a new computational framework for investigating the functional connectivity in the brain and recovering network structure while reducing the information loss inherent in previous methods. For this purpose, pair-wise mutual information (MI) was extracted from all pixel time series within the brain on resting-state fMRI data. Non-metric topographic mapping of proximity (TMP) data was subsequently applied to recover network structure from the pair-wise MI analysis. Our computational framework is demonstrated in the task of identifying regions of the primary motor cortex network on resting state fMRI data. For ground truth comparison, we also localized regions of the primary motor cortex associated with hand movement in a task-based fMRI sequence with a finger-tapping stimulus function. The similarity between our pair-wise MI clustering results and the ground truth is evaluated using the dice coefficient. Our results show that non-metric clustering with the TMP algorithm, as performed on pair-wise MI analysis, was able to detect the primary motor cortex network and achieved a dice coefficient of 0.53 in terms of overlap with the ground truth. Thus, we conclude that our computational framework can extract and visualize valuable information concerning the underlying network structure between different regions of the brain in resting state fMRI.

Characterizing Trabecular Bone structure for Assessing Vertebral Fracture Risk on Volumetric Quantitative Computed Tomography

While the proximal femur is preferred for measuring bone mineral density (BMD) in fracture risk estimation, the introduction of volumetric quantitative computed tomography has revealed stronger associations between BMD and spinal fracture status. In this study, we propose to capture properties of trabecular bone structure in spinal vertebrae with advanced second-order statistical features for purposes of fracture risk assessment. For this purpose, axial multi-detector CT (MDCT) images were acquired from 28 spinal vertebrae specimens using a whole-body 256-row CT scanner with a dedicated calibration phantom. A semi-automated method was used to annotate the trabecular compartment in the central vertebral slice with a circular region of interest (ROI) to exclude cortical bone; pixels within were converted to values indicative of BMD. Six second-order statistical features derived from gray-level co-occurrence matrices (GLCM) and the mean BMD within the ROI were then extracted and used in conjunction with a generalized radial basis functions (GRBF) neural network to predict the failure load of the specimens; true failure load was measured through biomechanical testing. Prediction performance was evaluated with a root-mean-square error (RMSE) metric. The best prediction performance was observed with GLCM feature 'correlation' (RMSE = 1.02 ± 0.18), which significantly outperformed all other GLCM features (p < 0.01). GLCM feature correlation also significantly outperformed MDCT-measured mean BMD (RMSE = 1.11 ± 0.17) (p < 10^-4). These results suggest that biomechanical strength prediction in spinal vertebrae can be significantly improved through characterization of trabecular bone structure with GLCM-derived texture features.

Functional Connectivity Analysis in Resting State fMRI with Echo-State Networks and Non-Metric Clustering for Network Structure Recovery

Echo state networks (ESN) are recurrent neural networks where the hidden layer is replaced with a fixed reservoir of neurons. Unlike feed-forward networks, neuron training in ESN is restricted to the output neurons alone thereby providing a computational advantage. We demonstrate the use of such ESNs in our mutual connectivity analysis (MCA) framework for recovering the primary motor cortex network associated with hand movement from resting state functional MRI (fMRI) data. Such a framework consists of two steps - (1) defining a pair-wise affinity matrix between different pixel time series within the brain to characterize network activity and (2) recovering network components from the affinity matrix with non-metric clustering. Here, ESNs are used to evaluate pair-wise cross-estimation performance between pixel time series to create the affinity matrix, which is subsequently subject to non-metric clustering with the Louvain method. For comparison, the ground truth of the motor cortex network structure is established with a task-based fMRI sequence. Overlap between the primary motor cortex network recovered with our model free MCA approach and the ground truth was measured with the Dice coefficient. Our results show that network recovery with our proposed MCA approach is in close agreement with the ground truth. Such network recovery is achieved without requiring low-pass filtering of the time series ensembles prior to analysis, an fMRI preprocessing step that has courted controversy in recent years. Thus, we conclude our MCA framework can allow recovery and visualization of the underlying functionally connected networks in the brain on resting state fMRI.

Integrating dimension reduction and out-of-sample extension in automated classification of ex vivo human patellar cartilage on phase contrast X-ray computed tomography

Phase contrast X-ray computed tomography (PCI-CT) has been demonstrated as a novel imaging technique that can visualize human cartilage with high spatial resolution and soft tissue contrast. Different textural approaches have been previously investigated for characterizing chondrocyte organization on PCI-CT to enable classification of healthy and osteoarthritic cartilage. However, the large size of feature sets extracted in such studies motivates an investigation into algorithmic feature reduction for computing efficient feature representations without compromising their discriminatory power. For this purpose, geometrical feature sets derived from the scaling index method (SIM) were extracted from 1392 volumes of interest (VOI) annotated on PCI-CT images of ex vivo human patellar cartilage specimens. The extracted feature sets were subject to linear and non-linear dimension reduction techniques as well as feature selection based on evaluation of mutual information criteria. The reduced feature set was subsequently used in a machine learning task with support vector regression to classify VOIs as healthy or osteoarthritic; classification performance was evaluated using the area under the receiver-operating characteristic (ROC) curve (AUC). Our results show that the classification performance achieved by 9-D SIM-derived geometric feature sets (AUC: 0.96 ± 0.02) can be maintained with 2-D representations computed from both dimension reduction and feature selection (AUC values as high as 0.97 ± 0.02). Thus, such feature reduction techniques can offer a high degree of compaction to large feature sets extracted from PCI-CT images while maintaining their ability to characterize the underlying chondrocyte patterns.

Nonlinear Functional Connectivity Network Recovery in the Human Brain with Mutual Connectivity Analysis (MCA): Convergent Cross-Mapping and Non-Metric Clustering

We explore a computational framework for functional connectivity analysis in resting-state functional MRI (fMRI) data acquired from the human brain for recovering the underlying network structure and understanding causality between network components. Termed mutual connectivity analysis (MCA), this framework involves two steps, the first of which is to evaluate the pair-wise cross-prediction performance between fMRI pixel time series within the brain. In a second step, the underlying network structure is subsequently recovered from the affinity matrix using non-metric network clustering approaches, such as the so-called Louvain method. Finally, we use convergent cross-mapping (CCM) to study causality between different network components. We demonstrate our MCA framework in the problem of recovering the motor cortex network associated with hand movement from resting state fMRI data. Results are compared with a ground truth of active motor cortex regions as identified by a task-based fMRI sequence involving a finger-tapping stimulation experiment. Our results regarding causation between regions of the motor cortex revealed a significant directional variability and were not readily interpretable in a consistent manner across subjects. However, our results on whole-slice fMRI analysis demonstrate that MCA-based model-free recovery of regions associated with the primary motor cortex and supplementary motor area are in close agreement with localization of similar regions achieved with a task-based fMRI acquisition. Thus, we conclude that our MCA methodology can extract and visualize valuable information concerning the underlying network structure between different regions of the brain in resting state fMRI.

Computer-aided diagnosis for phase-contrast X-ray computed tomography: quantitative characterization of human patellar cartilage with high-dimensional geometric features

Phase-contrast computed tomography (PCI-CT) has shown tremendous potential as an imaging modality for visualizing human cartilage with high spatial resolution. Previous studies have demonstrated the ability of PCI-CT to visualize (1) structural details of the human patellar cartilage matrix and (2) changes to chondrocyte organization induced by osteoarthritis. This study investigates the use of high-dimensional geometric features in characterizing such chondrocyte patterns in the presence or absence of osteoarthritic damage. Geometrical features derived from the scaling index method (SIM) and statistical features derived from gray-level co-occurrence matrices were extracted from 842 regions of interest (ROI) annotated on PCI-CT images of ex vivo human patellar cartilage specimens. These features were subsequently used in a machine learning task with support vector regression to classify ROIs as healthy or osteoarthritic; classification performance was evaluated using the area under the receiver-operating characteristic curve (AUC). SIM-derived geometrical features exhibited the best classification performance (AUC, 0.95 ± 0.06) and were most robust to changes in ROI size. These results suggest that such geometrical features can provide a detailed characterization of the chondrocyte organization in the cartilage matrix in an automated and non-subjective manner, while also enabling classification of cartilage as healthy or osteoarthritic with high accuracy. Such features could potentially serve as imaging markers for evaluating osteoarthritis progression and its response to different therapeutic intervention strategies.

Quantification of heterogeneity as a biomarker in tumor imaging: a systematic review

BACKGROUND

Many techniques are proposed for the quantification of tumor heterogeneity as an imaging biomarker for differentiation between tumor types, tumor grading, response monitoring and outcome prediction. However, in clinical practice these methods are barely used. This study evaluates the reported performance of the described methods and identifies barriers to their implementation in clinical practice.

METHODOLOGY

The Ovid, Embase, and Cochrane Central databases were searched up to 20 September 2013. Heterogeneity analysis methods were classified into four categories, i.e., non-spatial methods (NSM), spatial grey level methods (SGLM), fractal analysis (FA) methods, and filters and transforms (F&T). The performance of the different methods was compared.

PRINCIPAL FINDINGS

Of the 7351 potentially relevant publications, 209 were included. Of these studies, 58% reported the use of NSM, 49% SGLM, 10% FA, and 28% F&T. Differentiation between tumor types, tumor grading and/or outcome prediction was the goal in 87% of the studies. Overall, the reported area under the curve (AUC) ranged from 0.5 to 1 (median 0.87). No relation was found between the performance and the quantification methods used, or between the performance and the imaging modality. A negative correlation was found between the tumor-feature ratio and the AUC, which is presumably caused by overfitting in small datasets. Cross-validation was reported in 63% of the classification studies. Retrospective analyses were conducted in 57% of the studies without a clear description.

CONCLUSIONS

In a research setting, heterogeneity quantification methods can differentiate between tumor types, grade tumors, and predict outcome and monitor treatment effects. To translate these methods to clinical practice, more prospective studies are required that use external datasets for validation: these datasets should be made available to the community to facilitate the development of new and improved methods.

Using Anisotropic 3D Minkowski Functionals for Trabecular Bone Characterization and Biomechanical Strength Prediction in Proximal Femur Specimens

The ability of Anisotropic Minkowski Functionals (AMFs) to capture local anisotropy while evaluating topological properties of the underlying gray-level structures has been previously demonstrated. We evaluate the ability of this approach to characterize local structure properties of trabecular bone micro-architecture in ex vivo proximal femur specimens, as visualized on multi-detector CT, for purposes of biomechanical bone strength prediction. To this end, volumetric AMFs were computed locally for each voxel of volumes of interest (VOI) extracted from the femoral head of 146 specimens. The local anisotropy captured by such AMFs was quantified using a fractional anisotropy measure; the magnitude and direction of anisotropy at every pixel was stored in histograms that served as a feature vectors that characterized the VOIs. A linear multi-regression analysis algorithm was used to predict the failure load (FL) from the feature sets; the predicted FL was compared to the true FL determined through biomechanical testing. The prediction performance was measured by the root mean square error (RMSE) for each feature set. The best prediction performance was obtained from the fractional anisotropy histogram of AMF Euler Characteristic (RMSE = 1.01 ± 0.13), which was significantly better than MDCT-derived mean BMD (RMSE = 1.12 ± 0.16, p<0.05). We conclude that such anisotropic Minkowski Functionals can capture valuable information regarding regional trabecular bone quality and contribute to improved bone strength prediction, which is important for improving the clinical assessment of osteoporotic fracture risk.

Investigating the use of texture features for analysis of breast lesions on contrast-enhanced cone beam CT

Cone beam computed tomography (CBCT) has found use in mammography for imaging the entire breast with sufficient spatial resolution at a radiation dose within the range of that of conventional mammography. Recently, enhancement of lesion tissue through the use of contrast agents has been proposed for cone beam CT. This study investigates whether the use of such contrast agents improves the ability of texture features to differentiate lesion texture from healthy tissue on CBCT in an automated manner. For this purpose, 9 lesions were annotated by an experienced radiologist on both regular and contrast-enhanced CBCT images using two-dimensional (2D) square ROIs. These lesions were then segmented, and each pixel within the lesion ROI was assigned a label - lesion or non-lesion, based on the segmentation mask. On both sets of CBCT images, four three-dimensional (3D) Minkowski Functionals were used to characterize the local topology at each pixel. The resulting feature vectors were then used in a machine learning task involving support vector regression with a linear kernel (SVRlin) to classify each pixel as belonging to the lesion or non-lesion region of the ROI. Classification performance was assessed using the area under the receiver-operating characteristic (ROC) curve (AUC). Minkowski Functionals derived from contrast-enhanced CBCT images were found to exhibit significantly better performance at distinguishing between lesion and non-lesion areas within the ROI when compared to those extracted from CBCT images without contrast enhancement (p < 0.05). Thus, contrast enhancement in CBCT can improve the ability of texture features to distinguish lesions from surrounding healthy tissue.

Predicting the Biomechanical Strength of Proximal Femur Specimens with Minkowski Functionals and Support Vector Regression

Regional trabecular bone quality estimation for purposes of femoral bone strength prediction is important for improving the clinical assessment of osteoporotic fracture risk. In this study, we explore the ability of 3D Minkowski Functionals derived from multi-detector computed tomography (MDCT) images of proximal femur specimens in predicting their corresponding biomechanical strength. MDCT scans were acquired for 50 proximal femur specimens harvested from human cadavers. An automated volume of interest (VOI)-fitting algorithm was used to define a consistent volume in the femoral head of each specimen. In these VOIs, the trabecular bone micro-architecture was characterized by statistical moments of its BMD distribution and by topological features derived from Minkowski Functionals. A linear multi-regression analysis and a support vector regression (SVR) algorithm with a linear kernel were used to predict the failure load (FL) from the feature sets; the predicted FL was compared to the true FL determined through biomechanical testing. The prediction performance was measured by the root mean square error (RMSE) for each feature set. The best prediction result was obtained from the Minkowski Functional surface used in combination with SVR, which had the lowest prediction error (RMSE = 0.939 ± 0.345) and which was significantly lower than mean BMD (RMSE = 1.075 ± 0.279, p<0.005). Our results indicate that the biomechanical strength prediction can be significantly improved in proximal femur specimens with Minkowski Functionals extracted from on MDCT images used in conjunction with support vector regression.

Phase contrast imaging X-ray computed tomography: Quantitative characterization of human patellar cartilage matrix with topological and geometrical features

Current assessment of cartilage is primarily based on identification of indirect markers such as joint space narrowing and increased subchondral bone density on x-ray images. In this context, phase contrast CT imaging (PCI-CT) has recently emerged as a novel imaging technique that allows a direct examination of chondrocyte patterns and their correlation to osteoarthritis through visualization of cartilage soft tissue. This study investigates the use of topological and geometrical approaches for characterizing chondrocyte patterns in the radial zone of the knee cartilage matrix in the presence and absence of osteoarthritic damage. For this purpose, topological features derived from Minkowski Functionals and geometric features derived from the Scaling Index Method (SIM) were extracted from 842 regions of interest (ROI) annotated on PCI-CT images of healthy and osteoarthritic specimens of human patellar cartilage. The extracted features were then used in a machine learning task involving support vector regression to classify ROIs as healthy or osteoarthritic. Classification performance was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC). The best classification performance was observed with high-dimensional geometrical feature vectors derived from SIM (0.95 ± 0.06) which outperformed all Minkowski Functionals (p < 0.001). These results suggest that such quantitative analysis of chondrocyte patterns in human patellar cartilage matrix involving SIM-derived geometrical features can distinguish between healthy and osteoarthritic tissue with high accuracy.

Classification of small lesions on dynamic breast MRI: Integrating dimension reduction and out-of-sample extension into CADx methodology

OBJECTIVE

While dimension reduction has been previously explored in computer aided diagnosis (CADx) as an alternative to feature selection, previous implementations of its integration into CADx do not ensure strict separation between training and test data required for the machine learning task. This compromises the integrity of the independent test set, which serves as the basis for evaluating classifier performance.

METHODS AND MATERIALS

We propose, implement and evaluate an improved CADx methodology where strict separation is maintained. This is achieved by subjecting the training data alone to dimension reduction; the test data is subsequently processed with out-of-sample extension methods. Our approach is demonstrated in the research context of classifying small diagnostically challenging lesions annotated on dynamic breast magnetic resonance imaging (MRI) studies. The lesions were dynamically characterized through topological feature vectors derived from Minkowski functionals. These feature vectors were then subject to dimension reduction with different linear and non-linear algorithms applied in conjunction with out-of-sample extension techniques. This was followed by classification through supervised learning with support vector regression. Area under the receiver-operating characteristic curve (AUC) was evaluated as the metric of classifier performance.

RESULTS

Of the feature vectors investigated, the best performance was observed with Minkowski functional 'perimeter' while comparable performance was observed with 'area'. Of the dimension reduction algorithms tested with 'perimeter', the best performance was observed with Sammon's mapping (0.84±0.10) while comparable performance was achieved with exploratory observation machine (0.82±0.09) and principal component analysis (0.80±0.10).

CONCLUSIONS

The results reported in this study with the proposed CADx methodology present a significant improvement over previous results reported with such small lesions on dynamic breast MRI. In particular, non-linear algorithms for dimension reduction exhibited better classification performance than linear approaches, when integrated into our CADx methodology. We also note that while dimension reduction techniques may not necessarily provide an improvement in classification performance over feature selection, they do allow for a higher degree of feature compaction.

Classification of small lesions in dynamic breast MRI: Eliminating the need for precise lesion segmentation through spatio-temporal analysis of contrast enhancement over time

Characterizing the dignity of breast lesions as benign or malignant is specifically difficult for small lesions; they don't exhibit typical characteristics of malignancy and are harder to segment since margins are harder to visualize. Previous attempts at using dynamic or morphologic criteria to classify small lesions (mean lesion diameter of about 1 cm) have not yielded satisfactory results. The goal of this work was to improve the classification performance in such small diagnostically challenging lesions while concurrently eliminating the need for precise lesion segmentation. To this end, we introduce a method for topological characterization of lesion enhancement patterns over time. Three Minkowski Functionals were extracted from all five post-contrast images of sixty annotated lesions on dynamic breast MRI exams. For each Minkowski Functional, topological features extracted from each post-contrast image of the lesions were combined into a high-dimensional texture feature vector. These feature vectors were classified in a machine learning task with support vector regression. For comparison, conventional Haralick texture features derived from gray-level co-occurrence matrices (GLCM) were also used. A new method for extracting thresholded GLCM features was also introduced and investigated here. The best classification performance was observed with Minkowski Functionals area and perimeter, thresholded GLCM features f8 and f9, and conventional GLCM features f4 and f6. However, both Minkowski Functionals and thresholded GLCM achieved such results without lesion segmentation while the performance of GLCM features significantly deteriorated when lesions were not segmented (p < 0.05). This suggests that such advanced spatio-temporal characterization can improve the classification performance achieved in such small lesions, while simultaneously eliminating the need for precise segmentation.

Computer-aided diagnosis in phase contrast imaging X-ray computed tomography for quantitative characterization of ex vivo human patellar cartilage

Visualization of ex vivo human patellar cartilage matrix through the phase contrast imaging X-ray computed tomography (PCI-CT) has been previously demonstrated. Such studies revealed osteoarthritis-induced changes to chondrocyte organization in the radial zone. This study investigates the application of texture analysis to characterizing such chondrocyte patterns in the presence and absence of osteoarthritic damage. Texture features derived from Minkowski functionals (MF) and gray-level co-occurrence matrices (GLCM) were extracted from 842 regions of interest (ROI) annotated on PCI-CT images of ex vivo human patellar cartilage specimens. These texture features were subsequently used in a machine learning task with support vector regression to classify ROIs as healthy or osteoarthritic; classification performance was evaluated using the area under the receiver operating characteristic curve (AUC). The best classification performance was observed with the MF features perimeter (AUC: 0.94 ±0.08 ) and "Euler characteristic" (AUC: 0.94 ±0.07 ), and GLCM-derived feature "Correlation" (AUC: 0.93 ±0.07). These results suggest that such texture features can provide a detailed characterization of the chondrocyte organization in the cartilage matrix, enabling classification of cartilage as healthy or osteoarthritic with high accuracy.

Characterizing healthy and osteoarthritic knee cartilage on phase contrast CT with geometric texture features

The current approach to evaluating cartilage degeneration at the knee joint requires visualization of the joint space on radiographic images where indirect cues such as joint space narrowing serve as markers for osteoarthritis. A recent novel approach to visualizing the knee cartilage matrix using phase contrast imaging (PCI) with computed tomography (CT) was shown to allow direct examination of chondrocyte patterns and their subsequent correlation to osteoarthritis. This study aims to characterize chondrocyte cell patterns in the radial zone of the knee cartilage matrix in the presence and absence of osteoarthritic damage through texture analysis. Statistical features derived from gray-level co-occurrence matrices (GLCM) and geometric features derived from the Scaling Index Method (SIM) were extracted from 404 regions of interest (ROI) annotated on PCI images of healthy and osteoarthritic specimens of knee cartilage. These texture features were then used in a machine learning task to classify ROIs as healthy or osteoarthritic. A fuzzy k-nearest neighbor classifier was used and its performance was evaluated using the area under the Receiver Operating Characteristic (ROC) curve (AUC). The best classification performance was observed with high-dimensional geometrical feature vectors derived from SIM and GLCM correlation features. With the experimental conditions used in this study, both SIM and GLCM achieved a high classification performance (AUC value of 0.98) in the task of distinguishing between healthy and osteoarthritic ROIs. These results show that such quantitative analysis of chondrocyte patterns in the knee cartilage matrix can distinguish between healthy and osteoarthritic tissue with high accuracy.