The relevance of CT-based geometric and radiomics analysis of whole liver tumor burden to predict survival of patients with metastatic colorectal cancer.

OBJECTIVES: To investigate the prediction of 1-year survival (1-YS) in patients with metastatic colorectal cancer with use of a systematic comparative analysis of quantitative imaging biomarkers (QIBs) based on the geometric and radiomics analysis of whole liver tumor burden (WLTB) in comparison to predictions based on the tumor burden score (TBS), WLTB volume alone, and a clinical model. METHODS: A total of 103 patients (mean age: 61.0 ± 11.2 years) with colorectal liver metastases were analyzed in this retrospective study. Automatic segmentations of WLTB from baseline contrast-enhanced CT images were used. Established biomarkers as well as a standard radiomics model building were used to derive 3 prognostic models. The benefits of a geometric metastatic spread (GMS) model, the Aerts radiomics prior model of the WLTB, and the performance of TBS and WLTB volume alone were assessed. All models were analyzed in both statistical and predictive machine learning settings in terms of AUC. RESULTS: TBS showed the best discriminative performance in a statistical setting to discriminate 1-YS (AUC = 0.70, CI: [0.56, 0.90]). For the machine learning-based prediction for unseen patients, both a model of the GMS of WLTB (0.73, CI: [0.60, 0.84]) and the Aerts radiomics prior model (0.76, CI: [0.65, 0.86]) applied on the WLTB showed a numerically higher predictive performance than TBS (0.68, CI: [0.54, 0.79]), radiomics (0.65, CI: [0.55, 0.78]), WLTB volume alone (0.53, CI: [0.40. 0.66]), or the clinical model (0.56, CI: [0.43, 0.67]). CONCLUSIONS: The imaging-based GMS model may be a first step towards a more fine-grained machine learning extension of the TBS concept for risk stratification in mCRC patients without the vulnerability to technical variance of radiomics. KEY POINTS: • CT-based geometric distribution and radiomics analysis of whole liver tumor burden in metastatic colorectal cancer patients yield prognostic information. • Differences in survival are possibly attributable to the spatial distribution of metastatic lesions and the geometric metastatic spread analysis of all liver metastases may serve as robust imaging biomarker invariant to technical variation. • Imaging-based prediction models outperform clinical models for 1-year survival prediction in metastatic colorectal cancer patients with liver metastases.

Bifurcation matching for consistent cerebral vessel labeling in CTA of stroke patients.

PURPOSE: Vessel labeling is a prerequisite for comparing cerebral vasculature across patients, e.g., for straightened vessel examination or for localization. Extracting vessels from computed tomography angiography scans may come with a trade-off in segmentation accuracy. Vessels might be neglected or artificially created, increasing the difficulty of labeling. Related work mainly focuses on magnetic resonance angiography without stroke and uses trainable approaches requiring costly labels. METHODS: We present a robust method to identify major arteries and bifurcations in cerebrovascular models generated from existing segmentations. To localize bifurcations of the Circle of Willis, candidate paths for the adjacent vessels of interest are identified using registered landmarks. From those paths, the optimal ones are extracted by recursively maximizing an objective function for all adjacent vessels starting from a bifurcation to avoid erroneous paths and compensate for stroke. RESULTS: In 100 CTA stroke data sets for evaluation, 6 bifurcation locations are placed correctly in 85% of cases; 92.5% when allowing a margin of 5 mm. On average, 14 vessels of interest are found in 90% of the cases and traced correctly end-to-end in 73.5%. The baseline achieves similar detection rates but only 35.5% of the arteries are traced in full. CONCLUSION: Formulating the vessel labeling process as a maximization task for bifurcation matching can vastly improve accurate vessel tracing. The proposed algorithm only uses simple features and does not require expensive training data.

How scan parameter choice affects deep learning-based coronary artery disease assessment from computed tomography.

Recently, algorithms capable of assessing the severity of Coronary Artery Disease (CAD) in form of the Coronary Artery Disease-Reporting and Data System (CAD-RADS) grade from Coronary Computed Tomography Angiography (CCTA) scans using Deep Learning (DL) were proposed. Before considering to apply these algorithms in clinical practice, their robustness regarding different commonly used Computed Tomography (CT)-specific image formation parameters-including denoising strength, slab combination, and reconstruction kernel-needs to be evaluated. For this study, we reconstructed a data set of 500 patient CCTA scans under seven image formation parameter configurations. We select one default configuration and evaluate how varying individual parameters impacts the performance and stability of a typical algorithm for automated CAD assessment from CCTA. This algorithm consists of multiple preprocessing and a DL prediction step. We evaluate the influence of the parameter changes on the entire pipeline and additionally on only the DL step by propagating the centerline extraction results of the default configuration to all others. We consider the standard deviation of the CAD severity prediction grade difference between the default and variation configurations to assess the stability w.r.t. parameter changes. For the full pipeline we observe slight instability (± 0.226 CAD-RADS) for all variations. Predictions are more stable with centerlines propagated from the default to the variation configurations (± 0.122 CAD-RADS), especially for differing denoising strengths (± 0.046 CAD-RADS). However, stacking slabs with sharp boundaries instead of mixing slabs in overlapping regions (called true stack ± 0.313 CAD-RADS) and increasing the sharpness of the reconstruction kernel (± 0.150 CAD-RADS) leads to unstable predictions. Regarding the clinically relevant tasks of excluding CAD (called rule-out; AUC default 0.957, min 0.937) and excluding obstructive CAD (called hold-out; AUC default 0.971, min 0.964) the performance remains on a high level for all variations. Concluding, an influence of reconstruction parameters on the predictions is observed. Especially, scans reconstructed with the true stack parameter need to be treated with caution when using a DL-based method. Also, reconstruction kernels which are underrepresented in the training data increase the prediction uncertainty.

Radiomics Features of the Spleen as Surrogates for CT-Based Lymphoma Diagnosis and Subtype Differentiation.

The spleen is often involved in malignant lymphoma, which manifests on CT as either splenomegaly or focal, hypodense lymphoma lesions. This study aimed to investigate the diagnostic value of radiomics features of the spleen in classifying malignant lymphoma against non-lymphoma as well as the determination of malignant lymphoma subtypes in the case of disease presence-in particular Hodgkin lymphoma (HL), diffuse large B-cell lymphoma (DLBCL), mantle-cell lymphoma (MCL), and follicular lymphoma (FL). Spleen segmentations of 326 patients (139 female, median age 54.1 +/- 18.7 years) were generated and 1317 radiomics features per patient were extracted. For subtype classification, we created four different binary differentiation tasks and addressed them with a Random Forest classifier using 10-fold cross-validation. To detect the most relevant features, permutation importance was analyzed. Classifier results using all features were: malignant lymphoma vs. non-lymphoma AUC = 0.86 (p < 0.01); HL vs. NHL AUC = 0.75 (p < 0.01); DLBCL vs. other NHL AUC = 0.65 (p < 0.01); MCL vs. FL AUC = 0.67 (p < 0.01). Classifying malignant lymphoma vs. non-lymphoma was also possible using only shape features AUC = 0.77 (p < 0.01), with the most important feature being sphericity. Based on only shape features, a significant AUC could be achieved for all tasks, however, best results were achieved combining shape and textural features. This study demonstrates the value of splenic imaging and radiomic analysis in the diagnostic process in malignant lymphoma detection and subtype classification.

Unraveling the interplay of image formation, data representation and learning in CT-based COPD phenotyping automation: The need for a meta-strategy.

PURPOSE: In the literature on automated phenotyping of chronic obstructive pulmonary disease (COPD), there is a multitude of isolated classical machine learning and deep learning techniques, mostly investigating individual phenotypes, with small study cohorts and heterogeneous meta-parameters, e.g., different scan protocols or segmented regions. The objective is to compare the impact of different experimental setups, i.e., varying meta-parameters related to image formation and data representation, with the impact of the learning technique for subtyping automation for a variety of phenotypes. The identified associations of these parameters with automation performance and their interactions might be a first step towards a determination of optimal meta-parameters, i.e., a meta-strategy. METHODS: A clinical cohort of 981 patients (53.8 ± 15.1 years, 554 male) was examined. The inspiratory CT images were analyzed to automate the diagnosis of 13 COPD phenotypes given by two radiologists. A benchmark feature set that integrates many quantitative criteria was extracted from the lung and trained a variety of learning algorithms on the first 654 patients (two thirds) and the respective algorithm retrospectively assessed the remaining 327 patients (one third). The automation performance was evaluated by the area under the receiver operating characteristic curve (AUC). 1717 experiments were conducted with varying meta-parameters such as reconstruction kernel, segmented regions and input dimensionality, i.e., number of extracted features. The association of the meta-parameters with the automation performance was analyzed by multivariable general linear model decomposition of the automation performance in the contributions of meta-parameters and the learning technique. RESULTS: The automation performance varied strongly for varying meta-parameters. For emphysema-predominant phenotypes, an AUC of 93%-95% could be achieved for the best meta-configuration. The airways-predominant phenotypes led to a lower performance of 65%-85%, while smooth kernel configurations on average were unexpectedly superior to those with sharp kernels. The performance impact of meta-parameters, even that of often neglected ones like the missing-data imputation, was in general larger than that of the learning technique. Advanced learning techniques like 3D deep learning or automated machine learning yielded inferior automation performance for non-optimal meta-configurations in comparison to simple techniques with suitable meta-configurations. The best automation performance was achieved by a combination of modern learning techniques and a suitable meta-configuration. CONCLUSIONS: Our results indicate that for COPD phenotype automation, study design parameters such as reconstruction kernel and the model input dimensionality should be adapted to the learning technique and may be more important than the technique itself. To achieve optimal automation and prediction results, the interaction between input those meta-parameters and the learning technique should be considered. This might be particularly relevant for the development of specific scan protocols for novel learning algorithms, and towards an understanding of good study design for automated phenotyping.

Quantitative Imaging Biomarkers of the Whole Liver Tumor Burden Improve Survival Prediction in Metastatic Pancreatic Cancer.

Finding prognostic biomarkers with high accuracy in patients with pancreatic cancer (PC) remains a challenging problem. To improve the prediction of survival and to investigate the relevance of quantitative imaging biomarkers (QIB) we combined QIB with established clinical parameters. In this retrospective study a total of 75 patients with metastatic PC and liver metastases were analyzed. Segmentations of whole liver tumor burden (WLTB) from baseline contrast-enhanced CT images were used to derive QIBs. The benefits of QIBs in multivariable Cox models were analyzed in comparison with two clinical prognostic models from the literature. To discriminate survival, the two clinical models had concordance indices of 0.61 and 0.62 in a statistical setting. Combined clinical and imaging-based models achieved concordance indices of 0.74 and 0.70 with WLTB volume, tumor burden score (TBS), and bilobar disease being the three WLTB parameters that were kept by backward elimination. These combined clinical and imaging-based models have significantly higher predictive performance in discriminating survival than the underlying clinical models alone (p < 0.003). Radiomics and geometric WLTB analysis of patients with metastatic PC with liver metastases enhances the modeling of survival compared with models based on clinical parameters alone.

Semiautomated versus manual evaluation of liver metastases treated by radiofrequency ablation.

PURPOSE: To determine the accuracy of semiautomated volume and density measurements of liver metastases from colorectal and breast cancer before and after radiofrequency (RF) ablation compared with manual evaluation. MATERIALS AND METHODS: Twenty-five patients (mean age, 63.2 years +/- 10.7) with 50 known liver metastases from underlying primary breast (n = 15) or colorectal cancer (n = 35) underwent triphasic contrast-enhanced multidetector computed tomography (CT) to evaluate hepatic tumor load and localization before RF ablation and for postinterventional follow-up. Each lesion was quantified in terms of volume and CT value (in HU) with a semiautomated software tool and manually by an experienced radiologist before and 4 months after RF ablation. RESULTS: Before RF ablation, all 50 liver metastases, and after ablation, 49 of 50 ablation zones (98%), were correctly evaluated by the software. Mean lesion volumes before and after the intervention were 5.5 cm(3) and 22.4 cm(3), respectively. Corresponding concordance correlation coefficients between measurement techniques were 0.98 and 0.99, respectively, for volume; and 0.90 and 0.76, respectively, for CT value. CONCLUSIONS: Compared with manual measurements, semiautomated volumetric assessment of liver metastases before and after RF ablation demonstrated a high degree of correlation. Agreement of attenuation was slightly worse, particularly when assessing the postinterventional multidetector CT examination, probably because of the different regions of interest used for manual and semiautomated assessment of CT values.

Accuracy of radiomics for differentiating diffuse liver diseases on non-contrast CT.

PURPOSE: Radiomics help move cross-sectional imaging into the domain of quantitative imaging to assess the lesions, their stoma as well as in their temporal monitoring. We applied and assessed the accuracy of radiomics for differentiating healthy liver from diffuse liver diseases (cirrhosis, steatosis, amiodarone deposition, and iron overload) on non-contrast abdomen CT images in an institutional-reviewed board-approved, retrospective study. METHODS: Our study included 300 adult patients (mean age 63 ± 16 years; 171 men, 129 women) who underwent non-contrast abdomen CT and had either a healthy liver (n = 100 patients) or an evidence of diffuse liver disease (n = 200). The diffuse liver diseases included steatosis (n = 50), cirrhosis (n = 50), hyperdense liver due to amiodarone deposition (n = 50), or iron overload (n = 50). We manually segmented the liver in one section at the level of the porta hepatis (all 300 patients) and then over the entire liver volume (50 patients). Radiomics were estimated for the liver, and statistical comparison was performed with multiple logistic regression and random forest classifier. RESULTS: With random forest classifier, the AUC for radiomics ranged between 0.72 (iron overload vs. healthy liver) and 0.98 (hepatic steatosis vs. healthy liver) for differentiating diffuse liver disease from the healthy liver. Combined root mean square and gray-level co-occurrence matrix had the highest AUC (AUC:0.99, p < 0.01) for differentiating healthy liver from steatosis. Radiomics were more accurate for differentiating healthy liver from amiodarone (AUC:0.93) than from iron overload (AUC:0.79). CONCLUSION: Radiomics enable differentiation of healthy liver from hepatic steatosis, cirrhosis, amiodarone deposition, and iron overload from a single section of non-contrast abdominal CT. The high accuracy of radiomics coupled with rapid segmentation of the region of interest, radiomics estimation, and statistical analyses within the same prototype makes a compelling case for bringing radiomics to clinical use for improving reporting in evaluation of healthy liver and diffuse liver diseases.

Publisher Correction: The Technome - A Predictive Internal Calibration Approach for Quantitative Imaging Biomarker Research.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

The Technome - A Predictive Internal Calibration Approach for Quantitative Imaging Biomarker Research.

The goal of radiomics is to convert medical images into a minable data space by extraction of quantitative imaging features for clinically relevant analyses, e.g. survival time prediction of a patient. One problem of radiomics from computed tomography is the impact of technical variation such as reconstruction kernel variation within a study. Additionally, what is often neglected is the impact of inter-patient technical variation, resulting from patient characteristics, even when scan and reconstruction parameters are constant. In our approach, measurements within 3D regions-of-interests (ROI) are calibrated by further ROIs such as air, adipose tissue, liver, etc. that are used as control regions (CR). Our goal is to derive general rules for an automated internal calibration that enhance prediction, based on the analysed features and a set of CRs. We define qualification criteria motivated by status-quo radiomics stability analysis techniques to only collect information from the CRs which is relevant given a respective task. These criteria are used in an optimisation to automatically derive a suitable internal calibration for prediction tasks based on the CRs. Our calibration enhanced the performance for centrilobular emphysema prediction in a COPD study and prediction of patients' one-year-survival in an oncological study.

Semi-automated measurement of hyperdense, hypodense and heterogeneous hepatic metastasis on standard MDCT slices. Comparison of semi-automated and manual measurement of RECIST and WHO criteria.

As semi-automated measurement would be desirable for lesion quantification and therapy-response control, the purpose of this study was to compare semi-automated measurements with manual assessment of different types of hepatic metastases. Seventy-six patients with known liver metastases were analysed. All of them underwent contrast-enhanced 16-MDCT (16 x 0.75 mm collimation, 120 kV, 0.5 s rotation time, 160 mAs(eff)) for evaluation of follow-up status. On the basis of standard reconstructed 5-mm slices (in 4-mm increments), each lesion was quantified based on RECIST and WHO criteria using a semi-automated software tool (Syngo Oncology) and also manually by an experienced radiologist. Results from the software were compared to manual measurements. Statistical analysis was performed applying the concordance correlation coefficient, and results were represented graphically in Bland-Altman plots. A total of 52 hyperdense, 57 hypodense and 56 heterogeneous metastases were found and correctly measured by the software. All three lesion types revealed a strong correlation agreement between measurement techniques [RECIST diameter: 0.93 (hyperdense), 0.95(hypodense), 0.94 (heterogeneous); WHO area: 0.95, 0.98, 0.93]. Semi-automatic measurement of hyperdense, hypodense and heterogeneous liver metastases showed reliable results on standard axial reconstructions in comparison to manual quantification.

Full motion and flow field recovery from echo Doppler data.

We present a new computational method for reconstructing a vector velocity field from scattered, pulsed-wave ultrasound Doppler data. The main difficulty is that the Doppler measurements are incomplete, for they do only capture the velocity component along the beam direction. We thus propose to combine measurements from different beam directions. However, this is not yet sufficient to make the problem well posed because 1) the angle between the directions is typically small and 2) the data is noisy and nonuniformly sampled. We propose to solve this reconstruction problem in the continuous domain using regularization. The reconstruction is formulated as the minimizer of a cost that is a weighted sum of two terms: 1) the sum of squared difference between the Doppler data and the projected velocities 2) a quadratic regularization functional that imposes some smoothness on the velocity field. We express our solution for this minimization problem in a B-spline basis, obtaining a sparse system of equations that can be solved efficiently. Using synthetic phantom data, we demonstrate the significance of tuning the regularization according to the a priori knowledge about the physical property of the motion. Next, we validate our method using real phantom data for which the ground truth is known. We then present reconstruction results obtained from clinical data that originate from 1) blood flow in carotid bifurcation and 2) cardiac wall motion.

Multiscale motion mapping: a novel computer vision technique for quantitative, objective echocardiographic motion measurement independent of Doppler: first clinical description and validation.

BACKGROUND: Objective, quantitative, segmental noninvasive/bedside measurement of cardiac motion is highly desirable in cardiovascular medicine, but current technology suffers from significant drawbacks, such as subjectivity of conventional echocardiographic reading, angle dependence of tissue Doppler measurements, radiation exposure by computer tomography, and infrastructure requirements in MRI. We hypothesized that computer vision technology could represent a powerful new paradigm for quantification in echocardiography. METHODS AND RESULTS: We present multiscale motion mapping, a novel computer vision technology that is based on mathematical image processing and that exploits echocardiographic information in a fashion similar to the human visual system. It allows Doppler- and border-independent determination of motion and deformation in echocardiograms at arbitrary locations. Correctness of the measurements was documented in synthetic echocardiograms and phantom experiments. Exploratory case studies demonstrated its usefulness in a series of complex motion analyses that included abnormal septal motion and analysis of myocardial twisting. Clinical applicability was shown in a consecutive series of echocardiograms, in which good feasibility, good correlation with expert rating, and good intraobserver and interobserver concordance were documented. Separate assessment of 2D displacement and deformation at the same location was successfully applied to elucidate paradoxical septal motion, a common clinical problem. CONCLUSIONS: This is the first clinical report of multiscale motion mapping, a novel approach to echocardiographic motion quantification. For the first time, full 2D echocardiographic assessment of both motion and deformation is shown to be feasible. Overcoming current limitations, this computer vision-based technique opens a new door to objective analysis of complex heart motion.

Spatio-temporal nonrigid registration for ultrasound cardiac motion estimation.

We propose a new spatio-temporal elastic registration algorithm for motion reconstruction from a series of images. The specific application is to estimate displacement fields from two-dimensional ultrasound sequences of the heart. The basic idea is to find a spatio-temporal deformation field that effectively compensates for the motion by minimizing a difference with respect to a reference frame. The key feature of our method is the use of a semi-local spatio-temporal parametric model for the deformation using splines, and the reformulation of the registration task as a global optimization problem. The scale of the spline model controls the smoothness of the displacement field. Our algorithm uses a multiresolution optimization strategy to obtain a higher speed and robustness. We evaluated the accuracy of our algorithm using a synthetic sequence generated with an ultrasound simulation package, together with a realistic cardiac motion model. We compared our new global multiframe approach with a previous method based on pairwise registration of consecutive frames to demonstrate the benefits of introducing temporal consistency. Finally, we applied the algorithm to the regional analysis of the left ventricle. Displacement and strain parameters were evaluated showing significant differences between the normal and pathological segments, thereby illustrating the clinical applicability of our method.

Variational image reconstruction from arbitrarily spaced samples: a fast multiresolution spline solution.

We propose a novel method for image reconstruction from nonuniform samples with no constraints on their locations. We adopt a variational approach where the reconstruction is formulated as the minimizer of a cost that is a weighted sum of two terms: (1) the sum of squared errors at the specified points and (2) a quadratic functional that penalizes the lack of smoothness. We search for a solution that is a uniform spline and show how it can be determined by solving a large, sparse system of linear equations. We interpret the solution of our approach as an approximation of the analytical solution that involves radial basis functions and demonstrate the computational advantages of our approach. Using the two-scale relation for B-splines, we derive an algebraic relation that links together the linear systems of equations specifying reconstructions at different levels of resolution. We use this relation to develop a fast multigrid algorithm. We demonstrate the effectiveness of our approach on some image reconstruction examples.

Myocardial motion analysis from B-mode echocardiograms.

The quantitative assessment of cardiac motion is a fundamental concept to evaluate ventricular malfunction. We present a new optical-flow-based method for estimating heart motion from two-dimensional echocardiographic sequences. To account for typical heart motions, such as contraction/expansion and shear, we analyze the images locally by using a local-affine model for the velocity in space and a linear model in time. The regional motion parameters are estimated in the least-squares sense inside a sliding spatiotemporal B-spline window. Robustness and spatial adaptability is achieved by estimating the model parameters at multiple scales within a coarse-to-fine multiresoluion framework. We use a wavelet-like algorithm for computing B-spline-weighted inner products and moments at dyadic scales to increase computational efficiency. In order to characterize myocardial contractility and to simplify the detection of myocardial dysfunction, the radial component of the velocity with respect to a reference point is color coded and visualized inside a time-varying region of interest. The algorithm was first validated on synthetic data sets that simulate a beating heart with a speckle-like appearance of echocardiograms. The ability to estimate motion from real ultrasound sequences was demonstrated by a rotating phantom experiment. The method was also applied to a set of in vivo echocardiograms from an animal study. Motion estimation results were in good agreement with the expert echocardiographic reading.

Multiresolution moment filters: theory and applications.

We introduce local weighted geometric moments that are computed from an image within a sliding window at multiple scales. When the window function satisfies a two-scale relation, we prove that lower order moments can be computed efficiently at dyadic scales by using a multiresolution wavelet-like algorithm. We show that B-splines are well-suited window functions because, in addition to being refinable, they are positive, symmetric, separable, and very nearly isotropic (Gaussian shape). We present three applications of these multiscale local moments. The first is a feature-extraction method for detecting and characterizing elongated structures in images. The second is a noise-reduction method which can be viewed as a multiscale extension of Savitzky-Golay filtering. The third is a multiscale optical-flow algorithm that uses a local affine model for the motion field, extending the Lucas-Kanade optical-flow method. The results obtained in all cases are promising.

Automatic contrast phase estimation in CT volumes.

We propose an automatic algorithm for phase labeling that relies on the intensity changes in anatomical regions due to the contrast agent propagation. The regions (specified by aorta, vena cava, liver, and kidneys) are first detected by a robust learning-based discriminative algorithm. The intensities inside each region are then used in multi-class LogitBoost classifiers to independently estimate the contrast phase. Each classifier forms a node in a decision tree which is used to obtain the final phase label. Combining independent classification from multiple regions in a tree has the advantage when one of the region detectors fail or when the phase training example database is imbalanced. We show on a dataset of 1016 volumes that the system correctly classifies native phase in 96.2% of the cases, hepatic dominant phase (92.2%), hepatic venous phase (96.7%), and equilibrium phase (86.4%) in 7 seconds on average.