1,721,362 research outputs found
Performances of Different Algorithms for Tracer Kinetics Parameters Estimation in Breast DCE-MRI
No full textObjective of this study was to evaluate the performances of different algorithms for tracer kinetics parameters estimation in breast Dynamic Contrast Enhanced-MRI. We considered four algorithms: two non-iterative algorithms based on impulsive and linear approximation of the Arterial Input Function respectively; and two iterative algorithms widely used for non-linear regression (Levenberg-Marquardt, LM and VARiable PROjection, VARPRO). Per each value of the kinetic parameters within a physiological range, we simulated 100 noisy curves and estimated the parameters with all algorithms. Sampling time, total duration and noise level have been chosen as in a typical breast examination. We compared the performances with respect to the Cramer-Rao Lower Bound (CRLB). Moreover, in order to gain further insight we applied the algorithms to a real breast examination. Accuracy of all the methods depends on the specific value of the parameters. The methods are in general biased: however, VARPRO showed small bias in a region of the parameter space larger than the other methods; moreover, VARPRO approached CRLB and the number of iterations were smaller than LM. In the specific conditions analyzed, VARPRO showed better performances with respect to LM and to non-iterative algorithms
The Use of the Levenberg-Marquardt and Variable Projection Curve-Fitting Algorithm in Intravoxel Incoherent Motion Method for DW-MRI Data Analysis
No full textThe objective of this study was to evaluate the performances of different algorithms for diffusion parameters estimation in intravoxel incoherent motion method for diffusion-weighted magnetic resonance imaging (DW-MRI) data analysis. Traditionally, the method of non-linear least squares analysis by means of Levenberg–Marquardt algorithms has been used to estimate the parameters obtained from exponential decay data. In this study, we evaluated the Variable Projection curve-fitting algorithm and the performance of two non-linear regression methods when single and multiple starting points were used. Analysis was done on simulation data to which different amounts of Gaussian noise had been added. The performance of two non-linear regression methods was compared using the residual sum of squares and the number of failures in data fitting. We conclude that the VarPro algorithm is superior to the LM algorithm for curve fitting in intravoxel incoherent motion method for DW-MRI data analysis
A geometrical perspective on the 3TP method in DCE-MRI
No full textThe 3TP (three time points) method has been proposed for producing high-resolution pseudo-coloured maps related to the angiogenic activity in breast DCE-MRI. In the original formulation of the method, the three time-points have been chosen on an empirical basis: the algorithm for benign/malignant/uncertain classification of a voxel was to be as simple as possible and the corresponding regions of the parameters space (according to the Tofts' model) were to occupy approximately the same area. Since its inception, the method has been largely used in clinical environment, due to its simplicity and soundness. However, as only three time-points are used to evaluate the characteristics of the complex time-course of the contrast medium within capillaries, noise can result in voxel misclassification, as we show in this study.In this paper we analysed the performances of the method from a geometrical perspective, based on the concept of confidence region, and we proposed an 'optimal' choice of the three time-points, in order to reduce the misclassification to a minimum. Comparing the original 3TP method with our proposal on the basis of misclassification rate, our results show that the modified 3TP method can lead to better performance. Preliminary results on real data have been also reported. Moreover, our proposal has a sounding mathematical basis and is easily generalisable to the case of more than two parameters and to other modalities such as DCE-CT. (C) 2014 Elsevier Ltd. All rights reserved
Breast contrast-enhanced MR imaging: semiautomatic detection of vascular map
No full textBackground
The diagnostic value of breast vascular maps using contrast-enhanced MR imaging has recently been explored. We propose a semiautomatic method to obtain breast vascular maps and to measure the number of blood vessels in the breast.
Methods
From January 2011 to December 2013, 188 patients underwent breast contrast-enhanced MRI; patients with unilateral and histopathologically confirmed breast lesions were included in this study; 123 patients had malignant lesions and 65 patients had benign tissue diagnoses. Breast semiautomatic vascular map detection was performed using Hessian matrix-based method and morphologic operators. Blood vessels detection was compared with radiologic interpretation findings to evaluate algorithm goodness. Increase in vascularity associated with ipsilateral cancer was also assessed. Chi square test was used to observe statistically significant difference.
Results
A total of 1315 blood vessels were identified using semiautomatic procedure; 1034 were correctly classified (78.7 %), 261 (19.8 %) were incorrectly classified, and 20 (1.5 %) were missing. A significant association was found between one-sided increased breast vascularity and ipsilateral malignancy (p < 0.001).
Conclusions
In conclusion, detection of vascularity increase as risk factor for developing breast cancer could be performed with semiautomatic vascular mapping of contrast-enhanced MR imaging
Dynamic contrast-enhanced MRI in breast cancer: A comparison between distributed and compartmental tracer kinetic models
No full textBackground/objectives: Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) is widely used in tumor
diagnosis, staging and assessment of therapy response for different types of tumors, thanks to its capability to provide
important functional information about tissue microvasculature. Tracer kinetic models used for estimating microcirculatory
parameters can be broadly categorized as conventional compartmental (CC) or distributed- parameter (DP)
models. While DP models seem to be more realistic, CC models (in particular the Tofts and the Brix models) have been
widely used in clinical investigations over the past two decades. However, to date there is no direct comparison of CC vs
DP models on real breast DCE-MRI data; moreover, a direct comparison between Tofts and Brix models, has not yet been
reported on real breast data. Therefore, the purpose of this study was two-fold: on the one hand we analyzed the
performance, on real breast DCE-MRI data, of CC vs DP models in terms of goodness-of-fit metrics; on the other hand we
compared Tofts and Brix models on the basis of real breast DCE-MRI data.
Methods: Three models were compared: two CC models (the Tofts and the Brix models) and one DP model (the ATH
model). We gathered data in two different scenarios: DCE-MRI with high temporal resolution obtained by means of a
k-space under-sampling and data sharing method known as Time-resolved angiography With Stochastic Trajectories
(TWIST) and DCE-MRI with low temporal resolution obtained by means of the Spoiled Gradient-Echo k-space scheme
known as Fast Low Angle Shot (FLASH). The performances of the three models were evaluated by means of three
goodness-of-fit metrics: the Residual Sum of Squares, the Bayesian Information Criterion and the Akaike Information
Criterion on four breast DCE-MRI examinations.
Results: Although not conclusive, the results of this study suggest that the ATH model can achieve better fit in comparison
to the Tofts and Brix models for TWIST data; and that the Brix model can achieve better fit with respect to the Tofts model
for FLASH data.
Conclusion: Given the current typical settings of clinical breast DCE-MRI examinations, there seems not to be a clear
advantage, in terms of goodness-of-fit, of ATH with respect to Tofts and Brix models; moreover, at lower temporal
resolution the Brix model can achieve better fit than the Tofts model
Electrocardiogram Pattern Recognition and Analysis Based on Artificial Neural Networks and Support Vector Machines: A Review
Full text linkComputer systems for Electrocardiogram (ECG) analysis support the clinician in tedious tasks (e.g., Holter ECG monitored in Intensive Care Units) or in prompt detection of dangerous events (e.g., ventricular fibrillation). Together with clinical applications (arrhythmia detection and heart rate variability analysis), ECG is currently being investigated in biometrics (human identification), an emerging area receiving increasing attention. Methodologies for clinical applications can have both differences and similarities with respect to biometrics. This paper reviews methods of ECG processing from a pattern recognition perspective. In particular, we focus on features commonly used for heartbeat classification. Considering the vast literature in the field and the limited space of this review, we dedicated a detailed discussion only to a few classifiers (Artificial Neural Networks and Support Vector Machines) because of their popularity; however, other techniques such as Hidden Markov Models and Kalman Filtering will be also mentioned
Influence of Parameterization on Tracer Kinetic Modeling in DCE-MRI
No full textTracer kinetic modeling in dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is commonly
performed using least squares algorithms. The convergence of such algorithms and the repeatability of the estimates are
affected by the curvature of the model’s expectation surface. An adequate choice of the parameterization can reduce
curvature and thus improve parameter estimation. This study analyzes the influence of two parameterizations on the
curvature of the Tofts model. The influences of the total acquisition time and the sampling period are evaluated.
Analysis results show that using (Ktrans, ve
) can significantly reduce the curvature in a large area of the parameter space,
suggesting that curvature analysis could guide the choice of the best local parameterization in Gauss-Newton-based
algorithms. In addition, increasing the total acquisition time and decreasing the sampling period reduce the curvature.
However, only slight improvements are obtained for a total time longer than about 6 min and a sampling period shorter
than approximately 10
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Segmentation and classification of breast lesions using dynamic features in Dynamic Contrast Enhanced-Magnetic Resonance Imaging
No full textThe aim of this study is to propose an approach, based on
Multi Layer Perceptron classification of dynamic and textural features, for breast lesions segmentation and classification using Dynamic Contrast Enhanced-Magnetic Resonance Imaging data. We compared the performance obtainable with dynamic, textural and spatio-temporal features. In particular, 98 dynamic features, 60 textural features and 72 spatio-temporal features were considered. The
dataset included 20 breast lesions, 10 benign and 10 malignant. The performance of lesion segmentation have been
evaluated with respect to manual segmentation provided by
an expert radiologist. Results of lesion classification were
compared to histological findings. Our results indicate that
Multi Layer Perceptron can achieve better results in terms
of sensitivity, specificity and accuracy when dynamic features are considered both for lesion segmentation and classification (accuracy of 91 % and 70 %, respectively)
- …
