Comparison of various methods for quantitative evaluation of myocardial infarct volume from magnetic resonance delayed enhancement data.

BACKGROUND: Late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) enables the estimation of myocardial infarct (MI) extent. Nevertheless, manual quantification is time consuming and subjective. We sought to assess MI volume with different quantitative methods in both acute (AMI) and chronic MI (CMI). METHODS: CMR was performed 50 ± 21 h after MI in 52 patients and was repeated 100 ± 21 days later in a subgroup of 34 patients. Then, necrosis volumes were quantified using: 1) manual delineation, 2) automated fuzzy c-means method, and 3) +2 to 6 SD thresholding approaches. Results were compared against peak values of serum Troponin I (TnI), creatine kinase (CK) and left ventricular (LV) functional parameters: LV ejection fraction (LVEF), indexed end-diastolic (EDVi), end-systolic volumes (ESVi) and the number of hypokinetic segments (NbHk). RESULTS: For CMI, quantitative evaluation of infarct size using manual, +2SD, +3 SD and fuzzy c-means provided equivalent results in terms of correlation coefficients for comparisons of MI volumes against LV function parameters (LVEF: r>0.79, p<0.0001; ESVi: r>0.82, p<0.0001, EDVi: r>0.67, p<0.0001, NbHk: r>0.54, p<0.0009). For AMI, +2SD and fuzzy c-means approaches provided higher correlations for comparisons of AMI volumes against biochemical markers (CK: r>0.79, p<0.0001,TnI: r>0.77, p<0.0001) and chronic LV function parameters (LVEF: r>0.82, p<0.0001, NbHk: r>0.59, p<0.0002). CONCLUSIONS: The fuzzy c-means and 2SD methods provided highest correlations with biochemical MI quantification as well as LV function parameters. The fuzzy c-means approach which does not require an arbitrary identification of the remote myocardium is fast and reproducible. It may be clinically useful in the evaluation of patients with MI.

An automatic respiratory gating method for the improvement of microcirculation evaluation: application to contrast-enhanced ultrasound studies of focal liver lesions.

Contrast-enhanced ultrasound (CEUS), with the recent development of both contrast-specific imaging modalities and microbubble-based contrast agents, allows noninvasive quantification of microcirculation in vivo. Nevertheless, functional parameters obtained by modeling contrast uptake kinetics could be impaired by respiratory motion. Accordingly, we developed an automatic respiratory gating method and tested it on 35 CEUS hepatic datasets with focal lesions. Each dataset included fundamental mode and cadence contrast pulse sequencing (CPS) mode sequences acquired simultaneously. The developed method consisted in (1) the estimation of the respiratory kinetics as a linear combination of the first components provided by a principal components analysis constrained by a prior knowledge on the respiratory rate in the frequency domain, (2) the automated generation of two respiratory-gated subsequences from the CPS mode sequence by detecting end-of-inspiration and end-of-expiration phases from the respiratory kinetics. The fundamental mode enabled a more reliable estimation of the respiratory kinetics than the CPS mode. The k-means algorithm was applied on both the original CPS mode sequences and the respiratory-gated subsequences resulting in clustering maps and associated mean kinetics. Our respiratory gating process allowed better superimposition of manually drawn lesion contours on k-means clustering maps as well as substantial improvement of the quality of contrast uptake kinetics. While the quality of maps and kinetics was satisfactory in only 11/35 datasets before gating, it was satisfactory in 34/35 datasets after gating. Moreover, noise amplitude estimated within the delineated lesions was reduced from 62 ± 21 to 40 ± 10 (p < 0.01) after gating. These findings were supported by the low residual horizontal (0.44 ± 0.29 mm) and vertical (0.15 ± 0.16 mm) shifts found during manual motion correction of each respiratory-gated subsequence. The developed technique could be used as a basis for accurate quantification of perfusion parameters for the evaluation and follow-up of patients under antiangiogenic therapies.

An automated quantification of the transmural myocardial infarct extent using cardiac DE-MR images.

Evaluating myocardial viability is an important prognostic factor in the follow-up of infarctions. Delayed Enhancement magnetic resonance (DE-MR) imaging allows precise delineation of the infarct transmural extent. Visual interpretation is the most commonly used method to assess the myocardial infarction (MI) transmural extent. This study proposes to automate the segmentation of the (DE) images prior to the estimation of the extent of infarcted tissue. Indeed the segmentation of the myocardium was performed using cine contraction images which present a high contrast between cavity and myocardium. After the segmentation, the segmental transmurality is estimated on a conventional five point scale. A head to head comparison was performed between visual and quantitative analysis of infarct transmurality on DE-MR imaging. Results on 921 sub-segments (9 patients) showed an absolute agreement of 80% and a relative agreement (with one point difference) of 97%.

[Tissue attenuation in small animals on contrast enhanced ultrasound]. / Prise en compte de l'atténuation ultrasonore en présence d'agents de contraste chez le petit animal.

Despite recent advances in contrast-enhanced ultrasound imaging, evaluation of tissue perfusion with contrast-enhanced ultrasound is still impaired by shadowing effects. These effects are particularly relevant in small animal studies due to high frequency imaging. Current methods of tissue attenuation correction are not suited for contrast-enhanced ultrasound examinations, because microbubble acoustic response to ultrasound waves is far more complex than that of tissues. A method allowing in vivo tissue attenuation correction in the presence of contrast agents is presented.

An automated myocardial segmentation in cardiac MRI.

In this paper we present an automatic approach to segment Cardiac Magnetic Resonance (CMR) images. A preprocessing step that consists in filtering the image using connected operators (area opening and closing filters) is applied in order to homogenize the cavity and solve the problems due to the papillary muscles. Thereby the GVF snake algorithm is applied with one point clicked in the cavity as initialization and an optimized tuning of parameters for the endocardial contour extraction. The epicardial border is then obtained using the endocardium as initialization. The performance of the proposed method was assessed by experimentation on thirty-nine CMR images. A high agreement between manual and automatic contours was obtained with correlation scores of 0.96 for the endocardium and 0.90 for the epicardium. Overlapping percentage, mean and maximum distances between the two contours show a good performance of the method.

On the ability of ultrasound parametric perfusion imaging to predict the area of infarction in acute ischemic stroke.

PURPOSE: Cerebral perfusion deficits in acute ischemic stroke can be detected by means of transcranial harmonic imaging after ultrasound contrast agent bolus injection. We evaluated five different parameters of the bolus kinetics as parametric images and correlated areas of disturbed perfusion with the area of definite infarction. MATERIALS AND METHODS: Perfusion harmonic imaging after SonoVue bolus injection (BHI) was used to investigate 22 patients suffering from acute internal carotid artery infarction. For each subject, we calculated five different images based on the following parameters from the time-intensity curve in each pixel: pixelwise peak intensity (PPI), area under the curve (AUC), positive gradient (PG), time to peak (TTP), and a three factor image from the factor analysis of medical image sequences (FAMIS). The findings in the diencephalic imaging plane were compared with the definite area of infarction, as diagnosed by cranial CT. RESULTS: In predicting the definite area of infarction in follow-up CT, we found the following sensitivities and positive predictive values (PPV): PPI (100 %/95 %), AUC (100 %/90 %), FAMIS (89 %/89 %), PG (84 %/94 %) and TTP (47 %/100 %). The areas of disturbed perfusion in all five types of parametric images correlated significantly with the area of infarction in CT. Images from the FAMIS algorithm and PPI images showed the highest Spearman rank correlation with the area of definite infarction as displayed in CT (both r = 0.76, p < 0.001). Images from the other parameters correlated as follows: PG: r = 0.62 (p = 0.003), AUC: r = 0.53 (p = 0.014), TTP: r = 0.50 (p = 0.021). CONCLUSION: BHI can detect disturbed perfusion in acute hemispheric stroke. In their ability to predict the development of an infarction, intensity-based parameters and FAMIS were determined to have a high sensitivity, and TTP was found to have a high PPV and specificity.

Velocity encoding versus acceleration encoding for pressure gradient estimation in MR haemodynamic studies.

Many methods have been proposed to extract pressure gradient maps from magnetic resonance (MR) images. They were based on the resolution of the haemodynamic model of Navier-Stokes and needed the flow acceleration to be known. Most used velocity data acquisition and computed acceleration from temporal and spatial derivatives of the velocity field. However, MR sequences have been developed in order to acquire the acceleration field directly. Here we compared direct MR measurements of acceleration field components with those calculated from MR velocity acquisitions. Two experimental phantoms were used to separately evaluate the inertial and convective components of the acceleration. Mathematical simulation of the convective phantom further explained the origin of the noise generated by the spatial and temporal derivatives of the velocity data, and the misregistration artefacts due to MR sequences. We found that direct measurement of the acceleration field generates less noise and fewer artefacts than calculation from velocity derivatives.

[Aortic arch shape and flow dynamics]. / Architecture de la crosse aortique et circulation des fluides.

Post-operative deformation of the aortic arch architecture is associated with an increased risk of hypertension following correction of coarctation. In addition to morphological analysis, MRI allows a functional analysis of the thoracic aorta. We report three examples which illustrate the direct relationship between aortic arch morphology and blood flow in the thoracic aorta.

A posteriori respiratory gating in contrast ultrasound for assessment of hepatic perfusion.

An original strategy is proposed to minimize the impact of respiratory motion on hepatic contrast-enhanced ultrasound studies. It is based on the a posteriori triggering of dynamic image sequences. It was tested on perfusion studies acquired with a high temporal resolution (8 images s-1) to enable parametric imaging. A respiratory component was first estimated by independent component analysis. The estimation of the local minima and maxima of this curve enabled us to select two subsets of frames, corresponding to the end-of-inspiration plane and to the end-of-expiration plane. Both subsets were simultaneously analysed using factor analysis of medical image sequences. This method identified the main contrast uptake kinetics and their associated localizations. The global strategy was validated firstly on a simulated study and then applied to 11 patients' studies. In both cases, the frame selection was judged relevant and a necessary preliminary step before applying methods of parametric imaging. In conclusion, the a posteriori gating method that is proposed is a first step towards local quantification of hepatic contrast-enhanced ultrasound studies.

Assessment of left ventricular contraction by parametric analysis of main motion (PAMM): theory and application for echocardiography.

The computerized study of the regional contraction of the left ventricle has undergone numerous developments, particularly in relation to echocardiography. A new method, parametric analysis of main motion (PAMM), is proposed in order to synthesize the information contained in a cine loop of images in parametric images. PAMM determines, for the intensity variation time curves (IVTC) observed in each pixel, two amplitude coefficients characterizing the continuous component and the alternating component; the variable component is generated from a mother curve by introducing a time shift coefficient and a scale coefficient. Two approaches, a PAMM data driven and a PAMM model driven (simpler and faster), are proposed. On the basis of the four coefficients, an amplitude image and an image of mean contraction time are synthesized and interpreted by a cardiologist. In all cases, both PAMM methods allow better IVTC adjustment than the other methods of parametric imaging used in echocardiography. A preliminary database comprising 70 segments is scored and compared with the visual analysis, taken from a consensus of two expert interpreters. The levels of absolute and relative concordance are 79% and 97%. PAMM model driven is a promising method for the rapid detection of abnormalities in left ventricle contraction.

Local reconstruction of stenosed sections of artery using multiple MRA acquisitions.

A method for reconstructing magnetic resonance angiography (MRA) volumes from successive acquisitions is described. The method is based on double oblique acquisitions of highly anisotropic MRA volumes, each of which corresponds to reduced k-space filling. These partial k-spaces are then combined to obtain a 3D k-space adapted to the frequency spread of the angiographic image of the stenosis. The SNR-resolution compromise of MRA is thus improved by focusing the acquisition on the most relevant k-space regions. The reconstruction is performed directly in k-space by averaging the partial k-spaces. The feasibility of the method was demonstrated in studies on a Lucite stenosis phantom, on MRAs of carotid arteries using three bolus injections, and on MRAs of renal arteries using a single contrast injection.

Multigate doppler signal analysis using 3-D regularized long AR modelling.

Autoregressive (AR) modelling has already been proposed as an alternative to fast Fourier transform to process ultrasound (US) Doppler signals. Previous works introduced long AR models, set up under a regularization framework. The latter may be in 1-D (frequency) or 2-D (frequency and space or time). This study generalizes the spectrum regularization in the three dimensions frequency, space and time. The problem of the penalization function is addressed, and a new convex solution is proposed, taking into account possible nonstationarity of the Doppler signal. The parameter tuning is based on simulations using a standard Doppler signal model. The first results show that this processing improves the spectral estimation, and is well suited to flow interpretation.

Validation of myocardial perfusion reserve measurements using regularized factor images of H(2)(15)O dynamic PET scans.

UNLABELLED: The use of H(2)(15)O PET scans for the measurement of myocardial perfusion reserve (MPR) has been validated in both animal models and humans. Nevertheless, this protocol requires cumbersome acquisitions such as C(15)O inhalation or (18)F-FDG injection to obtain images suitable for determining myocardial regions of interest. Regularized factor analysis is an alternative method proposed to define myocardial contours directly from H(2)(15)O studies without any C(15)O or FDG scan. The study validates this method by comparing the MPR obtained by the regularized factor analysis with the coronary flow reserve (CFR) obtained by intracoronary Doppler as well as with the MPR obtained by an FDG acquisition. METHODS: Ten healthy volunteers and 10 patients with ischemic cardiopathy or idiopathic dilated cardiomyopathy were investigated. The CFR of patients was measured sonographically using a Doppler catheter tip placed into the proximal left anterior descending artery. The mean velocity was recorded at baseline and after dipyridamole administration. All subjects underwent PET imaging, including 2 H(2)(15)O myocardial perfusion studies at baseline and after dipyridamole infusion, followed by an FDG acquisition. Dynamic H(2)(15)O scans were processed by regularized factor analysis. Left ventricular cavity and anteroseptal myocardial regions of interest were drawn independently on regularized factor images and on FDG images. Myocardial blood flow (MBF) and MPR were estimated by fitting the H(2)(15)O time-activity curves with a compartmental model. RESULTS: In patients, no significant difference was observed among the 3 methods of measurement-Doppler CFR, 1.73 +/- 0.57; regularized factor analysis MPR, 1.71 +/- 0.68; FDG MPR, 1.83 +/- 0.49-using a Friedman 2-way ANOVA by ranks. MPR measured with the regularized factor images correlated significantly with CFR (y = 1.17x - 0.30; r = 0.97). In the global population, the regularized factor analysis MPR and FDG MPR correlated strongly (y = 0.99x; r = 0.93). Interoperator repeatability on regularized factor images was 0.126 mL/min/g for rest MBF, 0.38 mL/min/g for stress MBF, and 0.34 for MPR (19% of mean MPR). CONCLUSION: Regularized factor analysis provides well-defined myocardial images from H(2)(15)O dynamic scans, permitting an accurate and simple measurement of MPR. The method reduces exposure to radiation and examination time and lowers the cost of MPR protocols using a PET scanner.

Detection of motion in hybrid PET/SPECT imaging based on the correlation of partial sinograms.

This paper describes a motion detection method specific to hybrid positron emission tomography/single photon emission computed tomography systems. The method relies on temporal fractionation of the acquisition into three data sets followed by an algorithm based on the cross correlation (CC) of partial sinograms from successive sets at different rotations of the camera. Spatial inconsistencies due to motion are detected by decreases in the CC between two sets. This permits to separate data into premotion and postmotion sets of consistent data that are reconstructed independently then registered and summed. Rigid motions greater than 1-cm translation or 10 degrees rotation were detected with this method from experimental data obtained by manually moving phantoms made of radioactive spheres as well as from a patient lung study corrupted by artificial motion. The different motion studies showed that the image contrast does not seem to be a limiting factor and that the motion is best detected when the gantry is parallel to the direction of motion. The registration and fusion of the reconstructed premotion and postmotion sets lead in all cases to a reduction of the motion artifacts and an increase in signal-to-noise ratio.

Adaptive and self-evaluating registration method for myocardial perfusion assessment.

With the advent of ultra-fast MRI, it is now possible to assess non-invasively regional myocardial perfusion with multislice coverage and sub-second temporal resolution. First-pass contrast enhanced studies are acquired with ECG-triggering and breath holding. Nevertheless, some respiratory induced movements still remain. Myocardial perfusion can be assessed locally by parametric imaging methods such as Factor Analysis of Medical Image Sequences (FAMIS), provided that residual motion can be corrected. An a posteriori registration method implemented in the image domain is proposed. It is based on an adaptive registration model of the heart combining three elementary shapes (left ventricle, right ventricle and pericardium). The registration procedure is performed on a potential map derived from the distance map. To evaluate the quality of the registration procedure a superimposition score between the registration model and the contour automatically extracted in the sequence is proposed. Rigid transformation hypotheses and registration analysis provide an efficient and automatic method which allows the rejection of outlier images, such as: out of synchronisation images, out of plane acquisitions. When compared to a manual registration method, this approach reduces processing time and requires a minimal intervention from the operator. The proposed method performs registration with a subpixel accuracy. It has been successfully applied to simulated images and clinical data. It should facilitate the use of MR first-pass perfusion studies in clinical practice.

In vitro quantification of flow using continuous infusion of Levovist and pairs of harmonic power Doppler images.

To evaluate the potential of harmonic power Doppler to quantify perfusion using a continuous infusion of contrast, two dialysis cartridges were perfused with different flow rates adjusted between 0 to 300 mL/min, corresponding to flow ratios comprised between 300:0 and 150:150. The contrast agent (Levovist, Schering) was injected at constant rates (0.6 to 5 g/h). Sequential pairs of images showing simultaneously the cross-sections of the two filters were acquired with a HDI 5000 (ATL) and the Doppler data were processed with HDI lab software (ATL). The absolute values of the signal in the different regions-of-interest (ROI) were not closely related to flow rate. At the opposite, the rapid signal decrease between the first and the second image of each pair was inversely proportional to the flow rate. An index of perfusion [PerI = image 1/(image 1 -- image 2)] was defined. It correlated closely with the absolute and relative flow rates. For the latter, the slopes of regression were found to be independent of the infusion rate of Levovist. Thus, the use of pairs of images combined with a continuous infusion of Levovist provide a quantification of perfusion.

Multidimensional MR mapping of multiple components of velocity and acceleration by fourier phase encoding with a small number of encoding steps.

Previous studies have shown that the multi-step approach of velocity or acceleration encoding is highly efficient in terms of the signal-to-noise ratio per unit time. This work describes a multidimensional extension of this method for simultaneously measuring multiple components of velocity and acceleration with a few encoding steps. N flow dimensions were encoded with an ND-matrix, obtained by combining the various flow-encoding gradients. The small matrix obtained with as few as two encoding steps can be extended by zero-filling in all N dimensions and using ND-Fourier transformation to obtain the maximum of the resulting peak in the ND-matrix, which gives simultaneously all the components of velocity and/or acceleration. The processing time was shortened by using a method of phase computation that gives the same precision as Fourier transformation, but is much faster. A rotating disk was used to show that the velocity-to-noise ratio increases with the number of dimensions acquired, demonstrating the efficiency of multidimensional flow measurements. The feasibility of the method is illustrated by 3D maps of the myocardium velocity, and 2D measurement of velocity and acceleration in the ascending aorta-both obtained by multidimensional phase encoding in volunteers.

Improved estimation of velocity and flow rate using regularized three-point phase-contrast velocimetry.

We improved the three-point phase-contrast method by regularization of MR velocity data after acquisition of a low velocity-to-noise ratio (VNR) velocity image and a high VNR aliased velocity image. The phase unwrapping algorithm is based on the assumed correlation of the velocity of adjacent flow voxels on the low VNR and the unaliased high VNR images. We used Fourier encoding with eight velocity-encoding gradient steps to obtain reference velocity images of the aorta from five subjects (274 images) and compared them with the phase-contrast and three-point phase-contrast images with and without regularization. The VNR of the regularized velocity image was improved by 9.1 dB and the VNR of the three-point phase-contrast velocity image was improved by 0.7 dB with respect to the low first moment velocity image. Corresponding improvements of 9 dB and 3.7 dB were obtained for the estimations of instantaneous flow rate. Magn Reson Med 44:122-128, 2000.

Fluid mechanics of regurgitant jets and calculation of the effective regurgitant orifice in free or complex configurations.

The velocity fields of turbulent jets can be described using a single formula which includes two empirical constants: k(core) determining the length of the central core and k(turb) the jet widening. Flow models simulating jet adhesion, confinement and noncircular orifices were studied using laser Doppler anemometer and the modifications of the constants were derived from series of velocity profiles. In circular free jets, k(core) was found equal to 4.1 with a variability of 1.4%. In complex configurations, its variability was equal to 15.2%. For k(turb), the value for free circular jets was of 45.2 with a variability of 6.0% and this variability in complex configurations was significantly higher (30. 1%, p=0.025). The correlation between the actual orifice size and the jet extension was poor (r=0.52). However, the almost constant value of k(core) allowed to define a new algorithm calculating the regurgitant orifice diameter with the use of outlines of the jet image (r=0.89). In conclusion, the fluid mechanics of regurgitant jets is modified in complex configurations but, due to the relative independency of the central core, velocity fields could be used to evaluate the dimensions of the effective regurgitant orifice.

In vitro flow quantification with contrast power Doppler imaging.

To evaluate the effectiveness of contrast harmonic (power Doppler imaging) as an ultrasonic modality to quantify flow, an in vitro model of perfusion was studied using Optison, a second-generation ultrasound (US) contrast agent. The in vitro model was made of two dialysis cartridges placed parallel and allowed absolute and relative flow quantification on both tube (entry lines) and tissue (cartridges) simulations. Video intensity curves were generated using intermittent harmonic power Doppler imaging after bolus injection of contrast. Correlation between flow and different parameters extracted from time-intensity curves and previously defined as indicators of flow was established for both tissue and entry lines, for flow rates ranging from 0 to 400 mL/min. Single-compartment equations were also tested on the model. A good correlation for the tissue model was observed between absolute flow and onset time (O), time to maximal enhancement (TME), peak intensity (P), area under the curve (AUC), and maximal ascending slope (S) parameters, with a r = 0.94, 0.94, 0.91, 0.92 and 0.92, respectively. The correlation for O, TME, P and AUC parameters was r = 0.86, 0.90, 0.78 and 0.82, respectively for entry lines. The correlation for tissue model and entry line was slightly improved when comparing flow ratios with peak ratios (P1/P2) and slope ratios (S1/S2) (r = 0.95 and 0.94). Flow calculation using the gradient-relationship method also showed a good correlation (r = 0.88) with the experimental flow. The results obtained indicated that absolute and relative quantification of flow using PDI is feasible in tube and tissue models. Several clinical applications, namely in myocardial, hepatic and renal artery studies, could be derived from these results.