Laurent Sarry

Three-dimensional tracking of coronary arteries from biplane angiographic sequences using parametrically deformable models

A new method for coronary artery tracking in biplane digital subtraction is presented. The dynamic tracking of nonrigid objects from two views is achieved using a generalization of parametrically deformable models. Three-dimensional (3-D) Fourier descriptors used for shape representation are obtained from the two-dimensional (2-D) descriptors of the projections. A new constraint inferred from epipolar geometry is applied to the contour model. Direct 3-D tracking is compared with the classical approach in two steps: independent 2-D tracking in each of the two projection planes; 3-D reconstruction using the epipolar constraint. Convergence quality and accuracy of the 3-D reconstruction are analyzed for several sequences showing different displacement amplitudes, deformation rates and image contrasts.

In Vivo Supervised Analysis of Stent Reendothelialization From Optical Coherence Tomography

The aim of this study is to interactively assess reendothelialization of stents at an accuracy of down to a few micrometer by analyzing endovascular optical coherence tomography (OCT) sequences. Vessel wall and stent struts are automatically detected by using morphological, gradient, and symmetry operators coupled with active contour models; alerts are issued to ask for user supervision over some extreme irregular geometries caused by thrombotic lesions or dissections. A complete distance map is then computed from sparse distances measured between wall and struts. Missing values are interpolated by thin-plate spline (TPS) functions. Accuracy and robustness are increased by taking into account the inhomogeneity of data points and integrating in the same framework orthogonalized forward selection of support points, optimal selection of regularization parameters by generalized cross-validation, and rejection of detection outliers. Validation is performed on simulated data, phantom acquisitions and 11 typical in vivo OCT sequences. The comparison against manual expert measurements demonstrates a bias of the order of OCT resolution (less than 10 ¿m) and a standard deviation of the order of the strut width (less than 150 ¿m ).

Nonsupervised Ranking of Different Segmentation Approaches: Application to the Estimation of the Left Ventricular Ejection Fraction From Cardiac Cine MRI Sequences

A statistical methodology is proposed to rank several estimation methods of a relevant clinical parameter when no gold standard is available. Based on a regression without truth method, the proposed approach was applied to rank eight methods without using any a priori information regarding the reliability of each method and its degree of automation. It was only based on a prior concerning the statistical distribution of the parameter of interest in the database. The ranking of the methods relies on figures of merit derived from the regression and computed using a bootstrap process. The methodology was applied to the estimation of the left ventricular ejection fraction derived from cardiac magnetic resonance images segmented using eight approaches with different degrees of automation: three segmentations were entirely manually performed and the others were variously automated. The ranking of methods was consistent with the expected performance of the estimation methods: the most accurate estimates of the ejection fraction were obtained using manual segmentations. The robustness of the ranking was demonstrated when at least three methods were compared. These results suggest that the proposed statistical approach might be helpful to assess the performance of estimation methods on clinical data for which no gold standard is available.

Monitoring of polyethylene wear in nonmetal-backed acetubular cups by digitized anteroposterior pelvic radiography

The aim of this study was to assess polyethylene wear in a total hip prosthesis by digitized radiography of the whole pelvis in the anteroposterior (AP) plane. The three-dimensional (3-D) pose of the nonmetal-backed acetubular cup, materialized by its metal ring and the femoral head made of metal or ceramic, was estimated using iterative algebraic algorithms with inner bias correction and bootstrapping for variance reduction. Points of interest were obtained by maximizing the correlation between sampled density profiles and 3-D geometric models degraded by the modulation transfer function (MTF) of the radiographic system and the film scanner. The error in the maximal correlation estimate were inferred from noise power spectra (NPS) and allowed the calculation of the point covariance matrix. Both NPS and MTF were modeled for each stage and estimated using least-square fitting of the overall NPS model to the autospectral density function calculated in stationary regions. Comparison of the radiographic time series was made possible by the high accuracy level and 3-D matching from the cup orientation. The feasibility of the full 3-D measurement, the assumption of negligible lateral wear and its influence on AP wear are discussed on simulated and real radiographic data.

New electrophysiological mapping combined with MRI in parkinsonian's subthalamic region.

The subthalamic nucleus (STN) is the main target for deep brain stimulation in Parkinson’s disease. We analysed the relationships between magnetic resonance imaging (MRI) anatomy and spontaneous neuronal activity to confirm the potential of microelectrode recordings to assist in determining the optimal surgical target. Ten bilateral surgeries were performed after 1.5‐T (T2‐weighted) anatomical MRI identification of the STN, zona incerta (ZI), Forel’s field H2 (H2) and substantia nigra (SN). Spontaneous neuronal activity was recorded simultaneously along the distal 10 mm on a central track (optimally covering the STN) and a 2‐mm anterior track. We calculated off‐line mean firing rate and burst frequency on 248 neurons clustered according to anatomical structure. Subjective visual analysis of signal was also realized on‐line, during surgery, to classify patterns of activity. Mean firing rate and burst frequency increased from H2–ZI to SN. The mean firing rate was higher in SN only using paired comparison (SN vs. its neighbours). The burst frequency was lower in H2 than in SN; using comparison with neighbours, it was lower in H2 and ZI. An irregular high activity (type 2C) was more often detected in STN and SN than in H2 and ZI. Anatomical boundaries and unitary recordings appear to be linked, supporting the ability of MRI to provide a detailed anatomy. Electrophysiological mapping combined with MRI is a useful tool for precise targeting in the subthalamic region.

Improved Estimation of Cardiac Function Parameters Using a Combination of Independent Automated Segmentation Results in Cardiovascular Magnetic Resonance Imaging.

This work aimed at combining different segmentation approaches to produce a robust and accurate segmentation result. Three to five segmentation results of the left ventricle were combined using the STAPLE algorithm and the reliability of the resulting segmentation was evaluated in comparison with the result of each individual segmentation method. This comparison was performed using a supervised approach based on a reference method. Then, we used an unsupervised statistical evaluation, the extended Regression Without Truth (eRWT) that ranks different methods according to their accuracy in estimating a specific biomarker in a population. The segmentation accuracy was evaluated by estimating six cardiac function parameters resulting from the left ventricle contour delineation using a public cardiac cine MRI database. Eight different segmentation methods, including three expert delineations and five automated methods, were considered, and sixteen combinations of the automated methods using STAPLE were investigated. The supervised and unsupervised evaluations demonstrated that in most cases, STAPLE results provided better estimates than individual automated segmentation methods. Overall, combining different automated segmentation methods improved the reliability of the segmentation result compared to that obtained using an individual method and could achieve the accuracy of an expert.

Variational myocardial tracking from CINE-MRI with non-linear regularization

We present a new motion estimation approach for cardiac Magnetic Resonance Imaging (MRI) data from a variational framework. The improved performance of variational approach has been achieved by designing a new regularization term that properly handles motion discontinuities. This approach was applied to both synthetic and real data. The quantitative evaluation revealed the superior performance of the proposed method against reference approaches.

Variational myocardial tracking from cine-MRI with non-linear regularization: validation of radial displacements vs. tagged-MRI

We present a new motion estimation approach for cardiac Magnetic Resonance Imaging (Cine-MRI) data from variational framework. The improved performance of this variational approach has been achieved by designing a new regularization term that properly handles motion discontinuities. This approach was applied to both synthetic and real data. The quantitative evaluation revealed that the results of proposed method on cine-MRI correlates with the results given by inTag, reference approach on tagged-MRI.

Estimation of Myocardial Strain and Contraction Phase From Cine MRI Using Variational Data Assimilation

This paper presents a new method to estimate left ventricle deformations using variational data assimilation that combines image observations from cine MRI and a dynamic evolution model of the heart. The main contribution of the model is that it embeds parameters modeling the contraction / relaxation process. It estimates myocardial motion and contraction parameters simultaneously, providing accurate complementary information for diagnosis. The method was applied to synthetic datasets with known ground truth motion and to 47 patients MRI datasets acquired at three slice locations (base, mid-ventricle and apex). Radial and circumferential strain components were compared to those obtained with a reference tag tracking software, exhibiting good agreement with intraclass correlation coefficients (ICC) above 0.8. Results were also evaluated against wall motion score indices used to assess cardiac kinetics in clinical practice. The assimilation process overcame issues caused by temporal artifacts as a result of the dynamic model, compared to using the observation term alone. Moreover we found that the new dynamic model, consisting of a piecewise transport model acting independently on systole and diastole performed better than the standard continuous transport model, which oversmooths temporal variations. Estimated strain and contraction parameters significantly correlated to clinical scores, making them promising features for diagnosing not only hypokinesia but also dyskinesia.

Radiographic stereometry for non-metal-backed acetubular cups: 3D wear estimation and related uncertainty

The aim of this work was to assess the 3D wear of non-metal-backed acetubular cups from two conventional digitized radiographs. The centers of the femoral head and the metal ring in the equatorial plane of the cup are located using 3D pose techniques for spheres and circles from sampled points in the images. The method used to locate these points of interest and also estimate their covariance is fully described in previous work. The covariance is used to decrease the bias of the pose estimation, while bootstrapping decreases its variance and gives access to the directions of minimal variability between the two centers for each image. These directions are used to produce the final distance after reconstruction. Results are compared with the 2D technique working on a single anteroposterior (AP) radiograph and assuming null lateral wear. Validation is performed on acetubular cups: (a) simulated by Monte-Carlo, (b) implanted on a pelvic model, and (c) acquired on patients just after arthroplasty. The accuracy in wear for simulation increases from about one to four hundredths of a millimeter as caudal and cranial absolute angulations decreases from 45 degrees to 30 degrees . It is more difficult to assess for real prostheses, but results are shown to lie within the manufacturers dimensional tolerances. Globally, the access to the lateral wear is obtained at the expense of confidence in the global linear wear (0.06 mm), which is doubled with regard to 2D technique but still satisfying for clinical use, whereas the bias of the measurement is decreased.