M. Baroni

Digital curvature estimation for left ventricular shape analysis

Abstract Curvature estimation from noisy digital curves is addressed for shape analysis of angiographic left ventricle (LV) images. A simple and fast algorithm, based on Fourier series approximation, changes the harmonic number and filter window along the closed (or partially open) contours to optimize smoothness and reconstruction errors. The curvatures estimated on a set of ventricle-shaped models are significantly closer to their analytical values than those assessed by four other methods. Preliminary results on LV data are presented.

Effects of reconstructive surgery for left ventricular anterior aneurysm on ventriculoarterial coupling

Objective To investigate left ventricular elastance (Emax) and effective arterial elastance (Ea) in postinfarction left ventricular aneurysm and evaluate their role in left ventricular function improvement after endoventricular circular patch plasty (EVCPP). Ventriculoarterial coupling has never been studied in these patients. Patients 22 consecutive patients (49 to 73 years) with left ventricular anterior aneurysm. Methods Haemodynamic studies were done before and two weeks after EVCPP. Ventriculography was performed during atrial pacing (100 beats/min). Pressure/volume loops were analysed and derived parameters measured. Emax was estimated by applying the “single beat” method. Ea was calculated as end systolic pressure/stroke volume. Results Left ventricular volumes and Ea decreased after surgery: end diastolic volume index from mean (SD) 155 (53) to 106 (29); end systolic volume index from 112 (51) to 62 (30) ml/m2 (both p < 0.0001); Ea from 1.65 (0.70) to 1.39 (0.41) mm Hg/ml (p = 0.04). Ejection fraction and Emax increased, without significant changes in stroke volume and work. The decrease in Ea was directly correlated with its preoperative value. The time interval between left ventricular pressure upstroke and peak systolic pressure decreased, from 237 (39) to 191 (41) ms (p < 0.0001), paralleling morphological changes in pressure tracings. Conclusions After EVCPP, ventriculoarterial coupling improves because of the fall in Ea caused by end systolic pressure reduction. The improvement is related to aortic pressure waveform changes and improved relaxation.

Texture Classification of Retinal Layers in Optical Coherence Tomography

This work investigates the ability of texture analysis to yield discrimination of retinal tissue layers in the images provided by Optical Coherence Tomography (OCT). In fact, this relatively new imaging technology allows noninvasive visualization of retinal layers. Their segmentation is a prerequisite for any computer method that aims to objectively extract valuable information, regarding the condition and the progression of disease and therapy. Since the regularities of biological tissue can be captured by texture analysis in a straightforward way, a computer approach is proposed based on co-occurence matrices and artificial neural networks (ANN) for the classification and analysis of single retinal layers. A subset of ten normal eyes has been used for the training phase, and another subset of ten normal eyes has been used for testing the system performance. For inner retinal layers, accuracy was 79%, specificity about 71% and sensibility was 87%. Slightly lower values were obtained for outer retinal layers. These preliminary results suggest that this approach may be useful as a prototype system for a quantitative characterization of retinal tissue.

Regional and Temporal Nonuniformity of Shape and Wall Movement in the Normal Left Ventricle

Asymmetry of left ventricular (LV) shape and asynchrony of regional LV movement have been described in the normal human heart, but never correlated to each other. In 16 normal subjects, right anterior oblique ventriculography was used to obtain volumes, regional wall motion (centerline method) and curvature (windowed Fourier series approximation of contours) over the entire cardiac cycle. The apex had the greatest curvature, while the posterior wall had a negative end-diastolic curvature that decreased further at early-systole and became positive at end-systole. The anterior region had the greatest and the anteroapical region the least fractional shortening. Asynchrony was evident as a delayed contraction of the infero- and anteroapical regions, and as a greater rate of late-systolic shortening of the anterior wall than that of the apex. Shape changes and shortening were dyssynchronous in the apical regions where the greatest changes occurred at early diastole. Temporal and regional nonuniformity of shape and movement exists in normal subjects. Dyssynchrony between shape and regional contraction of the apical regions deserves further studies.

Left Ventricular Shape Abnormalities in Inferior Wall Myocardial Infarction

Abstract Many unanswered questions remain concerning the physiopathology of inferior wall myocardial infarction. Expansion and thinning of the involved area (frequently observed in the anterior infarct) are seldom demonstrated in the inferior infarct. 1–3 According to some observations, wall expansion appears related more to the site rather than to the size of the pathologic lesion. 4 Furthermore, wall motion abnormalities are usually less severe in the inferior than anterior infarct, 3,5 even when the extent of myocardial damage and the depression of global left ventricular (LV) function are equivalent. 6 In explaining the higher frequency of wall expansion in the anterior infarct, an important role has been attributed to the involvement of the anteroapical region that mainly contributes to the stretching and thinning of the necrotic area. 7 The aim of this study was to analyze LV shape abnormalities in the inferior infarct by quantitative methods and to relate them to the extent of wall motion alterations and to the severity of LV dysfunction.

Computer evaluation of left ventricular wall motion by means of shape-based tracking and symbolic description

Relevant cardiac pathologies manifest themselves as abnormal movements of left ventricular (LV) myocardial wall. An objective quantification is usually accomplished by computer analysis of temporal sequences of LV contours, as obtained by angiography or echography. However the choice of the reference system for measuring motion is still open to discussion. Simple geometric models cannot deal with the non-uniform myocardial fibre structure, which gives rise to non-rigid movements, asynchronous even in normal subjects. Therefore, a new method, Curvature-Motion (CM), was developed for improving motion assessment. Since LV contour shape and position change smoothly throughout the cardiac cycle, the points of curvature extremes are tracked frame-to-frame and selected by exploiting physiologically-based assumptions; then, the points lying among these landmarks are mapped onto sequential contours, according to local displacements and curvature changes. In this way point-trajectories are allowed to be curvilinear and different in systole and diastole. CM gave no significant differences in estimating the known motion of computer-generated contours (R=0.88), unlike other methods commonly adopted (R<0.80). Moreover, for the evaluation of regional wall motion of a preclassified set of angiographic contours, CM showed a greater specificity (88%) and accuracy (90%) with respect to the centre-line method (respectively 83% and 87%). Finally, a fuzzy logic inference system is proposed for translating significant motion patterns from the quantitative form, as provided by the analysis method, into the linguistic terms used by cardiologists in their clinical examinations. This makes the interpretation of quantitative analysis easier and allows medical users to interact with the system for searching particular properties in a single clinical report as well as in a large database.

Elliptic Fourier approximation with application to left ventricular contour analysis

A computer program for the quantitative analysis of left ventricular wall motion, in standard angiography or echography, was developed based on the center-line model. In order to resolve the uncertainties of the previous implementations, the elliptic Fourier approximation was tested on a large set of left ventricular contour pairs, in various pathological conditions. Fitting a function to the center-line (in the middle of end diastolic and end systolic contours) makes it possible to perform multiple tasks at the same time, in an analytical way and without the need of arbitrary parameters, such as search windows. In doing so, the flexibility and accuracy of this method was demonstrated, more in general, for filtering, interpolation and differentiation of noisy 2D or 3D data. A comparison with standard B-spline was carried out, for both open and closed curves, with sparse or equispaced points. The resulting approximation errors were twice as high as those of the elliptic Fourier decomposition, still preserving smooth direction and curvature. Though already described, this methodology deserves more widespread application.

Left ventricular remodeling in chronic aortic regurgitation

Left ventricular (LV) shape in chronic volume overload due to aortic regurgitation is commonly described as rounder than in normal subjects. This statement derives from observations of qualitative nature or based on the measure of eccentricity index. We analyzed LV shape and function in 16 normal subjects (N) and in 24 patients with chronicpure aortic regurgitation (AR), without coronary artery disease or associated mitral regurgitation. LV cavity geometry was quantitatively evaluated from end-diastolic and end-systolic outlines obtained in 30° RAO angiographic projection, by calculating: 1. the eccentricity and circularity indexes, 2. the regional curvature at 90 equidistant points using a windowed Fourier series approximation of contours, in which the number of harmonics and filter-window were locally chosen in order to minimize the reconstruction errors and to maximize the smoothness of the curve, 3. by measuring the length of the anterior and posterior hemi-perimeter of LV outlines and 4. by performing a Fourier analysis of LV contours.Neither eccentricity nor circularity indexes were adequate to differentiate shape abnormalities, whereas Fourier geometric analysis indicated abnormalities of shape in AR. Regional curvature showed that diastolic outline of AR had a greater curvature of the anterobasal, anterolateral and inferoapical regions and a lower curvature of the anteroapical one. Systolic outline showed a greater curvature of the inferoapical region and a lower curvature of the anteroapical one. The angiographic apex, i.e. the point of the greatest curvature, was shifted towards the mitral plane (at end-diastole from point 48.4 in N to 51.5 in AR; p < 0.001, at end-systole from point 46.3 in N to 49.1 in AR; p=0.007), owing to a greater length of the anterior hemi-perimeter in respect to N. The increase in anterior hemi-perimeter length was significantly related to the decrease in pump function (increase in end-systolic volume index and decrease in ejection fraction).Conclusionin respect to normal subjects LV shape in aortic regurgitation is not simply more globular, but it definitely appears to be asymmetric because of the prevailing elongation of the anterior hemi-perimeter from the aortic corner to the apex suggesting a remodeling of the left ventricle with a prevailing expansion of the anterolateral regions. These alterations in cavity geometry correlate to the decrease in pump function.

Assessing LV wall motion by frame-to-frame curvature matching and optical flow estimation

This work seeks to describe left ventricle (LV) shape by optimizing the curvature computation. A method based on adaptive Fourier series approximation was developed and compared with three known methods. The authors also wish to assess wall motion by applying both the optical flow method and an improved shape matching algorithm, as well as to compare their results with each other and with the centerline method. First, the points of curvature extrema along each contour are tracked frame-to-frame, extrapolated if necessary throughout systole and selected by restrictive criteria. Second, the points lying among these markers are mapped onto the next contours proportionally to local displacement and curvature changes. Results are in closer agreement with physiology knowledge than those of both the centerline and optical flow methods.<<ETX>>

LV wall motion: from quantitative analysis to knowledge-based understanding

An expert system for computer evaluation of cardiac function from left ventricle (LV) contours is under development. A description of normal and pathological ventricles has been assessed by statistical analysis and fuzzy logic. For this purpose LV contours of 16 normal subjects and 67 patients were manually digitized from RAO cineangiograms on a frame-by-frame or two-frame basis. Wall motion was analyzed with eight known quantitative methods. Regional shape and its dynamic changes were quantified by curvature and Fourier analysis. A rule-based scheme combining partial evidence in order to test clinical hypotheses is proposed.<<ETX>>