Jérôme Garot

Segmentation of 4D cardiac MRI: Automated method based on spatio-temporal watershed cuts

Based on discrete mathematical morphology, we introduce in this paper a new watershed framework which allows for segmenting spatio-temporal images, that we apply to medical image analysis. Specifically, we propose a new automated and fast procedure to segment the left ventricular myocardium in 4D (3D+t) cine-MRI sequences. Both quantitative and qualitative evaluation are provided. Thanks to the comparison with manual segmentation performed by two cardiologists, we demonstrate the accuracy of the proposed method. The relevance of the ejection fraction and myocardium mass measured from segmentations is also assessed. Furthermore, we show that the proposed 4D procedure allows to keep the temporal coherency between the successive 3D segmentations obtained along the time sequence. In an effort to promote open science, both the data and the software are available on-line.

Effect of delayed percutaneous transluminal coronary angioplasty of occluded coronary arteries after acute myocardial infarction.

Whether angioplasty of occluded vessels after myocardial infarction may have beneficial effects on left ventricular function remains unknown. Patients with a first myocardial infarction and thrombolytic therapy who had an occluded infarct-related vessel at delayed coronary angiography were referred systematically for an elective coronary angioplasty performed between 3 and 4 weeks after the myocardial infarction. All patients underwent stress-redistribution-reinjection thallium-201 single-photon emission computed tomography for myocardial viability assessment. Prior angioplasty, a quantitative evaluation of global and regional left ventricular function, was performed. The study group consisted of 38 patients (aged 57 +/- 10 years); 18 had anterior wall infarctions and 20 inferior wall infarctions, but before angioplasty 3 had a patent artery and were excluded. Angioplasty was successful in 30 patients. At follow-up 13 patients (43%) had an occluded coronary artery. In contrast with patients with an occluded coronary artery at follow-up, those with a patent coronary artery had no left ventricular enlargement and had an improvement in both left ventricular ejection fraction (from 48 +/- 9% to 52 +/- 9.8%, p = 0.002) and regional wall motion index (delta = +0.95 SD, p <0.01). In patients with a patent vessel at follow-up, there was a positive correlation between the number of myocardial viable segments and improvement of the infarct zone wall motion (r = 0.52; p = 0.035), and the number of necrotic segments at baseline was positively correlated to the 4-month changes in end-diastolic volume indexes (r = 0.6; p = 0.04). Thus, elective revascularization of occluded coronary arteries with viable myocardium after myocardial infarction improves left ventricular function and lessens remodeling if the artery remains patent during follow-up.

Time course and relation to local viability of microvascular function and volume after reperfused acute myocardial infarction

We assessed the time course of the alterations of microvascular function and myocardial perfusion, as well their relation to local inotropic reserve (IR), in 21 patients who underwent successful primary coronary angioplasty for acute myocardial infarction and in whom local myocardial dysfunction persisted at hospital discharge. Coronary flow reserve (CFR) and myocardial perfusion were assessed immediately after angioplasty, and on day 1 and day 8 by intracoronary Doppler and myocardial contrast echocardiography, respectively. Dobutamine echocardiography was performed on day 7 for assessment of local IR. After angioplasty, CFR was severely altered in patients with (n = 14) and without (n = 7) IR (1.44 +/- 0.26 and 1.36 +/- 0.21, respectively; p = NS). Among patients with IR, CFR increased significantly at day 1 (2.26 +/- 0.62, p <0.005 vs acute stage) compared with those without IR (p = NS vs acute). In contrast, the extent of microvascular obstruction as defined by contrast echocardiography remained unchanged in all patients at day 1 compared with acute measurements. Microvascular obstruction decreased at day 8 in the sole subset of patients with local IR (p <0.05 vs acute stage). In patients treated by immediate coronary angioplasty for acute myocardial infarction, subsequent improvement of myocardial perfusion is associated with preexistent recruitable microvascular function in the infarct-related artery. The presence of reversible microvascular dysfunction at the early stage after acute myocardial infarction is associated with local tissue viability in humans.

Assessment of myocardial deformations from untagged cardiac cine MRI

This paper presents a method for the assessment of left ventricular myocardial deformations from untagged cardiac cine MRI. Firstly, a level set segmentation process is applied on the dataset to detect both endocardial and epicardial boundaries. Then the successive contours are matched using an original procedure that consists in an alignment followed by a morphing process. From the matched contours, we deduced an initial velocity of the contour points that will be used in a Thin-Plate Splines approximation method for estimating of the velocity flow on the complete myocardial structure. Finally, local measurements of ventricular deformations are derived from the velocity flow. The validation of the method is performed both mathematically and by comparing the measurements to those obtained on the same patients with the HARmonic Phase reference (HARP) method applied on matched tagged MR images.

MR/CT Multimodal Registration of Short-Axis Slices in CT Volumes

This paper deals with the registration of magnetic resonance (MR) and computerized tomography (CT) cardiac images. We use a multimodal iconic algorithm based on the maximisation of the mutual information of the joint histogram of two images. We apply it to the registration of MR and CT cardiac images. The purpose is to determine in a (3D+t) CT data volume, short-axis slices that correspond to the (2D+t) short-axis MR slices, acquired on the same patient. The couples of images should be similar according to their spatial position, orientation and cardiac cycle phase. We used the Powells direction set optimization method for the maximisation of the mutual information, which gives good results with an appropriate initialization.