J.G. Bosch

A computer-aided analysis system for the quantitative assessment of intravascular ultrasound images

In order to provide quantitative assessment of the artery lumen, a computer-aided analysis system was developed. The area measurements are performed by manual tracing of the boundaries of the intima, media and plaque. The mean diameter of the lumen obtained from ultrasound was compared with the mean diameter measured from X-ray angiography. A good correlation was found in nine measurements from three healthy vessels (r=0.97, p<0.0001). In eight patients, the change in the luminal area between systole and diastole was assessed; the luminal area increased by approximately 6% during the systolic period. From these results, it is concluded that the system is suitable for clinical applications such as the measurement of vascular stenosis, the evaluation of therapeutic intervention and the study of arterial wall compliance.<<ETX>>

Azimuthal Registration of Image Sequences Affected by Nonuniform Rotation Distortion

Imaging modalities that use a mechanically rotated endoscopic probe to scan a tubular volume, such as an artery, often suffer from image degradation due to nonuniform rotation distortion (NURD). In this paper, we present a new method to align individual lines in a sequence of images. It is based on dynamic time warping, finding a continuous path through a cost matrix that measures the similarity between regions of two frames being aligned. The path represents the angular mismatch corresponding to the NURD. The prime advantage of this novel approach compared to earlier work is the line-to-line continuity, which accurately captures slow intraframe variations in rotational velocity of the probe. The algorithm is optimized using data from a clinically available intravascular optical coherence tomography (OCT) instrument in a realistic vessel phantom. Its efficacy is demonstrated on an in vivo recording, and compared with conventional global rotation block matching. Intravascular OCT is a particularly challenging modality for motion correction because, in clinical situations, the image is generally undersampled, and correlation between the speckle in different lines or frames is absent. The algorithm can be adapted to ingest data frame-by-frame, and can be implemented to work in real time.

High-Definition Imaging of Carotid Artery Wall Dynamics

The carotid artery (CA) is central to cardiovascular research, because of the clinical relevance of CA plaques as culprits of stroke and the accessibility of the CA for cardiovascular screening. The viscoelastic state of this artery, essential for clinical evaluation, can be assessed by observing arterial deformation in response to the pressure changes throughout the cardiac cycle. Ultrasound imaging has proven to be an excellent tool to monitor these dynamic deformation processes. We describe how a new technique called high-frame-rate ultrasound imaging captures the tissue deformation dynamics throughout the cardiac cycle in unprecedented detail. Local tissue motion exhibits distinct features of sub-micrometer displacements on a sub-millisecond time scale. We present a high-definition motion analysis technique based on plane wave ultrasound imaging able to capture these features. We validated this method by screening a group of healthy volunteers and compared the results with those for two patients known to have atherosclerosis to illustrate the potential utility of this technique.

Design of a micro-beamformer for a 2D piezoelectric ultrasound transducer

This paper describes the design of a transesophageal probe using a matrix (2D array) ultrasonic transducer for 3D echocardiography. To obtain images with sufficient resolution, several thousand elements are required. To reduce the channel count from the transducer to the imaging system, it is necessary to include electronics close to the transducer and use smart signal processing for data reduction. A micro-beamforming method called pre-steering is proposed. All the groups have the same delay configuration, which will simplify the required electronics. For a correct design, delay steps and the maximum delay are of importance. Furthermore, the delay should be programmable for each direction. Simulations are performed to investigate the effect of pre-steering on the received field. The delay and sum operations are realized by an integrated circuit. Simulations show that, for the micro-beamformer of a matrix transducer, the pre-steering approach is an effective method to reduce the complexity of the electronics and the channel count, while maintaining an adequate receive field.

Sparse Registration for Three-Dimensional Stress Echocardiography

Three-dimensional (3-D) stress echocardiography is a novel technique for diagnosing cardiac dysfunction. It involves evaluating wall motion of the left ventricle, by visually analyzing ultrasound images obtained in rest and in different stages of stress. Since the acquisitions are performed minutes apart, variabilities may exist in the visualized cross-sections. To improve anatomical correspondence between rest and stress, aligning the images is essential. We developed a new intensity-based, sparse registration method to retrieve standard anatomical views from 3-D stress images that were equivalent to the manually selected views in the rest images. Using sparse image planes, the influence of common image artifacts could be reduced. We investigated different similarity measures and different levels of sparsity. The registration was tested using data of 20 patients and quantitatively evaluated based on manually defined anatomical landmarks. Alignment was best using sparse registration with two long-axis and two short-axis views; registration errors were reduced significantly, to the range of interobserver variabilities. In 91% of the cases, the registration result was qualitatively assessed as better than or equal to the manual alignment. In conclusion, sparse registration improves the alignment of rest and stress images, with a performance similar to manual alignment. This is an important step towards objective quantification in 3-D stress echocardiography.

Doppler flow velocity waveforms in the embryonic chicken heart at developmental stages corresponding to 5-8 weeks of human gestation

To obtain Doppler velocity waveforms from the early embryonic chicken heart by means of ultrasound biomicroscopy and to compare these waveforms at different stages of embryonic development.

Developments towards real-time frame-to-frame automatic contour detection on echocardiograms

An automated contour detection method has been developed for the assessment of left ventricular wall motion from short-axis transesophageal echocardiograms. The method is based on minimum-cost contour detection using an iterative approach and implicit knowledge about left ventricular geometry. The contour detection takes about 4 seconds per image on a 20-MHz Compaq 386. This automated method has been evaluated by comparison of automatically detected contours with contours drawn manually on 480 images. Inter- and intra-observer variabilities were determined for both types of contours. Performance of the manual drawing was only slightly better than automatic detection. This technique is now being extended towards frame-to-frame contour detection. Very promising results were obtained in a pilot study on a high-quality echo sequence.<<ETX>>

Automated endocardial contour detection in short-axis 2-D echocardiograms; methodology and assessment of variability

To obtain objective and reproducible data on left-ventricular (LV) wall motion from esophageal short-axis echocardiograms, an automated endocardial contour-detection technique has been developed. The method is based on minimum-cost contour detection using a novel iterative approach, which allows complex nonconvex contours to be found with a simple model. Required operator interaction is limited to the manual definition of the center of the LV cavity. To determine the inter- and intraoperator variabilities of this automated approach, an intermediate evaluation has been carried out on the end-systolic and end-diastolic images of 10 patient studies. The method was found to be minimally sensitive to small variations in the manually indicated LV center points. However, the algorithm requires further improvements for specific problematic image regions, such as at the sites of the papillary muscles.<<ETX>>

Semiautomatic frame-to-frame tracking of the luminal border from intravascular ultrasound

The authors describe a template-matching method which allows the measurement of the frame-to-frame changes in the luminal cross-sectional area from intravascular ultrasound images through the analysis of the regional wall displacement. The matching is performed by calculating a cross-correlation coefficient between the template and the test data. The optimal matching is determined using the minimum-cost algorithm. Comparison of the results from the template-matching method and those from manual tracing showed that the mean difference in the measure of the area change was very small. The beat-to-beat variations evaluated with data from the method were lower than with manual tracing. It is concluded that this method can be applied to wall compliance studies using intravascular ultrasound.<<ETX>>

P2A-6 Automatic Segmentation of the Left Ventricle in 3D Echocardiography Using Active Appearance Models

Assessment of left ventricular (LV) functional parameters, such as LV volume, ejection fraction and stroke volume, from real-time 3D echocardiography (RT3DE) is labor intensive and subjective, because in current analyses it requires input from the user. Automating these procedures will save valuable time in the analysis and will remove interobserver variability. We propose a fully automatic segmentation approach for the left ventricle in real-time 3D echocardiography, based on active appearance models (AAMs), using ultrasound specific grey value normalization. We evaluated shape and texture model generalization. Also, automatic segmentation has been preliminarily evaluated on transthoracic, apical acquisitions of 54 patients, acquired with the fast rotating ultrasound (FRU-) transducer (18 patients) and with the Philips Sonos 7500 (36 patients). The evaluations were done in a leave-N-out manner (with N=5). We evaluated point-to-surface (P2S) distances for the segmented endocardial contours to the expert manual contours. The generalization of the shape model was good, but texture model generalization was moderate, hampering the AAM matching We found preliminary segmentation errors (P2S) of 3.9plusmn 1.6 mm (N=54) for detection using AAM matching These results indicate that fully automatic segmentation of the LV in RT3DE using AAMs is feasible.