Intravascular Ultrasound Imaging Catheter Research Articles

Convolutional networks for the segmentation of intravascular ultrasound images: Evaluation on a multicenter dataset

Background and objectiveThe delineation of the lumen contour and external elastic lamina (EEL) in intravascular ultrasound (IVUS) images is crucial for the quantitative analysis of coronary atherosclerotic plaques. However, the presence of ultrasonic shadows and anatomical structures (such as bifurcations and calcified plaque) complicates the automatic delineation of the lumen contour and EEL. The purpose of this paper is to evaluate the IVUS segmentation performances of different convolutional networks and the impact factors on a large-scale multiple-center dataset. MethodsA total of 6516 cross-sectional images from 175 IVUS pullbacks acquired in different centers by different IVUS imaging catheters were screened from a corelab to evaluate the segmentation methods. The IVUS images included bifurcation, side branch ostia, and various image artifacts to reflect the general image characteristics in routine clinical acquisition. We compared three generic fully convolutional networks (FCNs) and two FCNs specifically designed for the segmentation of IVUS images and explored the factors impacting the segmentation performance, including the training images and the input of consecutive images to the models. The performance of the FCNs was evaluated by using the Dice similarity coefficient (DSC), the Jaccard index (JI), the Hausdorff distance (HD), linear regression and Bland-Altman analysis. ResultsThe 4-cascaded RefineNet and DeepLabv3+ outperformed U-net and IVUS-net in the segmentation of the lumen contour and EEL on IVUS images. DeepLabv3+ had the best segmentation performance, with DSCs of 0.927 and 0.944, JIs of 0.911 and 0.933, and HDs of 0.336 mm and 0.367 mm for delineation of the lumen and EEL, respectively. Excellent agreement between DeepLabv3+ and the manual delineation was found in the quantification of the coronary plaque area (r = 0.98). ConclusionsThe convolutional network architecture is effective in the automatic segmentation of IVUS images. It might contribute to the clinical application of quantitative IVUS analysis in real-world as well as the efficient assessment of coronary atherosclerosis.

Intravascular ultrasonic–photoacoustic (IVUP) endoscope with 2.2-mm diameter catheter for medical imaging

Intravascular ultrasound (IVUS) imaging is extremely important for detection and characterization of high-risk atherosclerotic plaques as well as gastrointestinal diseases. Recently, intravascular photoacoustic (IVPA) imaging has been used to differentiate the composition of biological tissues with high optical contrast and ultrasonic resolution. The combination of these imaging techniques could provide morphological information and molecular screening to characterize abnormal tissues, which would help physicians to ensure vital therapeutic value and prognostic significance for patients before commencing therapy. In this study, integration of a high-frequency IVUS imaging catheter (45MHz, single-element, unfocused, 0.7mm in diameter) with a multi-mode optical fiber (0.6mm in core diameter, 0.22 NA), an integrated intravascular ultrasonic–photoacoustic (IVUP) imaging catheter, was developed to provide spatial and functional information on light distribution in a turbid sample. Simultaneously, IVUS imaging was co-registered to IVPA imaging to construct 3D volumetric sample images. In a phantom study, a polyvinyl alcohol (PVA) tissue-mimicking arterial vessel phantom with indocyanine green (ICG) and methylene blue (MB) inclusion was used to demonstrate the feasibility of mapping the biological dyes, which are used in cardiovascular and cancer diagnostics. For the ex vivo study, an excised sample of pig intestine with ICG was utilized to target the biomarkers present in the gastrointestinal tumors or the atherosclerotic plaques with the proposed hybrid technique. The results indicated that IVUP endoscope with the 2.2-mm diameter catheter could be a useful tool for medical imaging.

Development of a catheter for combined intravascular ultrasound and photoacoustic imaging

Atherosclerosis is characterized by formation and development of the plaques in the inner layer of the vessel wall. To detect and characterize atherosclerotic plaques, we previously introduced the combined intravascular ultrasound (IVUS) and intravascular photoacoustic (IVPA) imaging capable of assessing plaque morphology and composition. The utility of IVUS/IVPA imaging has been demonstrated by imaging tissue-mimicking phantoms and ex vivo arterial samples using laboratory prototype of the imaging system. However, the clinical realization of a IVUS/IVPA imaging requires an integrated intravascular imaging catheter. In this paper, two designs of IVUS/IVPA imaging catheters--side fire fiber-based and mirror-based catheters--are reported. A commercially available IVUS imaging catheter was utilized for both pulse-echo ultrasound imaging and detection of photoacoustic transients. Laser pulses were delivered by custom-designed fiber-based optical systems. The optical fiber and IVUS imaging catheter were combined into a single device. Both designs were tested and compared using point targets and tissue-mimicking phantoms. The results indicate applicability of the proposed catheters for clinical use.

Intravascular ultrasound and photoacoustic imaging.

Catheter-based intravascular ultrasound (IVUS) is one of the imaging tools for the clinical evaluation of atherosclerosis. However, histopathological information obtained with IVUS imaging is limited. We developed a combined intravascular photoacoustic (IVPA) and IVUS imaging to characterize atherosclerotic plaques. The amplitude, spectral, and temporal characteristics of the photoacoustic signal are primarily determined by optical absorption properties of different types of tissues and can be used to differentiate the lipid, blood, fibrous, and fibro-cellular components of an inflammatory lesion. Furthermore, using bioconjugated contrast agents such as metal nanoconstructs, molecular IVPA imaging is possible since different molecules are overexpressed during various stages of atherosclerosis. Imaging experiments were performed using clinical IVUS imaging catheters interfaced with a pulsed laser system. The performance of the IVPA/IVUS imaging was assessed using vessel-mimicking phantoms. To detect the lipids in the plaque and to assess plaque composition, multi-wavelength (680–950 nm) spectroscopic IVPA imaging of a normal and an atherosclerotic rabbit aorta was performed. Molecular and cellular IVPA imaging was demonstrated using intravenously injected plasmonic nanoparticles. Finally, IVPA imaging was used to visualize coronary stents within the vessel wall. Our studies suggest that plaque detection, characterization, and even treatment can be improved using the combined IVPA.

In-vivo, cardiac-cycle related intimal displacement of coronary plaques assessed by 3-D ECG-gated intravascular ultrasound: exploring its correlate with tissue deformability identified by palpography

ECG-gated image acquisition of intravascular ultrasound (IVUS) has been shown to provide more accurate measurements at different phases of the cardiac cycle. We sought to explore the ability dynamic assessment of ECG-gated 3-D IVUS to identify deformable regions of coronary plaques, by testing the hypothesis that at a given pressure and region, a faster displacement of the intima would correspond to high strain (soft tissue) regions assessed by palpography. ECG-gated 3-D IVUS and palpograms were acquired using 30 and 20 MHz IVUS imaging catheters respectively. Frames with high and/or low strain spots identified by palpography were randomly selected and the spots were assigned to a respective quadrant within the cross section. A color-blinded side-by-side view was performed to enable the co-localization of the same region. Subsequently, the pressure driven displacement of the intima was established for each quadrant and a binary score (significant displacement or no displacement) was decided. One hundred and twenty-four quadrants were studied and the prevalence of highly deformable quadrants was low (n=7, 5.6% of the total). The sensitivity, specificity, positive predictive value and negative predictive value of 3-D ECG-gated IVUS to detect deformable quadrants as assessed by palpography were 42.9, 87.2, 16.7, and 96.2% respectively. In this pilot in vivo study, the intimal displacement velocity in the radial direction assessed by gray-scale 3-D ECG-gated IVUS failed to correlate with highly deformable regions. However, these preliminary findings suggest that the absence of significant displacement of the intima might be accurate to predict the absence of deformable tissue.

Efficient three-dimensional cylindrical-geometry ultrasound imaging

A number of new imaging modalities collect data from cylindrical platforms. In vivo imaging needles and intravascular ultrasound imaging catheters are examples of this geometry, where imager rotation and translation parallel to the cylinder axis are the only allowed motions. Efficient three-dimensional ultrasound image formation in these cases can be challenging when the aperture is small and/or highly curved. A frequency-domain imaging algorithm is obtained by approximating the free-space point spread function in cylindrical coordinates and obtaining its Fourier transform by analogy with the equivalent problem in Cartesian coordinates. We further propose an effective use of limited aperture by placing a focused transducer across the aperture, thereby creating a virtual source at the focus which is treated as a real, unfocused source by the imaging algorithm. This approach retains the simplicity and potential angular resolution of a small single element, yet permits full use of the available probe aperture and a higher energy output. Computer simulations and experimental ultrasonic results with wire targets show that this imaging technique attains the theoretical resolution dictated by the operating wavelength and transducer characteristics. [Work supported by NIH CA 079179.]

In VivoThrombogenicity of Intravascular Ultrasound Imaging Catheters

<i>In Vivo</i>Thrombogenicity of Intravascular Ultrasound Imaging Catheters

Geometric accuracy of intravascular ultrasound imaging.

In spite of the current interest in and clinical application of intravascular ultrasound imaging, there is very little published information on the limitations of this modality. In vitro studies on nine phantom vessels (diameter 4.76 to 12.7 mm) and 11 human arteries (diameter 3.7 to 6.2 mm) were performed to assess the potential sources of error of diagnostic intravascular ultrasound imaging. The effects of (1) blood flow velocity, (2) temperature, (3) eccentric, noncentral catheter placement, (4) alteration of the angle of incidence by 30 degrees, and (5) the effect of imaging in different mediums--saline solution, blood, and electrode gel--were studied. Variations in blood flow velocity (from 10 to 300 ml/min) and temperature (from 22 degrees C to 37 degrees C) resulted in a < 2% change in the lumen area measured by intravascular ultrasound imaging catheters. Eccentric catheter location had little effect on phantom or human arterial lumen shape or area when imaging was performed with optimized catheters. However, with used catheters circular lumens appeared elliptical with an eccentric index for phantoms from 0.88 to 1.15, (P < 0.05), and for human arteries from 0.88 to 1.11 (P < 0.05). The area ranged from 89% to 112% (P < 0.05) in phantoms and from 90% to 110% in human arteries compared with the lumen areas measured with a central catheter position (control). A 30-degree alteration in the angle of incidence resulted in 16.3% +/- 5.5% increase in lumen area for phantoms and 14.2% +/- 8.6% for human arteries in vitro. Ultrasonic-measured wall thickness of human vessels correlated closely with the actual measured thickness (r = 0.93) when a central catheter position was used. The wall thickness measured during adjacent (< 0.2 mm) and far-wall positioning (1.9 mm) of the catheter correlated closely (r = 0.96), but the far wall thickness with a 30-degree angle of incidence resulted in a 10.6% increase from control. Studies in saline solution resulted in significantly different measures of lumen area compared with imaging in blood. Compared with images recorded in blood, images in saline solution were 7.6% to 8.2% larger and 3.9% to 7.2% smaller in gel.(ABSTRACT TRUNCATED AT 400 WORDS)

Morphological effects of coronary balloon angioplasty in vivo assessed by intravascular ultrasound imaging.

Histological examination of the effects of balloon angioplasty have been described from in vitro experiments and a limited number of pathologic specimens. Intravascular ultrasound imaging permits real time cross-sectional observation of the effect of balloon dilation on the atherosclerotic plaque in vivo. The morphological effects of coronary angioplasty were visualized at 66 lesions in 47 patients immediately after balloon dilatation with an intravascular ultrasound imaging catheter. Cross-sectional images were obtained at 30 frames per second as the catheter passed along the length of the artery. Quantitative and qualitative assessments of the dilated atherosclerotic plaque were made from the angiograms and the ultrasound images. Six morphological patterns after angioplasty were appreciated by ultrasound imaging. Type A consists of a linear, partial tear of the plaque from the lumen toward the media (seven lesions); Type B is defined by a split in the plaque that extends to the media (12 lesions); Type C demonstrates a dissection behind the plaque that subtends an arc of up to 180 degrees around the circumference (18 lesions); Type D was a more extensive dissection that encompasses an arc of more than 180 degrees (four lesions); and Type E may be present in either concentric (Type E1, 14 lesions) or eccentric (Type E2, 11 lesions) plaque and is defined as an ultrasound study without any evidence of a fracture or a dissection in the plaque. There was a large amount of residual atheroma in each type of morphology (7.8 +/- 2.9 mm2, 61.6 +/- 15.4% of cross-sectional area); there was no difference, however, in lumen or atheroma cross-sectional area among these six patterns. There was a good correlation between ultrasound and angiography for the recognition of a dissection. Calcification was seen in only 14% of lesions on angiography, whereas most lesions (83%) revealed calcification on ultrasound imaging. As determined by intravascular ultrasound, calcified plaque was more likely to fracture in response to balloon dilatation than noncalcified plaque (p less than 0.01). Thirteen of 66 lesions (20%) developed clinical and angiographic restenosis. Restenosis was more likely to occur when the original dilatation left a concentric plaque without a fracture or dissection (Type E1, 50% incidence) compared with a mean restenosis rate of 12% in the remaining morphological patterns (p = 0.053). Intravascular ultrasound provides a more complete quantitative and qualitative description of plaque geometry and composition than angiography after balloon angioplasty. In addition, intravascular ultrasound identified a subset of atherosclerotic plaque that has a higher incidence of restenosis. This information could be used prospectively to consider other therapeutic options in this subset. Intravascular ultrasound provides a method to describe the effects of angioplasty that will be useful in comparing future coronary intervention studies.

Ultrasound guidance of laser atherectomy

Intravascular ultrasound imaging catheters and forward firing laser atherectomy systems are both being used in the coronary vascular tree of man. When used in conjunction, ultrasound imaging may reduce the probability of perforation and dissection related to laser atherectomy. A novel system is presented which combines ultrasound imaging and laser atherectomy capabilities on the same catheter. The system incorporates a holmium YAG laser directed perpendicular to the long axis of the catheter along with a co-directional ultrasound sensor. Ultrasound images are assembled by computer using the angle of the catheter tip and its linear advance associated with each ultrasound echo signal. In vivo experimental results are presented.

Histopathologic validation of angioscopy and intravascular ultrasound.

To establish a histopathologic basis for angioscopic and ultrasound image interpretation we studied 70 postmortem human arterial segments in vitro. We used 7- to 9-French fiber-optic angioscopes and 20- to 30-MHz intravascular ultrasound imaging catheters. Three observers assigned an angioscopic and ultrasound image classification to each vessel segment. The image and histological classification categories were then compared. The sensitivity, specificity, and accuracy of both methods separately or in combination for normal vessels were each greater than or equal to 95%. The predictive value was better for angioscopy than for ultrasound due to incorrect ultrasound interpretations of normal anatomy in the presence of thrombus. For stable atheroma the sensitivity, specificity, and accuracy of the individual methods were each greater than 90%. However, both angioscopy and ultrasound had classification errors in that disrupted atheroma was identified and classified as stable atheroma. Consequently, the predictive value was 74% for angioscopy and 78% for ultrasound. For disrupted atheroma the sensitivities for angioscopy and ultrasound were only moderate (73% and 81%, respectively), whereas the specificity, accuracy, and predictive value were each high (greater than 90%). For thrombus detection, the specificity, accuracy, and predictive value were high (greater than 93%) for each method. The sensitivity of angioscopy was 100%. However, sensitivity was lower for ultrasound (57%) due to false-negative interpretation of laminar clots in normal vessels and an inability to distinguish disrupted or stable atheroma from intraluminal thrombus. Contingency analyses showed that each imaging method alone or combined had significant agreement with the results obtained from histology (p less than 0.001). When assessing all cases in which angioscopy and ultrasound were concordant, there was a 92% agreement with the histological classification.

The Reproducibility of Intravascular Ultrasound Imaging In Vitro

To determine which factors may affect the image quality when an intravascular ultrasound catheter is used in vivo, the influence of blood, temperature change, and contrast media were evaluated. In addition, to confirm the reproducibility of intravascular ultrasound imaging to measure cross-sectional lumen area, intraobserver and interobserver variability were determined. The findings indicated that ultrasound images in blood are mildly attenuated, that changes from room temperature to body temperature do not have a significant impact on the image quality, that contrast media attenuates the image intensity in a dose-dependent manner, and that the intravascular ultrasound imaging catheter provides a reproducible method for measuring arterial lumen area with excellent intraobserver and interobserver correlation.

Assessment of normal and atherosclerotic arterial wall thickness with an intravascular ultrasound imaging catheter

A prototype intravascular ultrasound imaging catheter with a 20 MHz transducer was used to obtain 59 cross-sectional images in 14 segments of human atherosclerotic arteries. Three distinct components of the arterial wall were visualized on the ultrasound images: a highly reflective intima, an echolucent media, and a moderately reflective adventitia. Images were obtained at 1 mm increments in vitro and were compared with histologic sections at the same levels. Measurements of the arterial layers showed a close correlation between ultrasound images and histologic sections for the thickness of the intimal plaque (r = 0.91), the media (r = 0.83), and the total wall thickness (r = 0.85). The ultrasound images overestimated the mean intimal and total wall thickness by 0.3 mm and 0.7 mm compared to measurements in histologic sections (p less than 0.001). Intravascular imaging with high-frequency ultrasound is an accurate method for measuring microanatomic arterial dimensions and the extent of atheromatous involvement of the arterial wall. This method could represent an important adjunct to traditional angiographic techniques for assessing the severity of atherosclerosis.

Functional significance of arterial stenoses predicted by angiography and intravascular ultrasound imaging catheter

Functional significance of arterial stenoses predicted by angiography and intravascular ultrasound imaging catheter

Clinical percutaneous imaging of coronary anatomy using an over-the-wire ultrasound catheter system

This manuscript describes initial applications of a unique new intravascular ultrasound imaging catheter. This 5.5F catheter uses an over-the-wire design and incorporates a phased array transducer at its tip. There are no moving parts. A 360 degree image is produced perpendicular to the catheter axis using a 20 MHz center frequency. A dedicated minicomputer is used for initial image processing, as well as enhancement and analysis. Initial studies using phantoms demonstrated excellent accuracy for linear dimensions (r = 0.99, range 3.0 to 7.6 mm, image = 1.0 phantom + 0.1). Serial imaging of the same arterial segment in vitro showed good reproducibility (coefficients of variance 2.5-5.2%). Likewise, intra- and inter-observer variability in image analysis was minimal (r = 0.92-0.99). Initial in vivo studies were performed in dogs. The catheter was easily passed over a wire into mesenteric, cerebral and coronary vessels without evidence of significant vessel trauma. Subsequently, 20 patients had percutaneous coronary imaging performed during cardiac catheterization. Cardiac motion was rarely a problem and acceptable images were obtained in all but two patients. Areas of calcification, mild stenoses, branching vessels and graft atherosclerosis could be identified. We conclude that intracoronary ultrasound imaging will be useful for assessing vascular pathology, for studying both rapid change in vessel size as well as chronic progression or regression of atherosclerosis, and for assisting with new therapeutic interventions.