A Numerical Preoperative Planning Model to Predict Arterial Deformations in Endovascular Aortic Aneurysm Repair

Endovascular aneurysm repair is rapidly emerging as the primary preferred method for treating abdominal aortic aneurysm. In this image-guided interventional procedure, to obtain the roadmap and decrease contrast injections, preoperative CT images are overlaid onto live fluoroscopy images using various 2D/3D image fusion techniques. However, the structural changes due to the insertion of stiff tools degrade the fusion accuracy. To correct the mismatch and quantify the intraoperative deformations, we present a patient-specific biomechanical model of the aorto-iliac structure and its surrounding tissues. The predictive capability of the model was evaluated against intraoperative data for a group of four patients. Incorporating the perivascular tissues into the model significantly improved the results and the mean distance between the real and simulated endovascular tools was 2.99?±?1.78 mm on the ipsilateral side and 4.59?±?3.25 mm on the contralateral side. Moreover, the distance between the deformed iliac ostia and their corresponding landmarks on intraoperative images was 2.99?±?2.48 mm.     

Aortic valve closure for rapidly deteriorated aortic insufficiency after continuous flow left ventricular assist device implantation

A patient underwent aortic valve closure for de novo aortic insufficiency that had deteriorated to severe insufficiency during six months of support with a continuous flow left ventricular assist device (cf-LVAD). Aortic insufficiency was initially noted one month after LVAD implantation, and then deterioration quickly developed. Right heart catheterization revealed that when the rotational speed of the cf-LVAD was increased, the cardiac index was decreased by an increase in regurgitant volume, as shown by echocardiography. During surgery, fusion and shortening of the aortic leaflets as well as left coronary ostial occlusion were observed. Direct aortic closure improved hemodynamics. Thrombus formation on the aortic valve shown by echocardiography in the early postoperative period may be a trigger of aortic insufficiency. Control of the cf-LVAD rotational speed is likely required to prevent aortic insufficiency.     

2D/3D registration of endoscopic ultrasound to CT volume data

This paper describes a computer-aided navigation system using image fusion to support endoscopic interventions such as the accurate collection of biopsy specimens. An endoscope provides the physician with real-time ultrasound (US) and a video image. An image slice that corresponds to the corresponding image from the US scan head is derived from a preoperative computed tomography (CT) or magnetic resonance image volume data set using oblique reformatting and displayed side by side with the US image. The position of the image acquired by the US scan head is determined by a miniaturized electromagnetic tracking system (EMTS) after calibrating the endoscope's scan head. The transformation between the patient coordinate system and the preoperative data set is calculated using a 2D/3D registration. This is achieved by calibrating an intraoperative interventional CT slice with an optical tracking system (OTS) using the same algorithm as for the US calibration. The slice is then used for 2D/3D registration with the coordinate system of the preoperative volume. The fiducial registration error (FRE) for the US calibration was 2.0 mm +/- 0.4 mm; the interventional CT FRE was 0.36 +/- 0.12 mm; and the 2D/3D registration target registration error (TRE) was 1.8 +/- 0.3 mm. The point-to-point registration between the OTS and the EMTS had an FRE of 0.9 +/- 0.4 mm. Finally, we found an overall TRE for the complete system to be 3.9 +/- 0.6 mm.     

Evaluation of advancing the standard valve dysfunction by multidetector-row CT

We evaluated the diagnostic usefulness of electrocardiographically gated multidetector-row computed tomography (MDCT) for prosthetic valve dysfunction (PVD) of an ATS valve. Twenty-four patients underwent MDCT following echocardiography and cineradiography. Echocardiography and cineradiography showed normal valve function in 17 patients and PVD in 7. PVD included aortic prosthetic valve obstruction in 4 patients, an aortic annular aneurysm with paraprosthetic regurgitation in one, and a blocked leaflet in the mitral position in 2. Among the 7 patients, 5 received reoperation after MDCT. MDCT revealed a subprosthetic mass in all 5 patients with PVD and in 4 patients with normal valve function in the aortic position. In addition to a subprosthetic mass, an annular aneurysm was found in one. Valvular masses were detected in 2 patients with mitral PVD. At reoperation, subprosthetic pannus in the aortic position was detected in 2 patients, subprosthetic pannus and annular aneurysm with paraprosthetic leaks in one, and mitral valve thrombosis in 2. These findings confirmed at reoperation matched to the findings observed on MDCT. The mean CT attenuation of the subprosthetic mass in 6 patients was 152 ± 12 HU and that of the subprosthetic pannus in 3 patients was 163 ± 17 HU. CT attenuation of the thrombus in the mitral valve in the 2 patients was 60 and 99 HU. Our study demonstrates that MDCT is a valuable and reliable diagnostic technique for PVD in an ATS valve and that MDCT may identify an abnormality causing PVD.     

Performance of a feature-based algorithm for 3D-3D registration of CT angiography to cone-beam CT for endovascular repair of complex abdominal aortic aneurysms

Background

A crucial step in image fusion for intraoperative guidance during endovascular procedures is the registration of preoperative computed tomography angiography (CTA) with intraoperative Cone Beam CT (CBCT). Automatic tools for image registration facilitate the 3D image guidance workflow. However their performance is not always satisfactory. The aim of this study is to assess the accuracy of a new fully automatic, feature-based algorithm for 3D3D registration of CTA to CBCT.

Methods

The feature-based algorithm was tested on clinical image datasets from 14 patients undergoing complex endovascular aortic repair. Deviations in Euclidian distances between vascular as well as bony landmarks were measured and compared to an intensity-based, normalized mutual information algorithm.

Results

The results for the feature-based algorithm showed that the median 3D registration error between the anatomical landmarks of CBCT and CT images was less than 3 mm. The feature-based algorithm showed significantly better accuracy compared to the intensity-based algorithm (p?<?0.001).

Conclusion

A feature-based algorithm for 3D image registration is presented.

     

3D Printed Modeling of the Mitral Valve for Catheter-Based Structural Interventions

As catheter-based structural heart interventions become increasingly complex, the ability to effectively model patient-specific valve geometry as well as the potential interaction of an implanted device within that geometry will become increasingly important. Our aim with this investigation was to combine the technologies of high-spatial resolution cardiac imaging, image processing software, and fused multi-material 3D printing, to demonstrate that patient-specific models of the mitral valve apparatus could be created to facilitate functional evaluation of novel trans-catheter mitral valve repair strategies. Clinical 3D transesophageal echocardiography and computed tomography images were acquired for three patients being evaluated for a catheter-based mitral valve repair. Target anatomies were identified, segmented and reconstructed into 3D patient-specific digital models. For each patient, the mitral valve apparatus was digitally reconstructed from a single or fused imaging data set. Using multi-material 3D printing methods, patient-specific anatomic replicas of the mitral valve were created. 3D print materials were selected based on the mechanical testing of elastomeric TangoPlus materials (Stratasys, Eden Prairie, Minnesota, USA) and were compared to freshly harvested porcine leaflet tissue. The effective bending modulus of healthy porcine MV tissue was significantly less than the bending modulus of TangoPlus (p < 0.01). All TangoPlus varieties were less stiff than the maximum tensile elastic modulus of mitral valve tissue (3697.2 ± 385.8 kPa anterior leaflet; 2582.1 ± 374.2 kPa posterior leaflet) (p < 0.01). However, the slopes of the stress-strain toe regions of the mitral valve tissues (532.8 ± 281.9 kPa anterior leaflet; 389.0 ± 156.9 kPa posterior leaflet) were not different than those of the Shore 27, Shore 35, and Shore 27 with Shore 35 blend TangoPlus material (p > 0.95). We have demonstrated that patient-specific mitral valve models can be reconstructed from multi-modality imaging datasets and fabricated using the multi-material 3D printing technology and we provide two examples to show how catheter-based repair devices could be evaluated within specific patient 3D printed valve geometry. However, we recognize that the use of 3D printed models for the development of new therapies, or for specific procedural training has yet to be defined.     

Application of dynamic computed tomography for measurements of local aortic elastic modulus

A novel computed tomographic (CT) technique used for the instantaneous measurement of the dynamic elastic modulus of intact excised porcine aortic vessels subjected to physiological pressure waveforms is described. This system was comprised of a high resolution X-ray image intensifier based computed tomographic system with limiting spatial resolution of 3.2mm⁻¹ (for a 40mm field of view) and a computer-controlled flow simulator. Utilising cardiac gating and computer control, a time-resolved sequence of 1 mm thick axial tomographic slices was obtained for porcine aortic specimens during one simulated cardiac cycle. With an image acquisition sampling interval of 16.5 ms, the time sequences of CT slices were able to quantify the expansion and contraction of the aortic wall during each phase of the cardiac cycle. Through superficial tagging of the adventitial surface of the specimens with wire markers, measurement of wall strain in specific circumferential sectors and subsequent calculations of localised dynamic elastic modulus were possible. The precision of circumferential measurements made from the CT images utilising a cluster-growing segmentation technique was approximately ±0.25mm and allowed determination of the dynamic elastic modulus (E_dyn) with a precision of ±8kPa. Dynamic elastic modulus was resolved as a function of the harmonics of the physiological pressure waveform and as a function of the angular position around the vessel circumference. Application of this dynamic CT (DCT) technique to seven porcine thoracic aortic specimens produced a circumferential average (over all frequency components) E_dyn of 373±29kPa. This value was not statistically different (p<0.05) from the values of 430±77 and 390±47kPa obtained by uniaxial tensile testing and volumetric measurements respectively.     

Validation of a two- to three-dimensional registration algorithm for aligning preoperative CT images and intraoperative fluoroscopy images.

We present a validation of an intensity based two- to three-dimensional image registration algorithm. The algorithm can register a CT volume to a single-plane fluoroscopy image. Four routinely acquired clinical data sets from patients who underwent endovascular treatment for an abdominal aortic aneurysm were used. Each data set was comprised of two intraoperative fluoroscopy images and a preoperative CT image. Regions of interest (ROI) were drawn around each vertebra in the CT and fluoroscopy images. Each CT image ROI was individually registered to the corresponding ROI in the fluoroscopy images. A cross validation approach was used to obtain a measure of registration consistency. Spinal movement between the preoperative and intraoperative scene was accounted for by using two fluoroscopy images. The consistency and robustness of the algorithm when using two similarity measures, pattern intensity and gradient difference, was investigated. Both similarity measures produced similar results. The consistency values were rotational errors below 0.74 degree and in-plane translational errors below 0.90 mm. These errors approximately relate to a two-dimensional projection error of 1.3 mm. The failure rate was less than 8.3% for three of the four data sets. However, for one of the data sets a much larger failure rate (28.5%) occurred.     

Outflow graft anastomosis site design could be correlated to aortic valve regurgitation under left ventricular assist device support

Aortic valve regurgitation (AR) is a critical complication during circulatory support with a left ventricular assist device (LVAD). The time-course of AR and related factors, including outflow graft anastomosis site design, were investigated. Twenty-three patients who had continuous-flow LVAD implantation and were supported for more than 6 months were investigated. AR grade (none, 0; trivial, 0.5; mild, 1; mild-moderate, 1.5; moderate, 2; moderate-severe, 2.5; severe, 3) and aortic valve opening were evaluated with echocardiography. Computed tomography was performed to all the patients postoperatively. The angle of the outflow graft to the aorta (O-A angle, parallel 0; tangent 90°, 0–180°), aortic diameter at the anastomosis site, sino-tubular junction (STJ) diameter, distance between the STJ and the anastomosis site, and distance between the anastomosis site and the brachiocephalic artery were measured. The patients’ age was 38?±?11 years. Support duration was 686?±?354 days. Mean AR grade after continuous-flow LVAD implantation was increased to around mild and was maintained thereafter. No patient needed any intervention to the aortic valve. The aortic valves of 82.6% of patients were closed continuously. The O-A angle (83?±?14) was positively correlated with maximum AR grade (p?=?0.0095). The O-A angle was significantly smaller in patients with maximum AR grade of 1 or less (77?±?9°) than in those with 1.5 or greater (94?±?15°, p?=?0.021). The other CT measurements had no correlation with AR grade. In conclusion, the O-A angle was correlated with AR grade progression. The O-A angle appears to be one of the important factors related to AR under continuous-flow LVAD support.     

Evaluation of a Semi-automatic Right Ventricle Segmentation Method on Short-Axis MR Images

The purpose of this study was to evaluate a semi-automatic right ventricle segmentation method on short-axis cardiac cine MR images which segment all right ventricle contours in a cardiac phase using one seed contour. Twenty-eight consecutive short-axis, four-chamber, and tricuspid valve view cardiac cine MRI examinations of healthy volunteers were used. Two independent observers performed the manual and automatic segmentations of the right ventricles. Analyses were based on the ventricular volume and ejection fraction of the right heart chamber. Reproducibility of the manual and semi-automatic segmentations was assessed using intra- and inter-observer variability. Validity of the semi-automatic segmentations was analyzed with reference to the manual segmentations. The inter- and intra-observer variability of manual segmentations were between 0.8 and 3.2%. The semi-automatic segmentations were highly correlated with the manual segmentations (R² 0.79–0.98), with median difference of 0.9–4.8% and of 3.3% for volume and ejection fraction parameters, respectively. In comparison to the manual segmentation, the semi-automatic segmentation produced contours with median dice metrics of 0.95 and 0.87 and median Hausdorff distance of 5.05 and 7.35 mm for contours at end-diastolic and end-systolic phases, respectively. The inter- and intra-observer variability of the semi-automatic segmentations were lower than observed in the manual segmentations. Both manual and semi-automatic segmentations performed better at the end-diastolic phase than at the end-systolic phase. The investigated semi-automatic segmentation method managed to produce a valid and reproducible alternative to manual right ventricle segmentation.     

一种基于2D/3D配准的脊柱术中校正方法

背景：脊柱术前三维影像有助于诊断和治疗,术中患者体位变化将引起脊柱形态改变,致使术前影像不能反映术中实际情况,无法确保手术的顺利实施。 目的：利用脊髓手术中影像校正术前脊柱模型形态。 方法：实验提出了一种基于2D/3D配准的脊柱术中校正方法,借助数字影像重建技术完成术前X射线图像与CT体数据的2D/3D配准,进一步完成术中、术前X射线图像中独立椎段的特征匹配,利用上述配准结果实现术前脊柱CT模型的术中快速校正。 结果与结论：采用附有标记的颈椎标本进行实验,校正后可基本消除术前脊柱模型与术中形态的偏差,其误差可控制在1 mm以内,能够满足医学临床要求。     

Minimizing artifacts resulting from respiratory and cardiac motion by optimization of the transmission scan in cardiac PET/CT

The introduction of positron emission/computed tomography (PET/CT) systems coupled with multidetector CT arrays has greatly increased the amount of clinical information in myocardial perfusion studies. The CT acquisition serves the dual role of providing high spatial anatomical detail and attenuation correction for PET. However, the differences between the interaction of respiratory and cardiac cycles in the CT and PET acquisitions presents a challenge when using the CT to determine PET attenuation correction. Three CT attenuation correction protocols were tested for their ability to produce accurate emission images: gated, a step mode acquisition covering the diastolic heart phase; normal, a high-pitch helical CT; and slow, a low-pitch, low-temporal-resolution helical CT. The amount of cardiac tissue in the emission image that overlaid lung tissue in the transmission image was used as the measure of mismatch between acquisitions. Phantom studies simulating misalignment of the heart between the transmission and emission sequences were used to correlate the amount of mismatch with the artificial defect changes in the emission image. Consecutive patients were studied prospectively with either paired gated (diastolic phase, 120 kVp, 280 mA, 2.6 s) and slow CT (0.562:1 pitch, 120 kVp, Auto-mA, 16 s) or paired normal (0.938:1 pitch, 120 kVp, Auto-mA, 4.8 s) and slow CT protocols, prior to a Rb-82 perfusion study. To determine the amount of mismatch, the transmission and emission images were converted to binary representations of attenuating tissue and cardiac tissue and overlaid using their native registration. The number of cardiac tissue pixels from the emission image present in the CT lung field yielded the magnitude of misalignment represented in terms of volume, of where a small volume indicates better registration. Acquiring a slow CT improved registration between the transmission and emission acquisitions compared to the gated and normal CT protocols. The volume of PET cardiac tissue in the CT lung field was significantly lower (p < 0.03) for the slow CT protocol in both the rest and stress emission studies. Phantom studies showed that an overlaying volume greater than 2.6 mL would produce significant artificial defects as determined by a quantitative software package that employs a normal database. The percentage of patient studies with overlaying volume greater than 2.6 mL was reduced from 71% with the normal CT protocol to 28% with the slow CT protocol. The remaining 28% exhibited artifacts consistent with heart drift and patient motion that could not be corrected by adjusting the CT acquisition protocol. The low pitch of the slow CT protocol provided the best match to the emission study and is recommended for attenuation correction in cardiac PET/CT studies. Further reduction in artifacts arising from cardiac drift is required and warrants an image registration solution.     

基于投影图像和CT容积图像的三维冠状动脉运动跟踪新方法

目的利用双平面X射线投影图像序列和参考CT容积图像进行冠状动脉的三维运动跟踪建模。方法①提取投影图像序列中的冠状动脉树;②利用多尺度滤波和血管函数提取CT容积图像中的动脉血管;③采用基于B样条配准的方法进行三维运动建模。结果将双投影图像中血管的运动估计结果的三维重建形态与三维运动模型预测出的结果进行了量化比较,调整配准的B样条参数后,两者误差处于有效范围之内。结论由此表明采用投影图像和对应容积图像的联合先验知识进行冠状动脉的三维运动建模是有效的。     

Statistically Deformable 2D/3D Registration for Estimating Post-operative Cup Orientation from a Single Standard AP X-ray Radiograph

The widely used procedure of estimating post-operative cup orientation based on a single standard AP X-ray radiograph is known inaccurate, largely due to the wide variability in individual pelvic orientation relative to X-ray plate. CT-based 2D/3D rigid image registration methods have been developed to measure post-operative cup orientation. Although encouraging results have been reported, their extensive usage in clinical routine is still limited. This may be explained by their requirement of having a CT study of the patient at some point during treatment, which is not available for vast majority of Total Hip Arthroplasty procedures performed nowadays. To address this limitation, this article proposes a statistically deformable 2D/3D registration approach for estimating post-operative cup orientation. No CT study of the patient is required any more. Compared to ground truths established from post-operative CT images, the cup orientations measured by the present technique in a cadaver experiment showed differences of 1.7 ± 1.4° for anteversion and difference of 1.5 ± 1.5° for inclination. When the present technique was evaluated on patients’ datasets, differences of 2.2 ± 1.3° and differences of 2.0 ± 0.8° were found for the anteversion and the inclination, respectively. The experimental results, though still preliminary, demonstrated the efficacy of the present approach.     

The influence of transducer position on the M-mode echocardiographic measurement of the left atrial dimension

Eighty-five subjects were investigated to see if the left atrial dimension (LAD) determined by M-mode echocardiography varies with different placements of the transducer in the parasternal long-axis view. Satisfactory images from the third and fourth intercostal spaces were obtained in all subjects. A satisfactory image from the second, third and fourth intercostal spaces was obtained in 58 subjects and from the third, fourth and fifth intercostal spaces in 27 subjects. The mean LAD was significantly smaller (P less than 0.001) in the fourth intercostal space than in the third. The mean value obtained from the fifth intercostal space was smaller than that from the fourth. The beam pathway seen on the two-dimensional echocardiogram indicates that the transducer should preferably be placed in the third intercostal space. If the transducer is placed in a lower intercostal space the measured left atrial dimension might be underestimated.     

Automatic segmentation of the aortic root in CT angiography of candidate patients for transcatheter aortic valve implantation

Transcatheter aortic valve implantation is a minimal-invasive intervention for implanting prosthetic valves in patients with aortic stenosis. Accurate automated sizing for planning and patient selection is expected to reduce adverse effects such as paravalvular leakage and stroke. Segmentation of the aortic root in CTA is pivotal to enable automated sizing and planning. We present a fully automated segmentation algorithm to extract the aortic root from CTA volumes consisting of a number of steps: first, the volume of interest is automatically detected, and the centerline through the ascending aorta and aortic root centerline are determined. Subsequently, high intensities due to calcifications are masked. Next, the aortic root is represented in cylindrical coordinates. Finally, the aortic root is segmented using 3D normalized cuts. The method was validated against manual delineations by calculating Dice coefficients and average distance error in 20 patients. The method successfully segmented the aortic root in all 20 cases. The mean Dice coefficient was 0.95 ± 0.03, and the mean radial absolute error was 0.74 ± 0.39 mm, where the interobserver Dice coefficient was 0.95 ± 0.03 and the mean error was 0.68 ± 0.34?mm. The proposed algorithm showed accurate results compared to manual segmentations.     

Automatic segmentation of left ventricle cavity from short-axis cardiac magnetic resonance images

In this paper, a computational framework is proposed to perform a fully automatic segmentation of the left ventricle (LV) cavity from short-axis cardiac magnetic resonance (CMR) images. In the initial phase, the region of interest (ROI) is automatically identified on the first image frame of the CMR slices. This is done by partitioning the image into different regions using a standard fuzzy c-means (FCM) clustering algorithm where the LV region is identified according to its intensity, size and circularity in the image. Next, LV segmentation is performed within the identified ROI by using a novel clustering method that utilizes an objective functional with a dissimilarity measure that incorporates a circular shape function. This circular shape-constrained FCM algorithm is able to differentiate pixels with similar intensity but are located in different regions (e.g. LV cavity and non-LV cavity), thus improving the accuracy of the segmentation even in the presence of papillary muscles. In the final step, the segmented LV cavity is propagated to the adjacent image frame to act as the ROI. The segmentation and ROI propagation are then iteratively executed until the segmentation has been performed for the whole cardiac sequence. Experiment results using the LV Segmentation Challenge validation datasets show that our proposed framework can achieve an average perpendicular distance (APD) shift of 2.23 ± 0.50 mm and the Dice metric (DM) index of 0.89 ± 0.03, which is comparable to the existing cutting edge methods. The added advantage over state of the art is that our approach is fully automatic, does not need manual initialization and does not require a prior trained model.     

On the Significance of Systolic Flow Waveform on Aortic Valve Energy Loss

This study aims to quantitatively and qualitatively assess energy dissipation in the aortic valve as a function of systolic aortic flow waveform representing pathologies where flow time-to-peak is delayed. A bioprosthetic valve was tested in the aortic position of a left-heart simulator under physiological pressure and flow conditions. The flow loop piston pump was programmed to generate three different flow waveforms each with a different peak time annotated as early peak (EP) with a rapid acceleration, mid peak (MP) and late peak (LP) with a rapid deceleration. Energy dissipation was calculated from flow and pressure measurements while sinus vorticity dynamics were evaluated using time-resolved planar particle image velocimetry. Average pressure gradients during systole are found 30.2?±?0.19, 30.7?±?0.25 and 32.9?±?0.29 mmHg and average dissipation over systole is found 0.95?±?0.026, 1.05?±?0.034 and 1.25?±?0.043 W for EP, MP and LP respectively. As systole’s acceleration phase is slower, sinus vortices are more likely to form, necessitating more energy exchange from shear layers inducing more viscous dissipation. EP found in healthy individuals is superior in terms of reducing energy dissipation and increasing aortic valve efficiency. In the context of possible left ventricular dysfunction and aortic stenosis, this means that delayed time-to-peak in the aortic flow waveform seen is not compensatory.     

Pulse sequence generated oblique magnetic resonance imaging: applications to cardiac imaging

A pulse sequence procedure for producing oblique magnetic resonance images is described. Using this procedure we present a new, accurate method to obtain true short-axis views and true long-axis views (both parallel and perpendicular to the septal plane) of the heart. The method is accurate regardless of the orientation of patient's heart. The method does not require the patient to be rotated, nor otherwise moved, and does not require any additional hardware. The method is experimentally verified with both human and phantom studies. The phantom study indicates accuracy of approximately 1 degree with a commercial scanner that reports angular measurements to a precision of 1 degree. Application of the short-axis views to measurement of left ventricular volume, and possible advantages of Gauss-Legendre integration for this measurement are discussed. Finally, multiphase oblique cardiac images are presented.     

基于自由变形法的多模态医学图像的配准与融合

本研究提出了一种自动识别颈部PET-CT图像特征点的算法，它应用自由变形（FFD）方法以CT图像的特征点为参考使PET图像产生变形，再结合最大互信息法对颈部PET与CT图像进行非刚体配准，最后用改进的小波图像融合法把两者进行融合得出视觉效果比较理想的融合图像。经实际计算得出的变形PET图像与对应CT图像的互信息量大于原始PET图像，并且最后用改进的小波图像融合法得出的融合图像的信息量比一般小波融合大，由此证明本研究所用方法是有效的。