Five-dimensional MRI incorporating simultaneous resolution of cardiac and respiratory phases for volumetric imaging

Three-dimensional sector-wise golden angle-improved k-space uniformity after electrocardiogram binning

PURPOSE

To develop and evaluate a 3D sector-wise golden-angle (3D-SWIG) profile ordering scheme for cardiovascular MR cine imaging that maintains high k-space uniformity after electrocardiogram (ECG) binning.

METHOD

Cardiovascular MR (CMR) was performed at 1.5 T. A balanced SSFP pulse sequence was implemented with a novel 3D-SWIG radial ordering, where k-space was divided into wedges, and each wedge was acquired in a separate heartbeat. The high uniformity of k-space coverage after physiological binning can be used to perform functional imaging using a very short acquisition. The 3D-SWIG was compared with two commonly used 3D radial trajectories for CMR (i.e., double golden angle and spiral phyllotaxis) in numerical simulations. Free-breathing 3D-SWIG and conventional breath-held 2D cine were compared in patients (n = 17) referred clinically for CMR. Quantitative comparison was performed based on left ventricular segmentation.

RESULTS

Numerical simulations showed that 3D-SWIG both required smaller steps between successive readouts and achieved better k-space sampling uniformity after binning than either the double golden angle or spiral phyllotaxis trajectories. In vivo evaluation showed that measurements of left ventricular ejection fraction calculated from a 48 heart-beat free-breathing 3D-SWIG acquisition were highly reproducible and agreed with breath-held 2D-Cartesian cine (mean ± SD difference of -3.1 ± 3.5% points).

CONCLUSIONS

The 3D-SWIG acquisition offers a simple solution for highly improved k-space uniformity after physiological binning. The feasibility of the 3D-SWIG method is demonstrated in this study through whole-heart cine imaging during free breathing with an acquisition time of less than 1 min.

Does merged three-dimensional mapping improve contact force and long-term procedure outcome in atrial fibrillation ablation? (MICRO-AF study): a prospective randomized controlled study

Integration of electroanatomical map (EAM) with preacquired three-dimensional (3D) cardiac images provides detailed appreciation of the complex anatomy of the left atrium (LA) and pulmonary vein (PV). High-density (HD) multi-electrode mapping catheters have enabled creating more accurate EAM reflecting real-time volume-rendered LA-PV geometry during atrial fibrillation (AF) ablation. However, no study has compared the outcomes of AF ablation using HD-EAM versus 3D-merged map. We aimed to investigate the procedural and clinical outcomes of AF ablation with HD-EAM (HD-EAM group) versus 3D-merged map (Merge group). One hundred patients (59.5 ± 11.5 years, 53% with paroxysmal AF [PAF]) were randomly assigned (1:1) to HD-EAM or Merged group. HD multi-electrode mapping and contact force (CF)-sensing catheters were used to create virtual LA-PV chamber and to perform wide antral circumferential ablation (WACA), respectively. The two groups showed no significant differences in baseline characteristics and procedural data including ablation time, fluoroscopy time, LA voltage, and CF. PV isolation with a single WACA line was achieved in 21 (42%) and 27 (54%) patients in the Merge and HD-EAM groups, respectively (P = NS). CF was significantly lower in lesions with gap than lesions without gap after a single WACA (7.3 ± 7.3 g vs. 16.0 ± 8.3, respectively, P < 0.001). During the 12-month follow-up, no significant difference in AF recurrence was observed between two groups, irrespective of AF type. In multivariate analysis, non-PAF was an independent risk factor for AF recurrence. Integration of 3D cardiac imaging did not improve procedural and clinical outcomes. HD-EAM provides an accurate real-time LA geometry.

Medical decision making for 5D cardiac model: Template matching technique and simulation of the fifth dimension

The purpose of this paper is to develop a 5D cardiac model which is inspired from the 5D model for the lungs. This model depends on five variables: the anatomical structure of the 3D heart, temporal dimension and the function of blood flow as the fifth dimension. To test this hypothesis, we took the same mathematical modeling as a reference for the fifth dimension of pulmonary flow where r→_ρ(t)=r→_v(t)+r_f→(t) wherer→_v(t) is the displacement vectors with approximate magnitudes by linear functions of the tidal volume and r_f→(t) is the blood flow. The scans were acquired for 10 patients,in the 404 series for a total of 18,483 images studied in three cases: healthy patient, case of heart failure and aortic stenosis. Where r→_vand r→_f are the unit vectors along the volume of ejection and the blood flow axes, indicating the direction of motion of the object due to heart volume ejection and blood flow variations, respectively. The quantities of α and β coefficients are determined from real-time patient image data. The alpha and beta coefficients are derived from the following dimension equations[mm / ml] [mm*ms / ml] . Since the cardiac system has two diastolic and systolic phases, we have calculated α₁ and β₁ for telediatolic volume and α₂ and β₂ for telesystolic volume throughout the cardiac cycle as a function of the location of the cuts chosen randomly. Fifth-dimensional experiments are used to track, simulate the behavior of blood flow to detect preliminary indications for the identification of stenosis or valve leakage. The average discrepancy was tabulated as the global fraction of systolic ejection. The results shown in Fig. 3 detect a correspondence between the hunting chamber cut and the flow sequence through the orifice of aorta for this patient with suspicious of having an aortic stenosis disease and an ejection fraction about 71% with a maximum of velocity (Vmax) detected=250 (cm / ms) = 2.5 (m / 10-3 s). In this case this patient has a minor stenosis in the aorta. It should be referred that the normalization of this measures is classified such as : Minor stenosis: area 1.5 cm2, Vmax <3 m / moderate stenosis: area 1.0 - 1.5 cm2, Vmax 3 - 4 m / severe stenosis: area <1.0 cm2, Vmax> 4 m / s. For a patient who has an aortic stenosis the cloud of the points is accumulated comparing to the origin of the axis while the patient with a symptom of insufficiency the points are widened with a remarkable gap in the trajectory. To solve the issue of the bad prediction, the inaccuracy of the clouds points of the model 5D, the lack of the exact measurements to estimate the degree of cardiac insufficiency (leakage or stenosis), a solution of 5D imagery was depicted. Our main contribution is to test the validity of the template-matching algorithm and the fifth dimension simulation to provide more clues to detect the aortic stenosis and cardiac insufficiency in the context of medical decision support.

Extending Cardiac Functional Assessment with Respiratory-Resolved 3D Cine MRI

This study aimed to develop a cardiorespiratory-resolved 3D magnetic resonance imaging (5D MRI: x-y-z-cardiac-respiratory) approach based on 3D motion tracking for investigating the influence of respiration on cardiac ventricular function. A highly-accelerated 2.5-minute sparse MR protocol was developed for a continuous acquisition of cardiac images through multiple cardiac and respiratory cycles. The heart displacement along respiration was extracted using a 3D image deformation algorithm, and this information was used to cluster the acquired data into multiple respiratory phases. The proposed approach was tested in 15 healthy volunteers (7 females). Cardiac function parameters, including the end-systolic volume (ESV), end-diastolic volume (EDV), stroke volume (SV), and ejection fraction (EF), were measured for the left and right ventricle in both end-expiration and end-inspiration. Although with the proposed 5D cardiac MRI, there were no significant differences (p > 0.05, t-test) between end-expiration and end-inspiration measurements of the cardiac function in volunteers, incremental respiratory motion parameters that were derived from 3D motion tracking, such as the depth, expiration and inspiration distribution, correlated (p < 0.05, correlation coefficient, Mann-Whitney) with those volume-based parameters of cardiac function and varied between genders. The obtained initial results suggested that this new approach allows evaluation of cardiac function during specific respiratory phases. Thus, it can enable investigation of effects related to respiratory variability and better assessment of cardiac function for studying respiratory and/or cardiac dysfunction.

Projection-based respiratory-resolved left ventricular volume measurements in patients using free-breathing double golden-angle 3D radial acquisition

Objective

To refine a new technique to measure respiratory-resolved left ventricular end-diastolic volume (LVEDV) in mid-inspiration and mid-expiration using a respiratory self-gating technique and demonstrate clinical feasibility in patients.

Materials and methods

Ten consecutive patients were imaged at 1.5 T during 10 min of free breathing using a 3D golden-angle radial trajectory. Two respiratory self-gating signals were extracted and compared: from the k-space center of all acquired spokes, and from a superior–inferior projection spoke repeated every 64 ms. Data were binned into end-diastole and two respiratory phases of 15% respiratory cycle duration in mid-inspiration and mid-expiration. LVED volume and septal–lateral diameter were measured from manual segmentation of the endocardial border.

Results

Respiratory-induced variation in LVED size expressed as mid-inspiration relative to mid-expiration was, for volume, 1 ± 8% with k-space-based self-gating and 8 ± 2% with projection-based self-gating (P = 0.04), and for septal–lateral diameter, 2 ± 2% with k-space-based self-gating and 10 ± 1% with projection-based self-gating (P = 0.002).

Discussion

Measuring respiratory variation in LVED size was possible in clinical patients with projection-based respiratory self-gating, and the measured respiratory variation was consistent with previous studies on healthy volunteers. Projection-based self-gating detected a higher variation in LVED volume and diameter during respiration, compared to k-space-based self-gating.

Left ventricular volume measurements with free breathing respiratory self-gated 3-dimensional golden angle radial whole-heart cine imaging - Feasibility and reproducibility

PURPOSE

To develop and evaluate a free breathing respiratory self-gated isotropic resolution technique for left ventricular (LV) volume measurements.

METHODS

A 3D radial trajectory with double golden-angle ordering was used for free-running data acquisition during free breathing in 9 healthy volunteers. A respiratory self-gating signal was extracted from the center of k-space and used with the electrocardiogram to bin all data into 3 respiratory and 25 cardiac phases. 3D image volumes were reconstructed and the LV endocardial border was segmented. LV volume measurements and reproducibility from 3D free breathing cine were compared to conventional 2D breath-held cine.

RESULTS

No difference was found between 3D free breathing cine and 2D breath-held cine with regards to LV ejection fraction, stroke volume, end-systolic volume and end-diastolic volume (P<0.05 for all). The test-retest differences did not differ between 3D free breathing cine and 2D breath-held cine (P<0.05 for all).

CONCLUSION

3D free breathing cine and conventional 2D breath-held cine showed similar values and test-retest repeatability for LV volumes in healthy volunteers. 3D free breathing cine enabled retrospective sorting and arbitrary angulation of isotropic data, and could correctly measure LV volumes during free breathing acquisition.

Interaction between respiration and right versus left ventricular volumes at rest and during exercise: a real-time cardiac magnetic resonance study

Breathing-induced changes in intrathoracic pressures influence left ventricular (LV) and right ventricular (RV) volumes, the exact nature and extent of which have not previously been evaluated in humans. We sought to examine this “respiratory pump” using novel real-time cardiac magnetic resonance (CMR) imaging. Eight healthy subjects underwent serial multislice real-time CMR during normal breathing, breath holding, and the Valsalva maneuver. Subsequently, a separate cohort of nine subjects underwent real-time CMR at rest and during incremental exercise. LV and RV end-diastolic volume (EDV) and end-systolic volume (ESV) and diastolic and systolic eccentricity indexes were determined at peak inspiration and expiration. During normal breathing, inspiration resulted in an increase in RV volumes [RVEDV: +18 ± 8%, RVESV: +14 ± 12%, and RV stroke volume (SV): +21 ± 10%, P < 0.01] and an opposing decrease in LV volumes ( P < 0.0001 for interaction). During end-inspiratory breath holding, RV SV decreased by 9 ± 10% ( P = 0.046), whereas LV SV did not change. During the Valsalva maneuver, volumes decreased in both ventricles (RVEDV: −29 ± 11%, RVESV: −16 ± 14%, RV SV: −36 ± 14%, LVEDV: −22 ± 17%, and LV SV: −25 ± 17%, P < 0.01). The reciprocal effect of respiration on LV and RV volumes was maintained throughout exercise. The diastolic and systolic eccentricity indexes were greater during inspiration than during expiration, both at rest and during exercise ( P < 0.0001 for both). In conclusion, ventricular volumes oscillate with respiratory phase such that RV and LV volumes are maximal at peak inspiration and expiration, respectively. Thus, interpretation of RV versus LV volumes requires careful definition of the exact respiratory time point for proper interpretation, both at rest and during exercise.

Medical image processing on the GPU - past, present and future

Graphics processing units (GPUs) are used today in a wide range of applications, mainly because they can dramatically accelerate parallel computing, are affordable and energy efficient. In the field of medical imaging, GPUs are in some cases crucial for enabling practical use of computationally demanding algorithms. This review presents the past and present work on GPU accelerated medical image processing, and is meant to serve as an overview and introduction to existing GPU implementations. The review covers GPU acceleration of basic image processing operations (filtering, interpolation, histogram estimation and distance transforms), the most commonly used algorithms in medical imaging (image registration, image segmentation and image denoising) and algorithms that are specific to individual modalities (CT, PET, SPECT, MRI, fMRI, DTI, ultrasound, optical imaging and microscopy). The review ends by highlighting some future possibilities and challenges.

Effect of catheter tip-tissue surface contact on three-dimensional left atrial and pulmonary vein geometries: potential anatomic distortion of 3D ultrasound, fast anatomical mapping, and merged 3D CT-derived images

Anatomic Distortion of 3D Mapping. 

BACKGROUND

Although catheter tip-tissue contact is known as a reliable basis for mapping and ablation of atrial fibrillation (AF), the effects of different mapping methods on 3-dimensional (3D) map configuration remain unknown.

METHODS AND RESULTS

Twenty AF patients underwent Carto-based 3D ultrasound (US) evaluation. Left atrium (LA)/pulmonary vein (PV) geometry was constructed with the 3D US system. The resulting geometry was compared to geometries created with a fast electroanatomical mapping (FAM) algorithm and 3D US merged with computed tomography (merged 3D US-CT). The 3D US-derived LA volumes were smaller than the FAM- and merged 3D US-CT-derived volumes (75 ± 21 cm(3) vs 120 ± 20 cm(3) and 125 ± 25 cm(3) , P < 0.0001 for both). Differences in anatomic PV orifice fiducials between 3D US- and FAM- and merged 3D US-CT-derived geometries were 6.0 (interquartile range 0-9.3) mm and 4.1 (0-7.0) mm, respectively. Extensive encircling PV isolation guided by 3D US images with real-time 2D intracardiac echocardiography-based visualization of catheter tip-tissue contact generated ablation point (n = 983) drop-out at 1.9 ± 3.8 mm beyond the surface of the 3D US-derived LA/PV geometry. However, these same points were located 1.5 ± 5.4 and 0.4 ± 4.1 mm below the FAM- and merged 3D US-CT-derived surfaces.

CONCLUSIONS

Different mapping methods yield different 3D geometries. When AF ablation is guided by 3D US-derived images, ablation points fall beyond the 3D US surface but below the FAM- or merged 3D US-CT-derived surface. Our data reveal anatomic distortion of 3D images, providing important information for improving the safety and efficacy of 3D mapping-guided AF ablation. (J Cardiovasc Electrophysiol, Vol. 24, pp. 259-266, March 2013).

True 4D Image Denoising on the GPU

The use of image denoising techniques is an important part of many medical imaging applications. One common application is to improve the image quality of low-dose (noisy) computed tomography (CT) data. While 3D image denoising previously has been applied to several volumes independently, there has not been much work done on true 4D image denoising, where the algorithm considers several volumes at the same time. The problem with 4D image denoising, compared to 2D and 3D denoising, is that the computational complexity increases exponentially. In this paper we describe a novel algorithm for true 4D image denoising, based on local adaptive filtering, and how to implement it on the graphics processing unit (GPU). The algorithm was applied to a 4D CT heart dataset of the resolution 512  × 512  × 445  × 20. The result is that the GPU can complete the denoising in about 25 minutes if spatial filtering is used and in about 8 minutes if FFT-based filtering is used. The CPU implementation requires several days of processing time for spatial filtering and about 50 minutes for FFT-based filtering. The short processing time increases the clinical value of true 4D image denoising significantly.

Reconstruction from free-breathing cardiac MRI data using reproducing kernel Hilbert spaces

This paper describes a rigorous framework for reconstructing MR images of the heart, acquired continuously over the cardiac and respiratory cycle. The framework generalizes existing techniques, commonly referred to as retrospective gating, and is based on the properties of reproducing kernel Hilbert spaces. The reconstruction problem is formulated as a moment problem in a multidimensional reproducing kernel Hilbert spaces (a two-dimensional space for cardiac and respiratory resolved imaging). Several reproducing kernel Hilbert spaces were tested and compared, including those corresponding to commonly used interpolation techniques (sinc-based and splines kernels) and a more specific kernel allowed by the framework (based on a first-order Sobolev RKHS). The Sobolev reproducing kernel Hilbert spaces was shown to allow improved reconstructions in both simulated and real data from healthy volunteers, acquired in free breathing.

Three-dimensional cine MRI in free-breathing infants and children with congenital heart disease

BACKGROUND

Patients with congenital heart disease frequently have complex cardiac and vascular malformations requiring detailed non-invasive diagnostic evaluation including functional parameters.

OBJECTIVE

To evaluate the morphological and functional information provided by a novel 3-D cine steady-state free-precession (SSFP) sequence.

MATERIALS AND METHODS

Twenty consecutive children (mean age 2.2 years, nine boys) were examined using a 1.5-T MR system including 2-D cine gradient-recalled-echo sequences, static 3-D SSFP and 3-D cine SSFP sequences.

RESULTS

Measurement of ventricular structures and volumes showed close agreement between the 3-D cine SSFP sequence and the 2-D cine gradient-recalled-echo and static 3-D SSFP sequences (left ventricular volumes mean difference 1.0-1.9 ml and 8.8-11.4%, respectively; right ventricular volumes 1.7-2.1 ml and 9.9-16.9%, respectively). No systematic bias was observed.

CONCLUSION

3-D cine MRI provides anatomic as well as functional information with sufficient spatial and temporal resolution in free-breathing infants with congenital heart disease.

Three-Dimensional Ultrasound for Image-Guided Mapping and Intervention

Background— Multiple factors create discrepancies between electroanatomic maps and merged, preacquired computed tomographic images used in guiding atrial fibrillation ablation. Therefore, a Carto-based 3D ultrasound image system (Biosense Webster Inc) was validated in an animal model and tested in 15 atrial fibrillation patients.

Methods and Results— Twelve dogs underwent evaluation using a newly developed Carto-based 3D ultrasound system. After fiducial clip markers were percutaneously implanted at critical locations in each cardiac chamber, 3D ultrasound geometries, derived from a family of 2D intracardiac echocardiographic images, were constructed. Point-source error of 3D ultrasound-derived geometries, assessed by actual real-time 2D intracardiac echocardiographic clip sites, was 2.1�1.1 mm for atrial and 2.4�1.2 mm for ventricular sites. These errors were significantly less than the variance on CartoMerge computed tomographic images (atria: 3.3�1.6 mm; ventricles: 4.8�2.0 mm; P< 0.001 for both). Target ablation at each clip, guided only by 3D ultrasound-derived geometry, resulted in lesions within 1.1�1.1 mm of the actual clips. Pulmonary vein ablation guided by 3D ultrasound-derived geometry resulted in circumferential ablative lesions. Mapping in 15 patients produced modestly smaller 3D ultrasound versus electroanatomic map left atrial volumes (98�24 cm 3 versus 109�25 cm 3 , P< 0.05). Three-dimensional ultrasound-guided pulmonary vein isolation and linear ablation in these patients were successfully performed with confirmation of pulmonary vein entrance/exit block.

Conclusions— These data demonstrate that 3D ultrasound images seamlessly yield anatomically accurate chamber geometries. Image volumes from the ultrasound system are more accurate than possible with CartoMerge computed tomographic imaging. This clinical study also demonstrates the initial feasibility of this guidance system for ablation in patients with atrial fibrillation.

Advances in 4D medical imaging and 4D radiation therapy

This paper reviews recent advances in 4D medical imaging (4DMI) and 4D radiation therapy (4DRT), which study, characterize, and minimize patient motion during the processes of imaging and radiotherapy. Patient motion is inevitably present in these processes, producing artifacts and uncertainties in target (lesion) identification, delineation, and localization. 4DMI includes time-resolved volumetric CT, MRI, PET, PET/CT, SPECT, and US imaging. To enhance the performance of these volumetric imaging techniques, parallel multi-detector array has been employed for acquiring image projections and the volumetric image reconstruction has been advanced from the 2D to the 3D tomography paradigm. The time information required for motion characterization in 4D imaging can be obtained either prospectively or retrospectively using respiratory gating or motion tracking techniques. The former acquires snapshot projections for reconstructing a motion-free image. The latter acquires image projections continuously with an associated timestamp indicating respiratory phases using external surrogates and sorts these projections into bins that represent different respiratory phases prior to reconstructing the cyclical series of 3D images. These methodologies generally work for all imaging modalities with variations in detailed implementation. In 4D CT imaging, both multi-slice CT (MSCT) and cone-beam CT (CBCT) are applicable in 4D imaging. In 4D MR imaging, parallel imaging with multi-coil-detectors has made 4D volumetric MRI possible. In 4D PET and SPECT, rigid and non-rigid motions can be corrected with aid of rigid and deformable registration, respectively, without suffering from low statistics due to signal binning. In 4D PET/CT and SPECT/CT, a single set of 4D images can be utilized for motion-free image creation, intrinsic registration, and attenuation correction. In 4D US, volumetric ultrasonography can be employed to monitor fetal heart beating with relatively high temporal resolution. 4DRT aims to track and compensate for target motion during radiation treatment, minimizing normal tissue injury, especially critical structures adjacent to the target, and/or maximizing radiation dose to the target. 4DRT requires 4DMI, 4D radiation treatment planning (4D RTP), and 4D radiation treatment delivery (4D RTD). Many concepts in 4DRT are borrowed, adapted and extended from existing image-guided radiation therapy (IGRT) and adaptive radiation therapy (ART). The advantage of 4DRT is its promise of sparing additional normal tissue by synchronizing the radiation beam with the moving target in real-time. 4DRT can be implemented differently depending upon how the time information is incorporated and utilized. In an ideal situation, the motion adaptive approach guided by 4D imaging should be applied to both RTP and RTD. However, until new automatic planning and motion feedback tools are developed for 4DRT, clinical implementation of ideal 4DRT will meet with limited success. However, simplified forms of 4DRT have been implemented with minor modifications of existing planning and delivery systems. The most common approach is the use of gating techniques in both imaging and treatment, so that the planned and treated target localizations are identical. In 4D planning, the use of a single planning CT image, which is representative of the statistical respiratory mean, seems preferable. In 4D delivery, on-site CBCT imaging or 3D US localization imaging for patient setup and internal fiducial markers for target motion tracking can significantly reduce the uncertainty in treatment delivery, providing improved normal tissue sparing. Most of the work on 4DRT can be regarded as a proof-of-principle and 4DRT is still in its early stage of development.

k-t2 BLAST: exploiting spatiotemporal structure in simultaneously cardiac and respiratory time-resolved volumetric imaging

Multidimensional imaging resolving both the cardiac and respiratory cycles simultaneously has the potential to describe important physiological interdependences between the heart and pulmonary processes. A fully five-dimensional acquisition with three spatial and two temporal dimensions is hampered, however, by the long acquisition time and low spatial resolution. A technique is proposed to reduce the scan time substantially by extending the k-t BLAST framework to two temporal dimensions. By sampling the k-t space sparsely in a lattice grid, the signal in the transform domain, x-f space, can be densely packed, exploiting the fact that large regions in the field of view have low temporal bandwidth. A volumetric online prospective triggering approach with full cardiac and respiratory cycle coverage was implemented. Retrospective temporal interpolation was used to refine the timing estimates for the center of k-space, which is sampled for all cardiac and respiratory time frames. This resulted in reduced reconstruction error compared with conventional k-t BLAST reconstruction. The k-t(2) BLAST technique was evaluated by decimating a fully sampled five-dimensional data set, and feasibility was further demonstrated by performing sparsely sampled acquisitions. Compared to the fully sampled data, a fourfold improvement in spatial resolution was accomplished in approximately half the scan time.