Deep learning-based Accelerated and Noise-Suppressed Estimation (DANSE) of quantitative Gradient-Recalled Echo (qGRE) magnetic resonance imaging metrics associated with human brain neuronal structure and hemodynamic properties.

The purpose of the current study was to introduce a Deep learning-based Accelerated and Noise-Suppressed Estimation (DANSE) method for reconstructing quantitative maps of biological tissue cellular-specific, R2t*, and hemodynamic-specific, R2', metrics of quantitative gradient-recalled echo (qGRE) MRI. The DANSE method adapts a supervised learning paradigm to train a convolutional neural network for robust estimation of R2t* and R2' maps with significantly reduced sensitivity to noise and the adverse effects of macroscopic (B0 ) magnetic field inhomogeneities directly from the gradient-recalled echo (GRE) magnitude images. The R2t* and R2' maps for training were generated by means of a voxel-by-voxel fitting of a previously developed biophysical quantitative qGRE model accounting for tissue, hemodynamic, and B0 -inhomogeneities contributions to multigradient-echo GRE signal using a nonlinear least squares (NLLS) algorithm. We show that the DANSE model efficiently estimates the aforementioned qGRE maps and preserves all the features of the NLLS approach with significant improvements including noise suppression and computation speed (from many hours to seconds). The noise-suppression feature of DANSE is especially prominent for data with low signal-to-noise ratio (SNR ~ 50-100), where DANSE-generated R2t* and R2' maps had up to three times smaller errors than that of the NLLS method. The DANSE method enables fast reconstruction of qGRE maps with significantly reduced sensitivity to noise and magnetic field inhomogeneities. The DANSE method does not require any information about field inhomogeneities during application. It exploits spatial and gradient echo time-dependent patterns in the GRE data and previously gained knowledge from the biophysical model, thus producing high quality qGRE maps, even in environments with high noise levels. These features along with fast computational speed can lead to broad qGRE clinical and research applications.

Learning-based motion artifact removal networks for quantitative R2∗ mapping.

PURPOSE: To introduce two novel learning-based motion artifact removal networks (LEARN) for the estimation of quantitative motion- and B0 -inhomogeneity-corrected R2∗ maps from motion-corrupted multi-Gradient-Recalled Echo (mGRE) MRI data. METHODS: We train two convolutional neural networks (CNNs) to correct motion artifacts for high-quality estimation of quantitative B0 -inhomogeneity-corrected R2∗ maps from mGRE sequences. The first CNN, LEARN-IMG, performs motion correction on complex mGRE images, to enable the subsequent computation of high-quality motion-free quantitative R2∗ (and any other mGRE-enabled) maps using the standard voxel-wise analysis or machine learning-based analysis. The second CNN, LEARN-BIO, is trained to directly generate motion- and B0 -inhomogeneity-corrected quantitative R2∗ maps from motion-corrupted magnitude-only mGRE images by taking advantage of the biophysical model describing the mGRE signal decay. RESULTS: We show that both CNNs trained on synthetic MR images are capable of suppressing motion artifacts while preserving details in the predicted quantitative R2∗ maps. Significant reduction of motion artifacts on experimental in vivo motion-corrupted data has also been achieved by using our trained models. CONCLUSION: Both LEARN-IMG and LEARN-BIO can enable the computation of high-quality motion- and B0 -inhomogeneity-corrected R2∗ maps. LEARN-IMG performs motion correction on mGRE images and relies on the subsequent analysis for the estimation of R2∗ maps, while LEARN-BIO directly performs motion- and B0 -inhomogeneity-corrected R2∗ estimation. Both LEARN-IMG and LEARN-BIO jointly process all the available gradient echoes, which enables them to exploit spatial patterns available in the data. The high computational speed of LEARN-BIO is an advantage that can lead to a broader clinical application.

Quantitative Gradient Echo MRI Identifies Dark Matter as a New Imaging Biomarker of Neurodegeneration that Precedes Tisssue Atrophy in Early Alzheimer's Disease.

BACKGROUND: Currently, brain tissue atrophy serves as an in vivo MRI biomarker of neurodegeneration in Alzheimer's disease (AD). However, postmortem histopathological studies show that neuronal loss in AD exceeds volumetric loss of tissue and that loss of memory in AD begins when neurons and synapses are lost. Therefore, in vivo detection of neuronal loss prior to detectable atrophy in MRI is essential for early AD diagnosis. OBJECTIVE: To apply a recently developed quantitative Gradient Recalled Echo (qGRE) MRI technique for in vivo evaluation of neuronal loss in human hippocampus. METHODS: Seventy participants were recruited from the Knight Alzheimer Disease Research Center, representing three groups: Healthy controls [Clinical Dementia Rating® (CDR®) = 0, amyloid ß (Aß)-negative, n = 34]; Preclinical AD (CDR = 0, Aß-positive, n = 19); and mild AD (CDR = 0.5 or 1, Aß-positive, n = 17). RESULTS: In hippocampal tissue, qGRE identified two types of regions: one, practically devoid of neurons, we designate as "Dark Matter", and the other, with relatively preserved neurons, "Viable Tissue". Data showed a greater loss of neurons than defined by atrophy in the mild AD group compared with the healthy control group; neuronal loss ranged between 31% and 43%, while volume loss ranged only between 10% and 19%. The concept of Dark Matter was confirmed with histopathological study of one participant who underwent in vivo qGRE 14 months prior to expiration. CONCLUSION: In vivo qGRE method identifies neuronal loss that is associated with impaired AD-related cognition but is not recognized by MRI measurements of tissue atrophy, therefore providing new biomarkers for early AD detection.

In vivo evaluation of heme and non-heme iron content and neuronal density in human basal ganglia.

Non-heme iron is an important element supporting the structure and functioning of biological tissues. Imbalance in non-heme iron can lead to different neurological disorders. Several MRI approaches have been developed for iron quantification relying either on the relaxation properties of MRI signal or measuring tissue magnetic susceptibility. Specific quantification of the non-heme iron can, however, be constrained by the presence of the heme iron in the deoxygenated blood and contribution of cellular composition. The goal of this paper is to introduce theoretical background and experimental MRI method allowing disentangling contributions of heme and non-heme irons simultaneously with evaluation of tissue neuronal density in the iron-rich basal ganglia. Our approach is based on the quantitative Gradient Recalled Echo (qGRE) MRI technique that allows separation of the total R2* metric characterizing decay of GRE signal into tissue-specific (R2t*) and the baseline blood oxygen level-dependent (BOLD) contributions. A combination with the QSM data (also available from the qGRE signal phase) allowed further separation of the tissue-specific R2t* metric in a cell-specific and non-heme-iron-specific contributions. It is shown that the non-heme iron contribution to R2t* relaxation can be described with the previously developed Gaussian Phase Approximation (GPA) approach. qGRE data were obtained from 22 healthy control participants (ages 26-63 years). Results suggest that the ferritin complexes are aggregated in clusters with an average radius about 100nm comprising approximately 2600 individual ferritin units. It is also demonstrated that the concentrations of heme and non-heme iron tend to increase with age. The strongest age effect was seen in the pallidum region, where the highest age-related non-heme iron accumulation was observed.

Deep learning using a biophysical model for robust and accelerated reconstruction of quantitative, artifact-free and denoised R2* images.

PURPOSE: To introduce a novel deep learning method for Robust and Accelerated Reconstruction (RoAR) of quantitative and B0-inhomogeneity-corrected R2* maps from multi-gradient recalled echo (mGRE) MRI data. METHODS: RoAR trains a convolutional neural network (CNN) to generate quantitative R2∗ maps free from field inhomogeneity artifacts by adopting a self-supervised learning strategy given (a) mGRE magnitude images, (b) the biophysical model describing mGRE signal decay, and (c) preliminary-evaluated F-function accounting for contribution of macroscopic B0 field inhomogeneities. Importantly, no ground-truth R2* images are required and F-function is only needed during RoAR training but not application. RESULTS: We show that RoAR preserves all features of R2* maps while offering significant improvements over existing methods in computation speed (seconds vs. hours) and reduced sensitivity to noise. Even for data with SNR = 5 RoAR produced R2* maps with accuracy of 22% while voxel-wise analysis accuracy was 47%. For SNR = 10 the RoAR accuracy increased to 17% vs. 24% for direct voxel-wise analysis. CONCLUSIONS: RoAR is trained to recognize the macroscopic magnetic field inhomogeneities directly from the input magnitude-only mGRE data and eliminate their effect on R2∗ measurements. RoAR training is based on the biophysical model and does not require ground-truth R2* maps. Since RoAR utilizes signal information not just from individual voxels but also accounts for spatial patterns of the signals in the images, it reduces the sensitivity of R2* maps to the noise in the data. These features plus high computational speed provide significant benefits for the potential usage of RoAR in clinical settings.

A convenient, reliable, and fast acoustic pressure field measurement method for magnetic resonance-guided high-intensity focused ultrasound systems with phased array transducers.

BACKGROUND: With the expanding applications of magnetic resonance-guided high-intensity focused ultrasound (MR-HIFU), there is an urgent need for a convenient, reliable, and fast acoustic pressure field measurement method to aid treatment protocol design, ensure consistent and safe operation of the transducer, and facilitate regulatory approval of new techniques. Herein, we report a method for acoustic pressure field characterization of MR-HIFU systems with multi-element phased array transducers. This method integrates fiber-optic hydrophone measurements and electronic steering of the ultrasound beam with MRI-assisted HIFU focus alignment to the fiber tip. METHODS: A clinical MR-HIFU system (Sonalleve V2, Profound Medical Inc., Mississauga, Canada) was used to assess the proposed method. A fiber-optic hydrophone was submerged in a degassed water bath, and the fiber tip location was traced using MRI. Subsequently, the nominal transducer focal point indicated on the MR-HIFU therapy planning software was positioned at the fiber tip, and the HIFU focus was electronically steered around the fiber tip within a 3D volume for 3D pressure field mapping, eliminating the need for an additional, expensive, and MRI-compatible 3D positioning stage. The peak positive and negative pressures were measured at the focus and validated using a standard hydrophone measurement setup outside the MRI magnet room. RESULTS: We found that the initial MRI-assisted HIFU focus alignment had an average offset of 2.23 ± 1.33 mm from the fiber tip as identified by the 3D pressure field mapping. MRI guidance and electronic beam steering allowed 3D focus localization within ~ 1 h, i.e., faster than the typical time required using the standard laboratory setup (~ 3-4 h). Acoustic pressures measured using the proposed method were not significantly different from those obtained with the standard laboratory hydrophone measurements. CONCLUSIONS: In conclusion, our method offers a convenient, reliable, and fast acoustic pressure field characterization tool for MR-HIFU systems with phased array transducers.

Focused ultrasound-enabled delivery of radiolabeled nanoclusters to the pons.

The goal of this study was to establish the feasibility of integrating focused ultrasound (FUS)-mediated delivery of 64Cu-integrated gold nanoclusters (64Cu-AuNCs) to the pons for in vivo quantification of the nanocluster brain uptake using positron emission tomography (PET) imaging. FUS was targeted at the pons for the blood-brain barrier (BBB) disruption in the presence of systemically injected microbubbles, followed by the intravenous injection of 64Cu-AuNCs. The spatiotemporal distribution of the 64Cu-AuNCs in the brain was quantified using in vivo microPET/CT imaging at different time points post injection. Following PET imaging, the accumulation of radioactivity in the pons was further confirmed using autoradiography and gamma counting, and the gold concentration was quantified using inductively coupled plasma-mass spectrometry (ICP-MS). We found that the noninvasive and localized BBB opening by the FUS successfully delivered the 64Cu-AuNCs to the pons. We also demonstrated that in vivo real-time microPET/CT imaging was a reliable method for monitoring and quantifying the brain uptake of 64Cu-AuNCs delivered by the FUS. This drug delivery platform that integrates FUS, radiolabeled nanoclusters, and PET imaging provides a new strategy for noninvasive and localized nanoparticle delivery to the pons with concurrent in vivo quantitative imaging to evaluate delivery efficiency. The long-term goal is to apply this drug delivery platform to the treatment of pontine gliomas.

Acoustic field characterization of a clinical magnetic resonance-guided high-intensity focused ultrasound system inside the magnet bore.

PURPOSE: With the expanding clinical application of magnetic resonance-guided high-intensity focused ultrasound (MR-HIFU), acoustic field characterization of MR-HIFU systems is needed for facilitating regulatory approval and ensuring consistent and safe power output of HIFU transducers. However, the established acoustic field measurement techniques typically use equipment that cannot be used in a magnetic resonance imaging (MRI) suite, thus posing a challenge to the development and execution of HIFU acoustic field characterization techniques. In this study, we developed and characterized a technique for HIFU acoustic field calibration within the MRI magnet bore, and validated the technique with standard hydrophone measurements outside of the MRI suite. METHODS: A clinical Philips MR-HIFU system (Sonalleve V2, Philips, Vantaa, Finland) was used to assess the proposed technique. A fiber-optic hydrophone with a long fiber was inserted through a 24-gauge angiocatheter and fixed inside a water tank that was placed on the HIFU patient table above the acoustic window. The long fiber allowed the hydrophone control unit to be placed outside of the magnet room. The location of the fiber tip was traced on MR images, and the HIFU focal point was positioned at the fiber tip using the MR-HIFU therapy planning software. To perform acoustic field mapping inside the magnet, the HIFU focus was positioned relative to the fiber tip using an MRI-compatible 5-axis robotic transducer positioning system embedded in the HIFU patient table. To perform validation measurements of the acoustic fields, the HIFU table was moved out of the MRI suite, and a standard laboratory hydrophone measurement setup was used to perform acoustic field measurements outside the magnetic field. RESULTS: The pressure field scans along and across the acoustic beam path obtained inside the MRI bore were in good agreement with those obtained outside of the MRI suite. At the HIFU focus with varying nominal acoustic powers of 10-500 W, the peak positive pressure and peak negative pressure measured inside the magnet bore were 3.87-68.67 MPa and 3.56-12.06 MPa, respectively, while outside the MRI suite the corresponding pressures were 3.27-67.32 MPa and 3.06-12.39 MPa, respectively. There was no statistically significant difference (P > 0.05) between measurements inside the magnet bore and outside the MRI suite for the p+ and p- at any acoustic power level. The spatial-peak pulse-average intensities (ISPPA ) for these powers were 312-17816 W/cm2 and 220-15698 W/cm2 for measurements inside and outside the magnet room, respectively. In addition, when the scanning step size of the HIFU focus was increased from 100 µm to 500 µm, the execution time for scanning a 4 × 4 mm2 area decreased from 210 min to 10 min, the peak positive pressure decreased by 14%, the peak negative pressure decreased by 5%, and the lateral full width at half maximum dimension of pressure profiles increased from 1.15 mm to 1.55 mm. CONCLUSIONS: The proposed hydrophone measurement technique offers a convenient and reliable method for characterizing the acoustic fields of clinical MR-HIFU systems inside the magnet bore. The technique was validated for use by measurements outside the MRI suite using a standard hydrophone calibration technique. This technique can be a useful tool in MR-HIFU quality assurance and acoustic field assessment.

Boiling histotripsy lesion characterization on a clinical magnetic resonance imaging-guided high intensity focused ultrasound system.

PURPOSE: High intensity focused ultrasound (HIFU) is a non-invasive therapeutic technique that can thermally ablate tumors. Boiling histotripsy (BH) is a HIFU approach that can emulsify tissue in a few milliseconds. Lesion volume and temperature effects for different BH sonication parameters are currently not well characterized. In this work, lesion volume, temperature distribution, and area of lethal thermal dose were characterized for varying BH sonication parameters in tissue-mimicking phantoms (TMP) and demonstrated in ex vivo tissues. METHODS: The following BH sonication parameters were varied using a clinical MR-HIFU system (Sonalleve V2, Philips, Vantaa, Finland): acoustic power, number of cycles/pulse, total sonication time, and pulse repetition frequency (PRF). A 3×3×3 pattern was sonicated inside TMP's and ex vivo tissues. Post sonication, lesion volumes were quantified using 3D ultrasonography and temperature and thermal dose distributions were analyzed offline. Ex vivo tissues were sectioned and stained with H&E post sonication to assess tissue damage. RESULTS: Significant increase in lesion volume was observed while increasing the number of cycles/pulse and PRF. Other sonication parameters had no significant effect on lesion volume. Temperature full width at half maximum at the end of sonication increased significantly with all parameters except total sonication time. Positive correlation was also found between lethal thermal dose and lesion volume for all parameters except number of cycles/pulse. Gross pathology of ex vivo tissues post sonication displayed either completely or partially damaged tissue at the focal region. Surrounding tissues presented sharp boundaries, with little or no structural damage to adjacent critical structures such as bile duct and nerves. CONCLUSION: Our characterization of effects of HIFU sonication parameters on the resulting lesion demonstrates the ability to control lesion morphologic and thermal characteristics with a clinical MR-HIFU system in TMP's and ex vivo tissues. We demonstrate that this system can produce spatially precise lesions in both phantoms and ex vivo tissues. The results provide guidance on a preliminary set of BH sonication parameters for this system, with a potential to facilitate BH translation to the clinic.

Investigation of polymer-shelled microbubble motions in acoustophoresis.

The objective of this paper is to explore the trajectory motion of microsize (typically smaller than a red blood cell) encapsulated polymer-shelled gas bubbles propelled by radiation force in an acoustic standing-wave field and to compare the corresponding movements of solid polymer microbeads. The experimental setup consists of a microfluidic chip coupled to a piezoelectric crystal (PZT) with a resonance frequency of about 2.8MHz. The microfluidic channel consists of a rectangular chamber with a width, w, corresponding to one wavelength of the ultrasound standing wave. It creates one full wave ultrasound of a standing-wave pattern with two pressure nodes at w/4 and 3w/4 and three antinodes at 0, w/2, and w. The peak-to-peak amplitude of the electrical potential over the PZT was varied between 1 and 10V. The study is limited to no-flow condition. From Gor'kov's potential equation, the acoustic contrast factor, Φ, for the polymer-shelled microbubbles was calculated to about -60.7. Experimental results demonstrate that the polymer-shelled microbubbles are translated and accumulated at the pressure antinode planes. This trajectory motion of polymer-shelled microbubbles toward the pressure antinode plane is similar to what has been described for other acoustic contrast particles with a negative Φ. First, primary radiation forces dragged the polymer-shelled microbubbles into proximity with each other at the pressure antinode planes. Then, primary and secondary radiation forces caused them to quickly aggregate at different spots along the channel. The relocation time for polymer-shelled microbubbles was 40 times shorter than that for polymer microbeads, and in contrast to polymer microbeads, the polymer-shelled microbubbles were actuated even at driving voltages (proportional to radiation forces) as low as 1V. In short, the polymer-shelled microbubbles demonstrate the behavior attributed to the negative acoustic contrast factor particles and thus can be trapped at the antinode plane and thereby separated from particles having a positive acoustic contrast factor, such as for example solid particles and cells. This phenomenon could be utilized in exploring future applications, such as bioassay, bioaffinity, and cell interaction studies in vitro in a well-controlled environment.

Unique pumping-out fracturing mechanism of a polymer-shelled contrast agent: an acoustic characterization and optical visualization.

This work describes the fracturing mechanism of air-filled microbubbles (MBs) encapsulated by a cross-linked poly(vinyl alcohol) (PVA) shell. The radial oscillation and fracturing events following the ultrasound exposure were visualized with an ultrahigh-speed camera, and backscattered timedomain signals were acquired with the acoustic setup specific for harmonic detection. No evidence of gas emerging from defects in the shell with the arrival of the first insonation burst was found. In optical recordings, more than one shell defect was noted, and the gas core was drained without any sign of air extrusion when several consecutive bursts of 1 MPa amplitude were applied. In acoustic tests, the backscattered peak-to-peak voltage gradually reached its maximum and exponentially decreased when the PVA-based MB suspension was exposed to approximately 20 consecutive bursts arriving at pulse repetition frequencies of 100 and 500 Hz. Taking into account that the PVA shell is porous and possibly contains large air pockets between the cross-linked PVA chains, the aforementioned acoustic behavior might be attributed to pumping gas from these pockets in combination with gas release from the core through shell defects. We refer to this fracturing mechanism as pumping-out behavior, and this behavior could have potential use for the local delivery of therapeutic gases, such as nitric oxide.

Assessment of the viscoelastic and oscillation properties of a nano-engineered multimodality contrast agent.

Combinations of microbubbles (MBs) and superparamagnetic iron oxide nanoparticles (SPIONs) are used to fabricate dual contrast agents for ultrasound and MRI. This study examines the viscoelastic and oscillation characteristics of two MB types that are manufactured with SPIONs and either anchored chemically on the surface (MBs-chem) or physically embedded (MBs-phys) into a polymer shell. A linearized Church model was employed to simultaneously fit attenuation coefficients and phase velocity spectra that were acquired experimentally. The model predicted lower viscoelastic modulus values, undamped resonance frequencies and total damping ratios for MBs-chem. MBs-chem had a resonance frequency of approximately 13 MHz and a damping ratio of approximately 0.9; thus, MBs-chem can potentially be used as a conventional ultrasound contrast agent with the combined functionality of MRI detection. In contrast, MBs-phys had a resonance frequency and damping of 28 MHz and 1.2, respectively, and requires further modification of clinically available contrast pulse sequences to be visualized.

On the interplay of shell structure with low- and high-frequency mechanics of multifunctional magnetic microbubbles.

Polymer-shelled magnetic microbubbles have great potential as hybrid contrast agents for ultrasound and magnetic resonance imaging. In this work, we studied US/MRI contrast agents based on air-filled poly(vinyl alcohol)-shelled microbubbles combined with superparamagnetic iron oxide nanoparticles (SPIONs). The SPIONs are integrated either physically or chemically into the polymeric shell of the microbubbles (MBs). As a result, two different designs of a hybrid contrast agent are obtained. With the physical approach, SPIONs are embedded inside the polymeric shell and with the chemical approach SPIONs are covalently linked to the shell surface. The structural design of hybrid probes is important, because it strongly determines the contrast agent's response in the considered imaging methods. In particular, we were interested how structural differences affect the shell's mechanical properties, which play a key role for the MBs' US imaging performance. Therefore, we thoroughly characterized the MBs' geometric features and investigated low-frequency mechanics by using atomic force microscopy (AFM) and high-frequency mechanics by using acoustic tests. Thus, we were able to quantify the impact of the used SPIONs integration method on the shell's elastic modulus, shear modulus and shear viscosity. In summary, the suggested approach contributes to an improved understanding of structure-property relations in US-active hybrid contrast agents and thus provides the basis for their sustainable development and optimization.

Acoustic characterization and contrast imaging of microbubbles encapsulated by polymeric shells coated or filled with magnetic nanoparticles.

The combination of superparamagnetic iron oxide nanoparticles with polymeric air-filled microbubbles is used to produce two types of multimodal contrast agents to enhance medical ultrasound and magnetic resonance imaging. The nanoparticles are either covalently linked to the shell or physically entrapped into the shell. In this paper, the characterization of the acoustic properties (backscattered power, fracturing pressure, attenuation and dispersion of the ultrasonic wave) and ultrasound imaging of the two types of magnetic microbubbles are presented. In vitro B-mode images are generated using a medical ultrasound scanner by applying a nonconventional signal processing technique that is suitable to detect polymeric bubbles and based on the combination of multipulse excitation and chirp coding. Even if both types of microbubbles can be considered to be effective ultrasound contrast agents, the different structure of the shell loaded with nanoparticles has a pronounced effect on the echogenicity and the detection sensitivity of the imaging technique. The best results are obtained using microbubbles that are externally coated with nanoparticles. A backscattered power of 20 dB was achieved at lower concentration, and an increment of 8 dB in the contrast-to-tissue ratio was observed with respect to the more rigid microbubbles with particles entrapped into the shell.

Magnetite nanoparticles can be coupled to microbubbles to support multimodal imaging.

Microbubbles (MBs) are commonly used as injectable ultrasound contrast agent (UCA) in modern ultrasonography. Polymer-shelled UCAs present additional potentialities with respect to marketed lipid-shelled UCAs. They are more robust; that is, they have longer shelf and circulation life, and surface modifications are quite easily accomplished to obtain enhanced targeting and local drug delivery. The next generation of UCAs will be required to support not only ultrasound-based imaging methods but also other complementary diagnostic approaches such as magnetic resonance imaging or computer tomography. This work addresses the features of MBs that could function as contrast agents for both ultrasound and magnetic resonance imaging. The results indicate that the introduction of iron oxide nanoparticles (SPIONs) in the poly(vinyl alcohol) shell or on the external surface of the MBs does not greatly decrease the echogenicity of the host MBs compared with the unmodified one. The presence of SPIONs provides enough magnetic susceptibility to the MBs to accomplish good detectability both in vitro and in vivo. The distribution of SPIONs on the shell and their aggregation state seem to be key factors for the optimization of the transverse relaxation rate.