Browsing by Subject "Magnetic resonance imaging (MRI)"

Now showing 1 - 16 of 16

Open Access
Absolute temperature monitoring using RF radiometry in the MRI scanner
(Institute of Electrical and Electronics Engineers, 2006) El-Sharkawy, A.-M. M.; Sotiriadis, P. P.; Bottomley, P. A.; Atalar, Ergin
Temperature detection using microwave radiometry has proven value for noninvasively measuring the absolute temperature of tissues inside the body. However, current clinical radiometers operate in the gigahertz range, which limits their depth of penetration. We have designed and built a noninvasive radiometer which operates at radio frequencies (64 MHz) with ∼100-kHz bandwidth, using an external RF loop coil as a thermal detector. The core of the radiometer is an accurate impedance measurement and automatic matching circuit of 0.05 Ω accuracy to compensate for any load variations. The radiometer permits temperature measurements with accuracy of ±0.1°K, over a tested physiological range of 28°C-40 °C in saline phantoms whose electric properties match those of tissue. Because 1.5 T magnetic resonance imaging (MRI) scanners also operate at 64 MHz, we demonstrate the feasibility of integrating our radiometer with an MRI scanner to monitor RF power deposition and temperature dosimetry, obtaining coarse, spatially resolved, absolute thermal maps in the physiological range. We conclude that RF radiometry offers promise as a direct, noninvasive method of monitoring tissue heating during MRI studies and thereby providing an independent means of verifying patient-safe operation. Other potential applications include titration of hyper- and hypo-therapies.
Open Access
Accelerated phase-cycled SSFP imaging with compressed sensing
(Institute of Electrical and Electronics Engineers Inc., 2015) Çukur, T.
Balanced steady-state free precession (SSFP) imaging suffers from irrecoverable signal losses, known as banding artifacts, in regions of large B0 field inhomogeneity. A common solution is to acquire multiple phase-cycled images each with a different frequency sensitivity, such that the location of banding artifacts are shifted in space. These images are then combined to alleviate signal loss across the entire field-of-view. Although high levels of artifact suppression are viable using a large number of images, this is a time costly process that limits clinical utility. Here, we propose to accelerate individual acquisitions such that the overall scan time is equal to that of a single SSFP acquisition. Aliasing artifacts and noise are minimized by using a variable-density random sampling pattern in k-space, and by generating disjoint sampling patterns for separate acquisitions. A sparsity-enforcing method is then used for image reconstruction. Demonstrations on realistic brain phantom images, and in vivo brain and knee images are provided. In all cases, the proposed technique enables robust SSFP imaging in the presence of field inhomogeneities without prolonging scan times. © 2014 IEEE.
Open Access
Constrained ellipse fitting for efficient parameter mapping with phase-cycled bSSFP MRI
(IEEE, 2021-08-05) Keskin, K.; Yılmaz, Uğur; Çukur, Tolga
Balanced steady-state free precession (bSSFP) imaging enables high scan efficiency in MRI, but differs from conventional sequences in terms of elevated sensitivity to main field inhomogeneity and nonstandard T2/T1 -weighted tissue contrast. To address these limitations, multiple bSSFP images of the same anatomy are commonly acquired with a set of different RF phase-cycling increments. Joint processing of phase-cycled acquisitions serves to mitigate sensitivity to field inhomogeneity. Recently phase-cycled bSSFP acquisitions were also leveraged to estimate relaxation parameters based on explicit signal models. While effective, these model-based methods often involve a large number of acquisitions (N ≈ 10-16), degrading scan efficiency. Here, we propose a new constrained ellipse fitting method (CELF) for parameter estimation with improved efficiency and accuracy in phase-cycled bSSFP MRI. CELF is based on the elliptical signal model framework for complex bSSFP signals; and it introduces geometrical constraints on ellipse properties to improve estimation efficiency, and dictionary-based identification to improve estimation accuracy. CELF generates maps of T1 , T2 , off-resonance and on-resonant bSSFP signal by employing a separate B1 map to mitigate sensitivity to flip angle variations. Our results indicate that CELF can produce accurate off-resonance and banding-free bSSFP maps with as few as N = 4 acquisitions, while estimation accuracy for relaxation parameters is notably limited by biases from microstructural sensitivity of bSSFP imaging.
Open Access
Efficient parameter mapping for magnetic resonance imaging
(2019-07) Keskin, Kübra
Balanced steady-state free precession (bSSFP) is a magnetic resonance imaging (MRI) sequence that enables high signal-to-noise ratios in short scan times. However, it has elevated sensitivity to main eld inhomogeneity, which leads to banding artifacts near regions of relatively large o -resonance shifts. To suppress these artifacts, multiple bSSFP images of the same anatomy are commonly acquired with a set of di erent RF phase-cycling increments. Joint processing of phase-cycled acquisitions has long been employed to eliminate banding artifacts due to eld inhomogeneity. Multiple bSSFP acquisitions can be further used for parameter mapping by exploiting the signal model of phase-cycled bSSFP. While model based approaches for mapping are e ective, they often need a large number of acquisitions, inherently limiting scan e ciency. In this thesis, we propose a new method for parameter mapping with improved e ciency and accuracy in phasecycled bSSFP MRI. The proposed method is based on the elliptical signal model framework for complex bSSFP signals; and introduces an observation about the signal's geometry to the constrained parameter mapping problem, such that the number of unknowns and thereby the required number of acquisitions can be reduced. It also leverages dictionary-based identi cation to improve estimation accuracy. Simulated, phantom and in vivo experiments demonstrate that the proposed method enables improved parameter mapping with fewer acquisitions.
Open Access
Implantable SUB-cm wireless resonators For MRI: from circuit theory to medical imaging
(2017-12) Gökyar, Sayım
Making implantable wireless resonators having small footprints is fundamentally challenging when using conventional designs that are subject to the inherent tradeo between their size and the achievable range of quality-factors (Q-factors). For clinical magnetic resonance imaging (MRI) frequencies (e.g., about 127 MHz for 3 T), conventional resonators either require a diameter of about 20 cm in chip size or o -the-chip lumped elements for successful operation, both of which practically prevent their use as implantable devices. At least two orders-of-magnitude reduction in footprint area is necessary to make on-chip resonators suitable for invivo applications. However, decreasing the size of such a conventional resonator chip comes at the expense of substantially decreased Q-factor. Thus, achieving high Q-factors with reduced footprints simultaneously entails a novel approach in implantable electronics. In this thesis work, to address this problem, we proposed, designed and demonstrated a new class of sub-wavelength, thin- lm loaded helical metamaterial structures for in-vivo applications including eld localization and signal-to-noise ratio (SNR) improvement in MRI. This implantable wireless architecture, implemented fully on chip with partially overlaid helicals on both sides of the chip interconnected by a through-chip-via, enables a wide range of resonant radio frequencies tunable on chip by design while achieving an extraordinarily small footprint area (<< 1 cm2) and ultra-thin geometry (< 30 m). The miniaturization of such microwave circuits to sub-cm range, together with their high Q-factors exceeding 30 in lossy soft tissues, allows for their use in vivo. The fabricated devices correspond to 1/1500th of their operating wavelength in size, rendering them deep sub-wavelength.For the proposed wireless resonant devices, equivalent circuit models were developed to understand their miniaturization property and the resulting high Q-factors are well explained by using these models. Additionally, full-wave numerical solutions of the proposed geometries were systematically carried out to verify the ndings of the developed equivalent circuit models. All of these theoretical and numerical studies were found in excellent agreement with the experimental RF characterization of the microfabricated devices. Retrieval analyses of the proposed architectures showed that these geometries lead to both negative relative permittivity and permeability simultaneously at their operating frequencies, which do not naturally exist together in nature, making these structures true metamaterials. These fabricated wireless devices were further shown to be promising for the in-vivo application of subdural electrode marking, along with SNR improvement and eld localization without causing excessive heating in MRI. MR images support that the proposed circuitry is also suitable for MRI marking of implants, high-resolution MR imaging and electric eld con nement for lossy medium. Although our demonstrations were for the purpose of marking subdural electrodes, RF characterization results suggest that the proposed device is not limited to MRI applications. Utilizing the same class of structures enabling strong eld localization, numerous wireless applications seem feasible, especially where miniaturization of the wireless devices is required and/or improving the performance of conventional structures is necessary. The ndings of this thesis indicate that the proposed implantable sub-cm wireless resonators will open up new possibilities for the next-generation implants and wireless sensing systems.
Open Access
Innovative designs of RF transmit array coils and RF heating analysis of patients with implanted DBS
(2020-09) Kazemivalipour, Ehsan
A safe and efficient magnetic resonance imaging (MRI) test would rely on informed specification, design, implementation, assessment, and application of appropriately selected radiofrequency (RF) coils. Towards these goals, this dissertation comprises three contributions to novel RF transmit array (TxArray) coil design techniques and two contributions to RF heating reduction of deep brain stimulation (DBS) implants in RF transmit coils. TxArray coils with multiple transmit elements provide the additional degrees of freedom that can be used to enhance field uniformity, accelerate acquisition time, enable RF shimming while intending to mitigate specific absorption rate (SAR) hotspots, and increase power efficiency. How a TxArray coil is designed can have a significant impact on its gain from parallel transmission technology. Thus, the first contribution of this dissertation is on the eigenmode analysis of the scattering matrix for the design of TxArray coils to obtain their efficient operation modes and achieve an efficient RF shimming in terms of power consumption. The algorithm is tested for the design of four 3T TxArray coils with 8 to 32 channels, and it is shown that it can enlarge the dimension of the excitation space by up to 50% compared with the commonly used design techniques. The next contribution is to establish a fast finetuning procedure to precisely design an imperfectly manufactured TxArray coil using its corresponding equivalent circuit model. By fitting the measured scattering parameters to a lumped circuit model, all inductances/resistances of an 8-channel 3T TxArray coil are estimated. The manufactured coil is then appropriately tuned only in a single iteration. As another contribution, a theoretical coil size optimization is introduced to minimize the magnetic coupling between non-adjacent transmit channels of a TxArray coil. By calculating all self/mutual-inductances of a 12-channel 3T TxArray coil and minimizing mutual-inductances, its sizes are determined. The finite element simulations are performed to demonstrate the feasibility of this approach. One of the safety considerations of RF transmit coils is the localized SAR amplification due to the interaction of metallic implants with the coil's electric fields. Therefore, the fourth contribution is on evaluating SAR mitigation performance at tips of patient-derived realistic DBS implants using a 3T patient-adjustable transmit coil, which uses a mechanically rotating linearly polarized birdcage resonator. The reconfigurable coil system decreases the SAR on average by 83% for unilateral leads and by 59% for bilateral leads in comparison to a conventional quadrature birdcage coil. In the final study, the local SAR amplification surrounding the tips of a large cohort of DBS implants with realistic lead trajectories in a commercially available vertical open-bore 1.2T coil and a standard horizontal closed-bore 1.5T birdcage coil is presented. On average, SAR is decreased by 31-fold in the 1.2T vertical coil compared to the 1.5T horizontal coil. Overall, this dissertation proposes innovative approaches for designing TxArray coils and heating assessment and SAR reduction of DBS patients, which mainly contribute to improving the performance of RF transmit coils in terms of power efficiency and patient safety.
Open Access
Innovative hybrid composite nanomaterials
(2016-09) Erdem, Zeliha Soran
Digital lighting and bio-imaging are two emerging crucial research fields. Nanotechnology stands in the center of these applications by providing nano-scale particles possessing large surface-to-volume ratios, high effciency, and low toxicity while allowing for functionalization, effcient quality lighting and improved biocompatible bio-imaging. Some of the frequently employed nanoparticles in optoelectronics and imaging are colloidal semiconductor quantum dots, colloidal conjugated polymer nanoparticles, and colloidal iron oxide nanoparticles, all of which we have studied using colloidal approaches to make hybrid composites for lighting and imaging in this thesis. Fluorescent inorganic nanoparticles of colloidal quantum dots (QDs) attract significant interest for many optoelectronic and biomedical applications. Although they possess numerous advantages including broad absorption band, high quantum yield, and narrow emission spectrum, there are serious concerns on their recycling due to their cadmium-based composition. Alternatively, relatively low toxic organic uorescent polymer nanoparticles or oligomer nanoparticles have stepped forward. However, their reduced emission effciency and stability in solid state is an important limitation for their use in wide-spread solid-state lighting applications. To address these problems, in the first part of this thesis, we proposed and demonstrated the design of new hybrid composite material systems of oligomer nanoparticles to be used in solid-state lighting. We first showed that the emission effciency and stability of the oligomer nanoparticles in solid state are significantly improved based on our proposed crystallization technique. Here, using this simple and low-cost approach, oligomer nanoparticle monoliths were obtained from the powders of these crystals. Despite the disadvantages of using QDs, their high quantum effciency and narrow-band emission still make them a valuable asset for solid-state lighting. However, the decrease in solid-film effciencies is still an important issue to be addressed. With this perspective, in this thesis we utilized the incorporation of QDs into crystalline matrices allowing for the nonradiative energy transfer (NRET) to improve the emission capability of the nano-emitters. Since it is an interesting crystalline semiconductor organic molecule, we employed anthracene as the host donor medium and incorporated the quantum dots being exciton acceptors. Here, we systematically investigated the NRET from each anthracene emission peak to QDs and demonstrated the use of this composite system on LEDs as color converters and the polarization ratio change of quantum dots within this crystal system. Magnetic resonance imaging (MRI), for which we also developed colloidal contrast agents using nanoparticles (NPs) as the second part of this thesis, is a powerful diagnostic tool providing good soft tissue contrast and high spatial resolution. It produces T1- and T2-weighted images, in which the region of interest is observed as brighter and darker contrast, respectively. Superparamagnetic iron oxide (IO) NPs are an important member of T2-weighted contrast agents possessing low toxicity. However, they suer from poor anatomic details due to their darker contrast. Therefore, combining T1- and T2-weighted features in a single IO NP (dual-modal contrast) is a major step for improving MRI contrast. In order to meet the requirement for dual-modal contrast agents, which possess both T1- and T2-weighted imaging capability, in this thesis we synthesized highly monodisperse superparamagnetic cubic IO NPs. Magnetic characterizations along with in vivo MRI experiments demonstrated that these nanoparticles hold great promise for dual-modal imaging. This increased dual-modal eect without paramagnetic material doping or decreasing the size of nanoparticles smaller than 5 nm directed us to understand the relation of the T1 and T2 relaxations depending on the IO NP size and shape. Here, we showed the presence of intrinsic paramagnetic phase in magnetite IO NPs. Moreover, we demonstrated that this contribution is higher in IO NPs possessing cubic shape compared to the spherical counterparts, which explains the increased dual-modal effect in the monodisperse superparamagnetic nanocubes.
Open Access
Modeling of electrodes and implantable pulse generator cases for the analysis of implant tip heating under MR imaging
(Wiley-Blackwell Publishing, Inc., 2015) Acikel, V.; Uslubas, A.; Atalar, Ergin
Purpose: The authors purpose is to model the case of an implantable pulse generator (IPG) and the electrode of an active implantable medical device using lumped circuit elements in order to analyze their effect on radio frequency induced tissue heating problem during a magnetic resonance imaging (MRI) examination. Methods: In this study, IPG case and electrode are modeled with a voltage source and impedance. Values of these parameters are found using the modified transmission line method (MoTLiM) and the method of moments (MoM) simulations. Once the parameter values of an electrode/IPG case model are determined, they can be connected to any lead, and tip heating can be analyzed. To validate these models, both MoM simulations and MR experiments were used. The induced currents on the leads with the IPG case or electrode connections were solved using the proposed models and the MoTLiM. These results were compared with the MoM simulations. In addition, an electrode was connected to a lead via an inductor. The dissipated power on the electrode was calculated using the MoTLiM by changing the inductance and the results were compared with the specific absorption rate results that were obtained using MoM. Then, MRI experiments were conducted to test the IPG case and the electrode models. To test the IPG case, a bare lead was connected to the case and placed inside a uniform phantom. During a MRI scan, the temperature rise at the lead was measured by changing the lead length. The power at the lead tip for the same scenario was also calculated using the IPG case model and MoTLiM. Then, an electrode was connected to a lead via an inductor and placed inside a uniform phantom. During a MRI scan, the temperature rise at the electrode was measured by changing the inductance and compared with the dissipated power on the electrode resistance. Results: The induced currents on leads with the IPG case or electrode connection were solved for using the combination of the MoTLiM and the proposed lumped circuit models. These results were compared with those from the MoM simulations. The mean square error was less than 9%. During the MRI experiments, when the IPG case was introduced, the resonance lengths were calculated to have an error less than 13%. Also the change in tip temperature rise at resonance lengths was predicted with less than 4% error. For the electrode experiments, the value of the matching impedance was predicted with an error less than 1%. Conclusions: Electrical models for the IPG case and electrode are suggested, and the method is proposed to determine the parameter values. The concept of matching of the electrode to the lead is clarified using the defined electrode impedance and the lead Thevenin impedance. The effect of the IPG case and electrode on tip heating can be predicted using the proposed theory. With these models, understanding the tissue heating due to the implants becomes easier. Also, these models are beneficial for implant safety testers and designers. Using these models, worst case conditions can be determined and the corresponding implant test experiments can be planned.
Open Access
mustGAN: multi-stream generative adversarial networks for MR image synthesis
(Elsevier BV, 2021-05) Yurt, Mahmut; Dar, Salman Uh; Erdem, A.; Erdem, E.; Oğuz, Kader K.; Çukur, Tolga
Multi-contrast MRI protocols increase the level of morphological information available for diagnosis. Yet, the number and quality of contrasts are limited in practice by various factors including scan time and patient motion. Synthesis of missing or corrupted contrasts from other high-quality ones can alleviate this limitation. When a single target contrast is of interest, common approaches for multi-contrast MRI involve either one-to-one or many-to-one synthesis methods depending on their input. One-to-one methods take as input a single source contrast, and they learn a latent representation sensitive to unique features of the source. Meanwhile, many-to-one methods receive multiple distinct sources, and they learn a shared latent representation more sensitive to common features across sources. For enhanced image synthesis, we propose a multi-stream approach that aggregates information across multiple source images via a mixture of multiple one-to-one streams and a joint many-to-one stream. The complementary feature maps generated in the one-to-one streams and the shared feature maps generated in the many-to-one stream are combined with a fusion block. The location of the fusion block is adaptively modified to maximize task-specific performance. Quantitative and radiological assessments on T1,- T2-, PD-weighted, and FLAIR images clearly demonstrate the superior performance of the proposed method compared to previous state-of-the-art one-to-one and many-to-one methods.
Open Access
Novel implantable distributively loaded flexible resonators for MRI
(2011) Gökyar, Sayım
Magnetic resonance imaging (MRI) is an enabling technology platform for imaging applications. In MRI, the imaging frequency falls within the radio frequency (RF) range where the tissue absorption of electromagnetic power is conveniently very low (e.g., compared to X-ray imaging), making MRI medically safe. As a result, MRI has evolved into a major imaging tool in medicine. However, in MRI, it is typically difficult to receive a magnetic resonance signal from tissue near a metallic implant, which hinders imaging of the implant device neighborhood to observe, monitor, and make assessment of the recovery and tissue compatibility. This can be accomplished by using locally resonating implants, but such implantable local resonators compatible with MRI that simultaneously feature reasonable chip size are currently not available (although there are some MRI-guided catheter applications). In this thesis, we proposed and developed a new class of implantable chip-scale local resonators that operate at radio frequencies of MRI, despite their small size, for the purposes of enhancing the signal-to-noise ratio (SNR) and thus the resolution in their vicinity. Here we addressed the scientific challenge of achieving low resonance frequency while maintaining chip-scale size suitable for potential MR-compatible implants. Using only biocompatible materials (gold, nitrides, and silicon or polyimide) within a substantially reduced footprint (miniaturized by 2 orders of magnitude), we demonstrated novel chip-scale designs based on the basic concept of split ring resonators (SRRs). Different than classical SRRs or those loaded with lumped elements (e.g., thin-film lumped loading), however, in our designs we loaded the SRR geometry in a distributive manner with a micro-fabricated dielectric thin-film layer to increase effective capacitance. For a proof-of-concept demonstration, we fabricated 20 mm ´ 20 mm resonators that operate at the resonance frequency of 130 MHz (compatible with 3 T MRI system) when distributively loaded with the capacitive film, which would otherwise operate around 1.2 GHz as a classical SRR of the same size if not loaded. It is worth noting that this resonance frequency of 130 MHz would normally require a classical SRR of 20 cm ´ 20 cm, a chip size 100-fold larger than ours. Designing and fabricating flexible thin-film resonators, we also showed that this architecture can be tuned by bending and is appropriate for non-planar surfaces, which is often the case for in vivo imaging. The phantom images indicated that, depending on the resonator configuration, these novel self-resonating structures increase SNR of the received signal by a maximum factor of 4 to 150 and over an enhancement penetration up to 10 mm into the phantom. This corresponds to a resolution enhancement in the 2D image by a factor of 2 to 12, respectively, under the same RF power. These in vitro experiments prove that it is possible to operate our local resonators at reduced frequencies via the help of distributive loading on the same chip. These findings suggest that proposed implantable resonator chips make promising candidates for self-resonating MR-compatible implants.
Open Access
Optimization and machine learning in MRI: applications in rapid MR image reconstruction and encoding models of cortical representations
(2020-02) Shahdloo, Mohammad
Magnetic Resonance Imaging (MRI) is a non-invasive medical imaging modality that is widely used by clinicians and researchers to picture body anatomy and neuronal function. However, long scan time remains a major problem. Recently, multiple techniques have emerged that reduce the acquired MRI signal samples, hence dramatically accelerating the acquisition. These techniques involve sophisticated signal reconstruction procedures that in essence require solving regularized optimization problems, and clinical adoption of accelerated MRI critically relies on self-tuning solutions for these problems. Further to this, recent experimental approaches in cognitive neuroscience favor employing naturalistic audio-visual stimuli that closely resemble humans’ daily-life experience. Yet, these modern paradigms inevitably lead to huge functional MRI (fMRI) datasets that require advanced statistical and computational techniques to uncover the large amount of embedded information. Here, we propose a novel eﬃcient datadriven self-tuning reconstruction method for accelerated MRI. We demonstrate superior performance of the proposed method across various simulated and in vivo datasets and under various scan conﬁgurations. Furthermore, we develop statistical analysis tools to investigate the neural representation of hundreds of action categories in natural movies in the brain via fMRI, and study their attentional modulations. Finally, we develop a model-based framework to estimate temporal extent of semantic information integration in the brain, and investigate its attentional modulations using fMRI data recorded during natural story listening. In short, the methodological and analytical approaches introduced in this thesis greatly beneﬁt clinical utility of accelerated MRI, and enhance our understanding of brain function in daily life.
Open Access
Patient’s body composition can significantly affect RF power deposition in the tissue around DBS implants: ramifications for lead management strategies and MRI field-shaping techniques
(Institute of Physics Publishing Ltd., 2021-01-13) Bhusal, B.; Keil, B.; Rosenow, J.; Kazemivalipour, Ehsan; Golestanirad, L.
Patients with active implants such as deep brain stimulation (DBS) devices have limited access to magnetic resonance imaging (MRI) due to risks associated with RF heating of implants in MRI environment. With an aging population and increased prevalence of neurodegenerative disease, the indication for MRI exams in patients with such implants increases as well. In response to this growing need, many groups have investigated strategies to mitigate RF heating of DBS implants during MRI. These efforts fall into two main categories: MRI field-shaping methods, where the electric field of the MRI RF coil is modified to reduce the interaction with implanted leads, and lead management techniques where surgical modifications in the trajectory reduces the coupling with RF fields. Studies that characterize these techniques, however, have relied either on simulations with homogenous body models, or experiments with box-shaped single-material phantoms. It is well established, however, that the shape and heterogeneity of human body affects the distribution of RF electric fields, which by proxy, alters the heating of an implant inside the body. In this contribution, we applied numerical simulations and phantom experiments to examine the degree to which variations in patient's body composition affects RF power deposition. We then assessed effectiveness of RF-heating mitigation strategies under variant patient body compositions. Our results demonstrated that patient's body composition substantially alters RF power deposition in the tissue around implanted leads. However, both techniques based on MRI field-shaping and DBS lead management performed well under variant body types.
Open Access
Prior-Guided image reconstruction for accelerated multi-contrast MRI via generative adversarial networks
(IEEE, 2020) Dar, Salman U.H.; Yurt, Mahmut; Shahdloo, Mohammad; Ildız, Muhammed Emrullah; Tınaz, Berk; Çukur, Tolga
Multi-contrast MRI acquisitions of an anatomy enrich the magnitude of information available for diagnosis. Yet, excessive scan times associated with additional contrasts may be a limiting factor. Two mainstream frameworks for enhanced scan efficiency are reconstruction of undersampled acquisitions and synthesis of missing acquisitions. Recently, deep learning methods have enabled significant performance improvements in both frameworks. Yet, reconstruction performance decreases towards higher acceleration factors with diminished sampling density at high-spatial-frequencies, whereas synthesis can manifest artefactual sensitivity or insensitivity to image features due to the absence of data samples from the target contrast. In this article, we propose a new approach for synergistic recovery of undersampled multi-contrast acquisitions based on conditional generative adversarial networks. The proposed method mitigates the limitations of pure learning-based reconstruction or synthesis by utilizing three priors: shared high-frequency prior available in the source contrast to preserve high-spatial-frequency details, low-frequency prior available in the undersampled target contrast to prevent feature leakage/loss, and perceptual prior to improve recovery of high-level features. Demonstrations on brain MRI datasets from healthy subjects and patients indicate the superior performance of the proposed method compared to pure reconstruction and synthesis methods. The proposed method can help improve the quality and scan efficiency of multi-contrast MRI exams.
Open Access
Reconfigurable MRI technology for low-SAR imaging of deep brain stimulation at 3T: application in bilateral leads, fully-implanted systems, and surgically modified lead trajectories
(Elsevier, 2019) Kazemivalipour, Ehsan; Keil, B.; Vali, A.; Rajan, S.; Elahi, B.; Atalar, Ergin; Wald, L.; Rosenow, J.; Pilitsis, J.; Golestanirad, L.
Patients with deep brain stimulation devices highly benefit from postoperative MRI exams, however MRI is not readily accessible to these patients due to safety risks associated with RF heating of the implants. Recently we introduced a patient-adjustable reconfigurable coil technology that substantially reduced local SAR at tips of single isolated DBS leads during MRI at 1.5 T in 9 realistic patient models. This contribution extends our work to higher fields by demonstrating the feasibility of scaling the technology to 3T and assessing its performance in patients with bilateral leads as well as fully implanted systems. We developed patient-derived models of bilateral DBS leads and fully implanted DBS systems from postoperative CT images of 13 patients and performed finite element simulations to calculate SAR amplification at electrode contacts during MRI with a reconfigurable rotating coil at 3T. Compared to a conventional quadrature body coil, the reconfigurable coil system reduced the SAR on average by 83% for unilateral leads and by 59% for bilateral leads. A simple surgical modification in trajectory of implanted leads was demonstrated to increase the SAR reduction efficiency of the rotating coil to >90% in a patient with a fully implanted bilateral DBS system. Thermal analysis of temperature-rise around electrode contacts during typical brain exams showed a 15-fold heating reduction using the rotating coil, generating <1C temperature rise during ∼4-min imaging with high-SAR sequences where a conventional CP coil generated >10C temperature rise in the tissue for the same flip angle.
Open Access
A simple analytical expression for the gradient induced potential on active implants during MRI
(2012) Turk, E.A.; Kopanoglu, E.; Guney, S.; Bugdayci, K.E.; Ider, Y. Z.; Erturk, V. B.; Atalar, Ergin
During magnetic resonance imaging, there is an interaction between the time-varying magnetic fields and the active implantable medical devices (AIMD). In this study, in order to express the nature of this interaction, simplified analytical expressions for the electric fields induced by time-varying magnetic fields are derived inside a homogeneous cylindrical volume. With these analytical expressions, the gradient induced potential on the electrodes of the AIMD can be approximately calculated if the position of the lead inside the body is known. By utilizing the fact that gradient coils produce linear magnetic field in a volume of interest, the simplified closed form electric field expressions are defined. Using these simplified expressions, the induced potential on an implant electrode has been computed approximately for various lead positions on a cylindrical phantom and verified by comparing with the measured potentials for these sample conditions. In addition, the validity of the method was tested with isolated frog leg stimulation experiments. As a result, these simplified expressions may help in assessing the gradient-induced stimulation risk to the patients with implants.
Open Access
Skip connections for medical image synthesis with generative adversarial networks
(IEEE, 2022-08-29) Mirza, Muhammad Usama; Dalmaz, Onat; Çukur, Tolga
Magnetic Resonance Imaging (MRI) is an imaging technique used to produce detailed anatomical images. Acquiring multiple contrast MRI images requires long scan times forcing the patient to remain still. Scan times can be reduced by synthesising unacquired contrasts from acquired contrasts. In recent years, deep generative adversarial networks have been used to synthesise contrasts using one-to-one mapping. Deeper networks can solve more complex functions, however, their performance can decline due to problems such as overfitting and vanishing gradients. In this study, we propose adding skip connections to generative models to overcome the decline in performance with increasing complexity. This will allow the network to bypass unnecessary parameters in the model. Our results show an increase in performance in one-to-one image synthesis by integrating skip connections.