Keywords: MR-Guided Interventions, Segmentation

Motivation: Localization of liver, liver vessels, and interventional needle on 3D magnetic resonance imaging (MRI) provides essential information for MR-guided interventions.

Goal(s): To develop a multi-class network for segmenting the three classes on intra-procedural 3D MRI.

Approach: 3D Swin UNEt Transformer (UNETR) with pre-trained model weights was trained with data augmentation. Needle localization was performed based on the predicted needle segmentation.

Results: In six-fold cross validation of 42 3D images, the multi-class model achieved median Dice scores of 0.87, 0.64, 0.76 for liver, liver vessels and needle. The needle tip localization showed improvements compared to a single-class 3D Swin UNETR model.

Impact: We trained the 3D Swin UNETR for 3D liver, liver vessel, and interventional needle segmentation on intra-procedural 3D MRI and showed that the needle localization performance can be improved using multi-class model compared to single-class model for needle localization.

Keywords: MR-Guided Interventions, MR-Guided Interventions, 0.55T

Motivation: Numerous studies have investigated the appearance of MRI-compatible needles for interventional MRI. Needle artifacts were evaluated manually, which is a time-consuming process and prone to inter-reader deviations. 

Goal(s): Our goal was to introduce an automated method for evaluating needle artifact diameter, aiming to simplify artifact assessment.

Approach: An automatic evaluation of needle artifacts was implemented and validated by conducting a systematic investigation of needle artifacts in a phantom study.

Results: A comprehensive chart of needle artifact diameter for various image parameters was generated automatically, serving as a tool for needle comparison. The results indicate excellent algorithm reliability. 

Impact: The proposed tool allows for a systematic characterization of needle artifact size with high reliability and efficiency, which could be used to guide the correct choice of needle in different use cases (balance between artifact and lesion visibility)

Keywords: Interventional Devices, Interventional Devices, Pulse sequence design

Motivation: MR-guided percutaneous interventions at closed-bore high-field systems profit from remote device manipulation with real-time needle tracking.

Goal(s): To integrate a sequence capable of acquiring two T2-weighted orthogonal slices simultaneously (Ortho-SSFP-Echo) into the device-assisted needle intervention workflow during needle insertion.

Approach: The assistance system (GantryMate) was coupled with a real-time POCC sequence to target a lesion. Once the lesion was identified, its position information was used in a real-time Ortho-SSFP-Echo acquisition to dynamically monitor the needle insertion.

Results: The Ortho-SSFP-Echo sequence enables simultaneous needle visualization in two planes, offering higher CNR than GRE and comparable to bSSFP, without banding artifacts.

Impact: Combining an Ortho-SSFP-Echo sequence with a compact needle assistance system provides a streamlined interventional workflow with decreased complexity and improved image contrast for MR-guided percutaneous interventions at high field.

Keywords: Interventional Devices, Interventional Devices, Arterial Spin Labeling

Motivation: Intra-Arterial-Spin Labeling (iASL) could offer an alternative for myocardial perfusion measurements without the use of contrast agents by labeling blood with a catheter transmit RF-coil.

Goal(s): To study the parameters influencing the iASL labeling efficiency inside the artery to improve the SNR.

Approach: The complex iASL process is investigated with simulations and in-vitro measurements for different coil geometries (solenoid, loop), driving currents, coil orientations relative to the magnetic field, and blood flow velocities.

Results: The simulations were in good agreement with measurements, show a maximal mean flip angle created by the labeling pulse of 105°, and a threshold behavior for the transmit current.

Impact: The improvement of the labeling process inside coronary arteries started in this study, is an important step towards making ASL a viable method for myocardial perfusion quantification.

Keywords: MR-Guided Interventions, MR-Guided Interventions

Motivation: Contemporary low-field MRI systems hold great promise for guiding interventions. However, due to inherently reduced polarization at lower field, it is more challenging to achieve high-spatiotemporal-resolution with sufficient SNR in interactive real-time imaging.

Goal(s): To improve real-time interactive imaging for MRI guided interventions at 0.55T by leveraging deep learning image reconstruction. 

Approach: We implemented deep learning image reconstruction for interactive real-time imaging, and compared its performance with conventional parallel imaging reconstruction and compressed-sensing on a biopsy phantom and a healthy volunteer. 

Results: Deep learning image reconstruction allows for accelerated interactive real-time imaging, achieving image quality that compared favorably with conventional reconstructions and compressed-sensing.

Impact: The proposed method has the potential to further empower 0.55T MRI as a viable interventional guidance platform by leveraging deep learning image reconstruction in accelerated interactive real-time imaging. 

Keywords: MR-Guided Interventions, Low-Field MRI, Interventional navigation, portable MRI, EPI,high resolution ,real time imaging

Motivation: Low-field MRI's portability and safety make it appealing for real-time interventional navigation and bedside evaluations, but it lacks efficient navigation strategies.

Goal(s): Our goal was to demonstrate the potential of low-field MRI in navigation by designing efficient and high-resolution echo-planar acquisition sequences.

Approach: We harnessed low-field MRI's long T₂ and short T₁ properties to design short-TR EPI sequences for real-time navigation and multi-shot EPI sequences for bedside postoperative assessments.

Results: With the short-TR 2D EPI sequence, real-time needle insertion images were successfully obtained, and was improved after denoising. In a self-made resolution phantom, the use of multi-shot EPI greatly improved the image resolution.

Impact: The combination of low-field portable EPI sequences and denoising algorithms highlights their time efficiency in navigation, thus expanding the prospects for low-field portable MRI in the field of navigation.

Keywords: MR-Guided Radiotherapy, MR-Guided Interventions, Particle therapy

Motivation: MR-guided particle therapy (MRgPT) could improve tumor control and reduce healthy tissue dose, thereby increasing quality of therapy compared to conventional CT-guided radiotherapy.

Goal(s): Development of a rotatable, radiation-transparent extremity radiofrequency (RF) coil suitable for MRgPT at a 0.25T C-shaped magnet.

Approach: The receive field homogeneity of the RF coil was evaluated for different rotation angles in electromagnetic field simulations, and the SNR was investigated in a homogeneous phantom in comparison to two commercial coils.

Results: The constructed extremity RF coil including a custom-built preamplifier achieves improved receive field homogeneity and SNR compared to two commercial RF coils and is compatible with MRgPT.

Impact: The rotatable, radiation-transparent RF coil enables MRgPT with through-coil irradiation and multi-angle access to the patient. The developed RF coil and preamplifier design provides better RF field homogeneity and SNR over a larger FOV compared to two commercial RF coils.

Keywords: MR-Guided Interventions, RF Arrays & Systems, Transcranial Magnetic Stimulation, Cocurrent TMS-fMRI

Motivation: Concurrent TMS-fMRI lacks whole-brain coverage or has low SNR.

Goal(s): Develop setups allowing for whole-brain TMS-fMRI with high spatiotemporal resolution and minimizing the interference between MRI and TMS during data collection and neuromodulation.

Approach: Combinations of 8- and 16-channel arrays enabled whole-brain TMS-MRI with 16 or 24 channels.

Results: Structural and functional images of high spatiotemporal resolution (1.5 mm with 2 s/volume or 5 mm with 0.1 s/volume) were demonstrated with an MRI-compatible TMS coil.

Impact: We developed concurrent TMS-MRI setups with whole-brain coverage and high spatiotemporal resolution using combinations of 8- and 16-channel coil arrays. Images of high spatiotemporal resolution were demonstrated.

Keywords: Interventional Devices, Interventional Devices

Motivation: Trajectory guides for MR-guided neurosurgeries have time-consuming, complex workflows.  Trajectories are oriented by iterating between imaging and device manipulation. Noniterative approaches could reduce complexity and anesthesia time.

Goal(s): To create and validate a trajectory guide that enables faster, accurate trajectory guidance in minimally-invasive neurosurgeries.

Approach: Using new hardware and software, a single scan approach was used to perform drill guidance and device insertion on phantoms and cadaver heads.

Results: The proposed methodology accurately guided needles to targets within cadaver brains using a single targeting scan. The initial design produced a radial error of 1.54±0.81mm.

Impact: The proposed device and software accelerate trajectory guidance in minimally-invasive neurosurgeries by reducing the number of acquired scans and procedural steps. This approach simplifies and accelerates procedures which minimizes time under anesthesia.

Keywords: Interventional Devices, MR-Guided Interventions, ThermalMR Brain Tumor

Motivation: Thermal magnetic resonance theranostics combines diagnostic MRI with targeted thermal therapy with an integrated radiofrequency applicator. Precise RF dosimetry is crucial for real-time treatment planning. 

Goal(s): Our goal is to evaluate a time-frequency multiplexing wideband RF beamforming method for precise targeting of small and large deep-seated brain tumors for Thermal Magnetic Resonance theranostics. 

Approach: We employed a multi-vector field shaping algorithm for optimizing RF channel settings of the RF applicator. 

Results: With time-frequency multiplex excitations, we achieved precise SAR_10g targeting in the tumor volume while minimizing RF exposure to healthy tissues. Our study advances thermal magnetic resonance theranostics efficacy promising improved outcomes.

Impact: Our approach of utilizing horse-shoe shaped RF applicator comprised of wideband SGBT dipole antenna can be conveniently adapted to individual patient's tumor position and geometry while maintaining the efficiency and quality of RF heating for ThermalMR theranostics of brain tumors.

Keywords: Interventional Devices, Simulations, Brain Tumor, UHF-MRI, Thermal Therapy, Multi-Target Evaluation

Motivation: Thermal Magnetic Resonance (ThermalMR) adds a thermal intervention dimension to MRI by focusing E-fields in the target region (TR). Today’s assessment of RF applicators uses TRs covering a few specific but arbitrary locations in the brain. This approach does not represent the full clinical picture. 

Goal(s): This work introduces a multi-target evaluation (MTE) framework and expands RF applicator evaluation to whole brain coverage. 

Approach: MTE using EMF simulations for a helmet and for an annular RF applicator.

Results: Our MTE approach provides a technical foundation for the development and objective evaluation of RF applicators tailored for ThermalMR.

Impact: MTE provides a technical foundation for objective RF applicator evaluation using whole brain coverage instead of a few specific but arbitrary target locations. Our framework presents a springboard for RF applicator design and for ThermalMR-based therapy of glioblastoma multiforme. 

Keywords: Other Interventional, fMRI (resting state), Brain Entropy, tDCS

Motivation: Specific impact of Transcranial Direct Current Stimulation (tDCS) on the human primary visual cortex (V1) remains unclear. 

Goal(s): We used fMRI-based brain entropy (BEN) method to investigate the effects of short- and long-term tDCS on brain activity, and dynamic changes of BEN during tDCS.

Approach: The resting-state fMRl data were collected before, during and after stimulation from 10 healthy subjects.

Results: We observed reduced BEN values after both short- and long-term tDCS in frontoparietal and occipital areas. During tDCS, the brain preferred to stay in a state with lower BEN values in the default mode network compared to other brain regions.

Impact: Short- and long-term tDCS on V1 both have a positive effect on improving cognitive functions in healthy and psychiatric disorder population. We also further validated the utility of BEN as an effective method for assessing tDCS effects.

Keywords: Interventional Devices, Stroke, flow diverter, 7T MRI

Motivation: Adverse interactions with static magnetic fields (e.g., displacement force and torque) and radiofrequency-induced heating during MRI have been resolved for 3T MR systems, but have yet to be assessed for 7T MRI.

Goal(s): The present study aimed to assess displacement force and torque in the 7T static magnetic field and to clarify radiofrequency-induced heating during 7T MRI for two types of FDs.

Approach: This study was conducted based on the ASTM standards.

Results: Magnetic field interactions and heating on FDs during 7T MRI are acceptable from a safety perspective.

Impact: This study based on ASTM standards showed magnetic field interactions and heating on FDs during 7T MRI are acceptable from a safety perspective.

Keywords: MR-Guided Interventions, Safety

Motivation: Neuromonitoring is critical during interventions near nerves. During MRI-guided interventions, a conventional system using metallic electrode needles has been used clinically, since MRI-conditional neuromonitoring systems are currently unavailable. This introduces elevated RF-heating risk. 

Goal(s): To identify factors that contribute to RF-heating of neuromonitoring electrodes and strategies to mitigate the risk.

Approach: Neuromonitoring electrodes were inserted into porcine tissue and imaged with various clinically relevant MRI sequences for various neuromonitoring equipment configurations with high RF-heating risks. Temperatures were recorded using fiber-optic sensors.

Results: Substantial temperature elevations were observed during MRI, and several RF-heating mitigation strategies were identified to enable neuromonitoring during MRI-guided procedures. 

Impact: The identified electrode RF-heating mitigation strategies can significantly reduce the RF-heating risk during MRI-guided ablations where neuromonitoring is critical. Understanding factors affecting RF-heating can also help identify high-risk procedures and guide risk-benefit analysis.

Keywords: MR-Guided Radiotherapy, Radiotherapy, Deep Learning, MR-only, Radiation Therapy Planning

Motivation: Deep Learning (DL) is an enabling technology for MR-only Radiation Therapy Planning in terms of 1) MR to synthetic CT conversion and 2) automated Organ-At-Risk (OAR) segmentation.  

Goal(s): To investigate the feasibility of DL based MR-only RT planning using accelerated MR protocols with <5mins scan time (not including tumor depiction).

Approach: Quantitative evaluation of synthetic CTs in terms of mean absolute error (MAE) and OAR segmentations in terms of Likert Score analysis.

Results: Preliminary results demonstrate the feasibility of DL-based MR-only RT in <5min scan time with only minor degradation of synthetic CT and OAR segmentation quality. 

Impact: Deep Learning is an enabling technology for MR-only Radiation Therapy Planning.  Here we demonstrate its capabilities for 1) synthetic CT conversion and 2) Organ-At-Risk (OAR) segmentation in <5min MR scan time. 

Keywords: Phantoms, Phantoms, Real-time, adaptive radiotherapy

Motivation: Real-time adaptive radiotherapy, which involves tracking a target during internal patient motion, may soon reach clinical implementation. Quality assurance methods will be needed, especially for deformable motion.

Goal(s): Develop a method of characterizing the relationship between input and target shape for deformable tumor model phantoms.

Approach: We acquired dynamic MRI and static CT of a deformable phantom subject to several input waveforms. We approximated the target as an ellipsoid and created a linear model to predict the ellipsoid parameters from the input.

Results: The ellipsoid parameters of the target subject to a new waveform could be predicted to within a millimeter.

Impact: We demonstrated a simple method to determine the relationship between input and target geometry for deformable tumor model phantoms. Patient-specific tumor motion can be recreated by inverting the input-target geometry relationship for quality assurance of real-time adaptive radiotherapy plans.