MR Image Quality Research Articles

MRI of the pancreas: Problem solving tool

Advances in MR hardware and pulse sequence design over the years have improved the quality and robustness of MR imaging of the pancreas. Today, MRI is an indispensible tool for studying the pancreas and can provide useful information not attainable with other noninvasive or minimally invasive imaging techniques. In the present review, specific cases are reviewed where the strengths of MRI demonstrate the utility of this imaging modality as a problem solving tool.

PET/MRI in head and neck cancer: initial experience

PurposeTo evaluate the feasibility of PET/MRI (positron emission tomography/magnetic resonance imaging) with FDG (18F-fluorodeoxyglucose) for initial staging of head and neck cancer.MethodsThe study group comprised 20 patients (16 men, 4 women) aged between 52 and 81 years (median 64 years) with histologically proven squamous cell carcinoma of the head and neck region. The patients underwent a PET scan on a conventional scanner and a subsequent PET/MRI examination on a whole-body hybrid system. FDG was administered intravenously prior to the conventional PET scan (267–395 MBq FDG, 348 MBq on average). The maximum standardized uptake values (SUVmax) of the tumour and of both cerebellar hemispheres were determined for both PET datasets. The numbers of lymph nodes with increased FDG uptake were compared between the two PET datasets.ResultsNo MRI-induced artefacts where observed in the PET images. The tumour was detected by PET/MRI in 17 of the 20 patients, by PET in 16 and by MRI in 14. The PET/MRI examination yielded significantly higher SUVmax than the conventional PET scanner for both the tumour (p < 0.0001) and the cerebellum (p = 0.0009). The number of lymph nodes with increased FDG uptake detected using the PET dataset from the PET/MRI system was significantly higher the number detected by the stand-alone PET system (64 vs. 39, p = 0.001).ConclusionThe current study demonstrated that PET/MRI of the whole head and neck region is feasible with a whole-body PET/MRI system without impairment of PET or MR image quality.

Effect of radiation induced current on the quality of MR images in an integrated linac‐MR system

In integrated linac-MRI systems, the RF coils are exposed to the linac's pulsed radiation, leading to a measurable radiation induced current (RIC). This work (1) visualizes the RIC in MRI raw data and determines its effect on the MR image signal-to-noise ratio (SNR) (b) examines the effect of linac dose rate on SNR degradations, (c) examines the RIC effect on different MRI sequences, (d) examines the effect of altering the MRI sequence timing on the RIC, and (e) uses a postprocessing method to reduce the RIC signal from the MR raw data. MR images were acquired on the linac-MR prototype system using various imaging sequences (gradient echo, spin echo, and bSSFP), dose rates (0, 50, 100, 150, 200, and 250 MU∕min) and repetition times (TR) with the gradient echo sequence. The images were acquired with the radiation beam either directly incident or blocked from the RF coils. The SNR was calculated for each of these scenarios, showing a loss in SNR due to RIC. Finally, a postprocessing method was applied to the image k-space data in order to remove partially the RIC signal and recover some of the lost SNR. The RIC produces visible spikes in the k-space data acquired with the linac's radiation incident on the RF coils. This RIC leads to a loss in imaging SNR that increases with increasing linac dose rate (15%-18% loss at 250 MU∕min). The SNR loss seen with increasing linac dose rate appears to be largely independent of the MR sequence used. Changing the imaging TR had interesting visual effects on the appearance of RIC in k-space due to the timing between the linac's pulsing and the MR sequence, but did not change the SNR loss for a given linac dose rate. The use of a postprocessing algorithm was able to remove much of the RIC noise spikes from the MR image k-space data, resulting in the recovery of a significant portion, up to 81% (Table II), of the lost image SNR. The presence of RIC in MR RF coils leads to a loss of SNR which is directly related to the linac dose rate. The RIC related loss in SNR is likely to increase for systems that are able to provide larger than 250 MU∕min dose. Some of this SNR loss can be recovered through the use of a postprocessing algorithm, which removes the RIC artefact from the image k-space.

Does the Use of an Immobilizer Provide a Quality MR Image of the Brain in Infants?

Magnetic resonance imaging (MRI) provides key clinical and diagnostic information for care of neonates and young infants. However, to obtain quality images, they need to be motionless in the scanner, often with the aid of sedation medications, which put them at risk for complications. An immobilizer device (an airtight, chambered device that safely and securely fastens around the infant without applying pressure) allows the infant to feel snug and warm, sleepy and motionless, and can be used as an alternative to sedation for this population. We conducted a retrospective, descriptive, comparative study at The Children’s Hospital of Philadelphia MRI Department in the Division of Radiology. Inclusion criteria were infants less than or equal to 90 days of age, weighing at least 2kg, requiring a MRI brain scan that is predicted to take no more than 60min to complete. We obtained data on 36 patients who received sedation medications for their brain MRI scan, and 36 patients who completed their brain MRI scan without the use of sedation medication, but rather were in the “feed and immobilize” group. Results of the study showed that brain MRIs on sedated infants took longer, and those infants were more likely to experience oxygen desaturation and require supplemental oxygen post-MRI. Most importantly, we found that the MR images were considered diagnostic in 100% of the sedated infants and 94% of the immobilized infants, although significantly more immobilized infants had artifact from motion than sedated infants. Implications for practice include potential cost saving and increased patient’s safety (e.g., more stable respiratory status).

Magnetic resonance imaging for the study of ovarian follicles in the mouse

Additional tools to analyze follicle development would be highly advantageous because current methods require sacrifice of animals at specific times and time-consuming sectioning of tissues for histologic analysis. Magnetic resonance imaging (MRI) may provide a less involved, faster and more cost-effective method to analyze follicles in whole ovaries. Fixed ovaries were collected at different stages of the estrus cycle and after stimulation with gonadotrophins (24 and 48 h post pregnant mares serum (PMSG), and 10 and 24 h post human chorionic gonadotrophin (hCG)) with or without administration of the contrast agent gadodiamide. The MR images were generated using a vertical-bore, 11.7 Tesla MR system. Analysis of the MR images revealed large antral follicles in fixed ovaries with the oocyte and cumulus mass identifiable within preovulatory follicles. The use of gadodiamide had no impact on the quality of MR images obtained. The fixed ovaries were paraffin embedded, sectioned, and hematoxylin stained. Follicles were counted using the MR images and the histology sections. Preovulatory follicle numbers determined using MR images were comparable to those using histology; however counts of smaller follicles were inconsistent. MRI of gonadotrophin-stimulated ovaries in situ did not reveal discernable ovarian structures. Therefore, MRI is a useful tool for studying whole fixed ovaries leaving the ovary intact for additional analyses or for selection of samples based on morphology. The MRI is also useful for identifying preovulatory follicles, although analysis of smaller follicles is not possible, and thus the potential exists for cyst analysis in mouse models of polycystic ovarian syndrome (PCOS).

Improved parallel MR imaging using a coefficient penalized regularization for GRAPPA reconstruction

A novel coefficient penalized regularization method for generalized autocalibrating partially parallel acquisitions (GRAPPA) reconstruction is developed for improving MR image quality. In this method, the fitting coefficients of the source data are weighted with different penalty factors, which are highly dependent upon the relative displacements from the source data to the target data in k-space. The imaging data from both phantom testing and in vivo MRI experiments demonstrate that the coefficient penalized regularization method in GRAPPA reconstruction is able to reduce noise amplification to a greater degree. Therefore, the method enhances the quality of images significantly when compared to the previous least squares and Tikhonov regularization methods.

Parallel Imaging of the Cervical Spine at 3T: Optimized Trade-Off between Speed and Image Quality

Patients with cervical spine syndrome often experience pain during the MR examination. Our aim was to compare the quality of cervical spine MR images obtained by parallel imaging with those of nonaccelerated images, with the goal of shortening the examination time while preserving adequate image quality. A phantom study and examinations of 10 volunteers and 26 patients were conducted on a clinical 3T scanner. Acquisitions included axial T2WI, sagittal T2WI, T1WI, and T2TIRM sequences. Nonaccelerated sequences and accelerated sequences with different numbers of averages and different accelerations, with a scanning time reduction of 67%, were performed. For quantitative analysis, the SNR was obtained from the phantom measurements, and the NU was calculated from the volunteer measurements. For qualitative analysis, 3 independent readers assessed the delineation of anatomic structures in volunteers and the visibility of degenerative disease in patients. In the phantom study, as expected, the SNR of the nonaccelerated images was higher than the SNR of the same sequence with parallel imaging. In vivo, the NU was higher when applying fewer averages or parallel imaging, compared with the nonaccelerated images. The analysis of the subjective parameters in the volunteers and patients showed that a scanning time of 48% of the original protocol could be obtained by combining the following sequences: sagittal T1WI with 1 average; sagittal T2WI with acceleration factor 3; sagittal T2TIRM with acceleration factor 2; and axial T2* GRE with acceleration factor 2. Parallel imaging of the cervical spine at 3T allows shortening of the examination time by 52%, preserving adequate image quality.

Magnetic resonance colonography in severe attacks of ulcerative colitis

To evaluate the ability of MR colonography (MRC) to detect lesions in severe attacks of ulcerative colitis (UC) and to assess its concordance with rectosigmoidoscopy. Eighteen patients underwent MRC and rectosigmoidoscopy. MRC consisted of a water-filled colonic procedure followed by T1/T2w images. Image quality was recorded. Inflammatory lesions and the existence of signs of severity were analysed. We calculated MR accuracy in the diagnosis of inflammatory lesions, as well as per segment and per patient concordance depending on the presence or absence of severe lesions. The MR image quality of the 108 segments was satisfactory. Endoscopy was used to study 36 segments (rectum and sigmoid). MRC had a positive predictive value of 100% and a sensitivity of 64% in the diagnosis of inflammatory lesions. Concordance for the diagnosis of severe lesions was excellent for the rectum (k = 0.85) and good for the sigmoid (k = 0.64). MRC diagnosed signs of severity in all patients affected at endoscopy. MRC also disclosed signs of severity located higher in the colon in four patients with nonsevere lesions at rectosigmoidoscopy. MRC can accurately diagnose inflammatory lesions in severe attacks of UC and significantly correlates with rectosigmoidoscopy in the diagnosis of severe lesions.

Technical performance evaluation of a human brain PET/MRI system

Technical performance evaluation of a human brain PET/MRI system. The magnetic field compatible positron emission tomography (PET) insert is based on avalanche photodiode (APD) arrays coupled with lutetium oxyorthosilicate (LSO) crystals and slip-fits into a slightly modified clinical 3-T MRI system. The mutual interference between the two imaging techniques was minimised by the careful design of the hardware to maintain the quality of the B (0) and B (1) field homogeneity. The signal-to-noise ratio (SNR) and the homogeneity of the MR images were minimally influenced by the presence of the PET. Measurements according to the Function Biomedical Informatics Research Network (FBIRN) protocol proved the combined system's ability to perform functional MRI (fMRI). The performance of the PET insert was evaluated according to the National Electrical Manufacturers Association (NEMA) standard. The noise equivalent count rate (NEC) peaked at 30.7 × 10(3) counts/s at 7.3 kBq/mL. The point source sensitivity was greater than 7 %. The spatial resolution in the centre field of view was less than 3 mm. Patient data sets clearly revealed a noticeably good PET and MR image quality. PET and MRI phantom tests and first patient data exhibit the device's potential for simultaneous multiparametric imaging. • Combination of PET and MRI is a new emerging imaging technology. • Evaluated brain PET/MRI enables uncompromised imaging performance. • PET/MRI aims to provide multiparametric imaging allowing acquisition of morphology and metabolism.

Overview on the efficacy and safety of gadobutrol: an MRI contrast agent for the CNS, body and vessels

Contrast-enhanced MRI has become a mainstay of clinical imaging with unlimited applications. However, a limitation of conventional contrast agents in MRI is the insufficient signal-to-noise ratios due to relaxivity. Gadobutrol (1.0 mmol/ml), an extracellular macrocyclic contrast agent of high stability, is indicated for contrast-enhanced MRI of adults, adolescents and children 2 years of age and older. Gadobutrol offers increased T1 relaxivity, which improves MR image quality and clinical use of the technique.

MRI guided electrophysiological intervention with a voltage-based electro-anatomic mapping system

Background MRI visualizes luminal & vessel-wall anatomy, and identifies edema & scar tissue, contributing to improved electrophysiological (EP) ablative procedures for treatment of Ventricular Tachycardia & Atrial Fibrillation. MRI-guided EP interventions will be performed for the foreseeable future partially in & outside MRI, due to the need for X-ray/Ultrasound-compliant devices. Electromagnetically tracked catheter procedures, today’ sn orm for most EP procedure phases; vascular navigation, Electro-Anatomic-Mapping (EAM, the diagnostic and therapeutic phases), can only be performed outside MRI. Separate MRI tracking is required in MRI, complicating EP procedures which require moving in & out of the bore [1,2]. Continuous catheter tracking using a single system would allow registration-free EAM in & outside MRI. The goal was developing an MR-compatible St. Jude Medical (SJM) EnSite NavX (ESN) voltage-based tracking [3]. ESN applies 5.8/8.0 kHz voltage bursts between 3 pairs of electrodes on the chest, detecting a catheter’s position [4], so a challenge for intra-MRI use is MR gradient ramps which interfere with ESN operation. Minimal MR Image Quality (IQ) reduction also needs to be insured, as well as <2oC patient-skin heating due to components in MRI.

Simultaneous PET/MR imaging in a human brain PET/MR system in 50 patients—Current state of image quality

ObjectivesThe present work illustrates the current state of image quality and diagnostic accuracy in a new hybrid BrainPET/MR. Materials and methods50 patients with intracranial masses, head and upper neck tumors or neurodegenerative diseases were examined with a hybrid BrainPET/MR consisting of a conventional 3T MR system and an MR-compatible PET insert. Directly before PET/MR, all patients underwent a PET/CT examination with either [18F]-FDG, [11C]-methionine or [68Ga]-DOTATOC. In addition to anatomical MR scans, functional sequences were performed including diffusion tensor imaging (DTI), arterial spin labeling (ASL) and proton-spectroscopy. Image quality score of MR imaging was evaluated using a 4-point-scale. PET data quality was assessed by evaluating FDG-uptake and tumor delineation with [11C]-methionine and [68Ga]-DOTATOC. FDG uptake quantification accuracy was evaluated by means of ROI analysis (right and left frontal and temporo-occipital lobes). The asymmetry indices and ratios between frontal and occipital ROIs were compared. ResultsIn 45/50 patients, PET/MR examination was successful. Visual analysis revealed a diagnostic image quality of anatomical MR imaging (mean quality score T2 FSE: 1.27±0.54; FLAIR: 1.38±0.61). ASL and proton-spectroscopy was possible in all cases. In DTI, dental artifacts lead to one non-diagnostic dataset (mean quality score DTI: 1.32±0.69; ASL: 1.10±0.31). PET datasets of PET/MR and PET/CT offered comparable tumor delineation with [11C]-methionine; additional lesions were found in 2/8 [68Ga]-DOTATOC-PET in the PET/MR. Mean asymmetry index revealed a high accordance between PET/MR and PET/CT (1.5±2.2% vs. 0.9±3.6%; mean ratio (frontal/parieto-occipital) 0.93±0.08 vs. 0.96±0.05), respectively. ConclusionsThe hybrid BrainPET/MR allows for molecular, anatomical and functional imaging with uncompromised MR image quality and a high accordance of PET results between PET/MR and PET/CT. These results justify the application of this technique in further clinical studies and may contribute to the transfer into whole-body PET/MR systems.

Improving quality of arterial spin labeling MR imaging at 3 tesla with a 32‐channel coil and parallel imaging

To compare 12-channel and 32-channel phased-array coils and to determine the optimal parallel imaging (PI) technique and factor for brain perfusion imaging using Pulsed Arterial Spin labeling (PASL) at 3 Tesla (T). Twenty-seven healthy volunteers underwent 10 different PASL perfusion PICORE Q2TIPS scans at 3T using 12-channel and 32-channel coils without PI and with GRAPPA or mSENSE using factor 2. PI with factor 3 and 4 were used only with the 32-channel coil. Visual quality was assessed using four parameters. Quantitative analyses were performed using temporal noise, contrast-to-noise and signal-to-noise ratios (CNR, SNR). Compared with 12-channel acquisition, the scores for 32-channel acquisition were significantly higher for overall visual quality, lower for noise and higher for SNR and CNR. With the 32-channel coil, artifact compromise achieved the best score with PI factor 2. Noise increased, SNR and CNR decreased with PI factor. However mSENSE 2 scores were not always significantly different from acquisition without PI. For PASL at 3T, the 32-channel coil at 3T provided better quality than the 12-channel coil. With the 32-channel coil, mSENSE 2 seemed to offer the best compromise for decreasing artifacts without significantly reducing SNR, CNR.

Safety and performance of TCI pumps in a magnetic resonance imaging environment*

Target controlled infusion (TCI) devices can be associated with significant safety concerns when used during magnetic resonance imaging (MRI). We tested the safety and compatibility of newer TCI systems in a 3-Tesla MRI environment. Two Asena PK and two Agilia TCI pumps were used to administer TCI propofol (at target blood concentrations of 0.5 and 6.0 μg.ml⁻¹) using the Marsh model under magnetic fields of up to 50 G with a T2-weighted sequence. We assessed the devices for projectile risk, accuracy of drug delivery, alarm function and effects on MR image quality. Both devices did not demonstrate any significant deflection at the tested field strengths, and performed within acceptable limits (cumulative error in total delivered volume < 3%; maximum 10-min interval error < 10%). The Asena pump caused minor artefacts on MR images. The TCI pumps tested perform well and safely implement pharmacokinetic software in a high magnetic field.

A Novel MRI-SIM Technique for Localization of Head and Neck Cancer Patients Using Immobilization and Surface Coils: Initial Clinical Experience

Purpose/Objective(s): To investigate the feasibility of developing a geometrically accurate MR imaging protocol at 1.5T for patients immobilized with the type S frame for head and neck radiotherapy. Materials/Methods: Protocol optimization was undertaken using a GE Optima 450 70cm wide bore MR-SIM unit commissioned for radiotherapy planning. MRI images (from the base of brain to below the clavicles) were acquired in five volunteers using an 8 channel phased array coil placed over the head and neck region while each subject was immobilized in the treatment position with a five point thermoplastic device. Pulse sequence parameters for 3DT1, T2 FRFSE and T2 CUBE (including matrix, slice thickness, bandwidth) were optimized to maximize SNR while minimizing the effect of chemical shift. MR image quality was rated by two independent Radiation Oncologists benchmarked against “gold standard” reference diagnostic images. Phantom measurements were then performed to quantify system related residual geometric distortions after application of a commercial gradient distortion correction algorithm for the effective FOV and the pulse sequences utilized. The optimized protocols were used to image five patients undergoing head and neck radiotherapy. MR and planning CT images were fused in the ECLIPSE treatment planning system using mutual information algorithm. Image quality, distortion and accuracy of image registration using internal anatomical reference landmarks (multiple regions of interest in the head, neck and thorax) were evaluated. Results: A high resolution MR-SIM imaging protocol (3DT1, T2 FRFSE and T2 CUBE) was developed with 3mm slice thickness/no gap that covers the base of brain to below the clavicles. From phantom studies, residual distortions were found to be < 1.0 mm within a 8 cm radius and < 1.5mm within a 12 cm radius of the scan centre. Image deformation was minimized in the head and neck region due to the reproducibility of image localization between CT and MRI. In terms of registration accuracy, anatomic landmark displacements in the x, y and z direction relative to CT were <2mm for all landmarks in the head and neck region but up to 6 mm in the thorax. Conclusions: We have developed a novel MRI-SIM technique that provides good image quality for patients immobilized for head and neck radiotherapy and is now in clinical use.

Polyelectrolyte coating of iron oxide nanoparticles for MRI-based cell tracking

Iron oxide-based magnetic nanoparticles (MNPs) offer unique properties for cell tracking by magnetic resonance imaging (MRI) in cellular immunotherapy. In this study, we investigated the uptake of chemically engineered NPs into antigen-presenting dendritic cells (DCs). DCs are expected to perceive MNPs as foreign antigens, thus exhibiting the capability to immunologically sense MNP surface chemistry. To systematically evaluate cellular uptake and T2/T2⁎ MR imaging properties of MNPs, we synthesized polymer-based MNPs by employing layer-by-layer (LbL) technology. Thereby, we achieved modification of particle shell parameters, such as size, surface charge, and chemistry. We found that subcellular packaging of MNPs rather than MNP content in DCs influences MR imaging quality. Increased local intracellular electron density as inferred from transmission electron microscopy (TEM) strongly correlated with enhanced contrast in MRI. Thus, LbL-tailoring of MNP shells using polyelectrolytes that impact on uptake and subcellular localization can be used for modulating MR imaging properties. From the Clinical EditorIn this study, layer-by-layer tailoring of magnetic NP shells was performed using polyelectrolytes to improve uptake by dendritic cells for cell-specific MR imaging. The authors conclude that polyelectrolyte modified NP-s can be used for modulating improving MR imaging quality by increasing subcellular localization.

Noise Removal in Magnetic Resonance Images using Hybrid KSL Filtering Technique

In Medical Diagnostic, Magnetic Resonance Images play a major role. Magnetic Resonance images are normally corrupted by random noise from the measurement process complicating the automatic feature extraction and analysis of clinical data. Because of this reason noise removal methods have been customarily applied to improve MR image quality. This work proposed a new scheme based on applying a series of filters, each used to modify the estimate into greater agreement, so that the output converges to a stable estimate providing noise free image. In this work, we have introduced a novel hybrid filter to reduce random noise in MR images by the combination of Kernel, Sobel and Low-pass (KSL) filtering techniques. The proposed method has been implemented using Matlab and compared with related state of art methods over synthetic and real clinical MR images showing a superior performance in all cases analyzed.

Assessing the image quality of pelvic MR images acquired with a flat couch for radiotherapy treatment planning

To improve the integration of MRI with radiotherapy treatment planning, our department fabricated a flat couch top for our MR scanner. Setting up using this couch top meant that the patients were physically higher up in the scanner and, posteriorly, a gap was introduced between the patient and radiofrequency coil. Phantom measurements were performed to assess the quantitative impact on image quality. A phantom was set up with and without the flat couch insert in place, and measurements of image uniformity and signal to noise were made. To assess clinical impact, six patients with pelvic cancer were recruited and scanned on both couch types. The image quality of pairs of scans was assessed by two consultant radiologists. The use of the flat couch insert led to a drop in image signal to noise of approximately 14%. Uniformity in the anteroposterior direction was affected the most, with little change in right-to-left and feet-to-head directions. All six patients were successfully scanned on the flat couch, although one patient had to be positioned with their arms by their sides. The image quality scores showed no statistically significant change between scans with and without the flat couch in place. Although the quantitative performance of the coil is affected by the integration of a flat couch top, there is no discernible deterioration of diagnostic image quality, as assessed by two consultant radiologists. Although the flat couch insert moved patients higher in the bore of the scanner, all patients in the study were successfully scanned.

MO-E-211-02: Breast MRI: Image Quality, Artifacts &amp; Quality Control

The quality of clinical breast MR images is influenced by the design and performance of the MR system, the breast coil, the breast MRI protocol, proper patient positioning, as well as patient-specific factors. MR system and coil quality control, protocol optimization, and technologist training are important to ensure high quality breast MR images. Common image quality issues encountered in clinical breast MR image include signal non-uniformity and inadequate breast tissue coverage, both of which can be influenced by the coil design. Inadequate fat or silicone saturation, wrap-around artifacts, and parallel imaging artifacts can often be minimized through technologist training. Breast MR protocol modification may be necessary to achieve improvements in signal-to-noise ratio (SNR), high contrast spatial resolution, or artifact reduction. Low SNR or non- uniform breast MR images may be coil performance related. Therefore it is important that breast coils be tested annually and whenever coil performance issues are suspected, as part of a comprehensive MR system quality control program. Educational objectives: 1. Describe common artifacts and image quality issues encountered in clinical breast MR images. 2. Discuss methods for addressing breast MR image quality problems. 3. Describe quality control procedures and accreditation programs specific to breast MRI.

SU‐D‐110‐04: Visual Biofeedback Combined with MRI for Respiratory‐Gated MR Imaging

Purpose: Irregular motion results in poor image quality, radiation targeting and tumor control. The aim of this study is to develop a novel respiratory motion control system using visual biofeedback combined with MRI and to demonstrate reduced respiratory motion artifacts and increase respiratory gating efficiency in MRI using the system. Methods: A visual biofeedback system has been combined with an MRI respiratory bellows belt at a 1.5T G.E. MRI system and at a 3T G.E. MRI system. The visual biofeedback system utilized (1) the external position information of the abdomen using an MR respiratory bellows belt to visually guide a patientˈs regular breathing within the magnet and (2) a respiratory gated Fast Spin Echo MR pulse sequence for MRI. The feasibility of respiratory motion control using the visual biofeedback within the MRI system has been tested in four volunteer studies without MRI scanning. Two MRI studies have been performed for image quality evaluation with/without the visual biofeedback system. All studies included a free breathing and a visual biofeedback breathing in a same day. Results: By using the visual biofeedback system with an MRI respiratory bellows belt, the respiratory motion reproducibility has been increased. Average RMSE (root mean square error) of breathing period has been reduced from 0.69s for free breathing to 0.38s for visual biofeedback breathing, while average RMSE of displacement changes insignificantly from 0.50cm of free breathing to 0.49cm of visual biofeedback breathing. In the obtained MR images, the image using visual biofeedback shows significant reduction of motion artefacts but respiratory gating efficiency improvement is insignificant with healthy volunteers. Conclusions: The study demonstrated the feasibility of respiratory motion control using visual biofeedback within the MRI system. This technique provides clinically applicable motion control in MRI scanning and improved MR image quality.