Low-field MRI System Research Articles

Feasibility and reproducibility of the PsAMRIS-H score for psoriatic arthritis in low-field-strength dedicated extremity magnetic resonance imaging

Objectives: The psoriatic arthritis magnetic resonance imaging scoring system (PsAMRIS-H) for the evaluation of inflammatory and destructive changes in PsA hands was validated on 0.6-T scanners. The applicability of the PsAMRIS-H on a low-field MRI system as a well-accepted, low-cost imaging modality was evaluated.Method: In 65 consecutive patients (31 males, median age 52 years), 73 scans on a 0.2-T dedicated extremity MRI system were obtained for evaluation of PsA. Images were scored for synovitis, tenosynovitis, periarticular enhancement, bone erosion, bone oedema, and proliferation, and the PsAMRIS-H score was calculated. The intraclass correlation coefficient (ICC) was calculated and the paired t-test conducted.Results: Intra-reader reliability for the total PsAMRIS-H score was good, with an ICC of 0.81 and 0.77 for readers 1 and 2, respectively, and inter-reader agreement was moderate (0.57 for each reader). However, the PsAMRIS-H score differed significantly between the two readers (22 vs. 31; p < 0.05). When individual components of the PsAMRIS were evaluated, intra- and inter-reader agreement was poor to moderate, especially for tenosynovitis and periarticular inflammation.Conclusions: Low-field 0.2-T MRI is capable of quantifying the PsAMRIS-H with good intra-reader reproducibility. However, low signal-to-noise ratio (SNR), low spatial resolution, and system artefacts limit the application of the PsAMRIS-H, leading to low inter- and intra-reader agreement for individual features.

Comparison of inversion recovery gradient echo with inversion recovery fast spin echo techniques for magnetic resonance imaging detection of navicular bone marrow lesions in horses

To compare navicular bone marrow lesion (BML) conspicuity in the feet of horses as determined via 2 fat-suppressed MRI techniques, including standard short tau inversion recovery (STIR) and inversion recovery gradient echo (IRGE). Feet (n = 150) of horses with lameness referable to the distal portion of the digit. STIR and IRGE sequences were obtained prospectively in all feet with a standing low-field equine MRI system. Presence of a BML was ascertained by identification of a characteristic combination of marrow alterations in T1-weighted, T2*-weighted, T2-weighted, and STIR images. Signal-to-noise and contrast-to-noise ratios were calculated on STIR and IRGE sequences in 56 feet with a navicular BML. Signal-to-noise and contrast-to-noise ratios of both sequences correlated linearly (r = 0.87 and r = 0.92, respectively) but were significantly higher for STIR images (mean ± SD, 22.6 ± 12.7 and 12.4 ± 11.4, respectively), compared with IRGE images (13.7 ± 8.0 and 5.9 ± 7.2, respectively). Results suggested that the IRGE sequence revealed BMLs significantly less conspicuously, compared with the standard STIR sequence. The 2 techniques cannot be used interchangeably, and IRGE is therefore not recommended as the sole fat-suppressed sequence for routine equine standing MRI protocols.

고온초전도 선재와 정합회로를 이용한 RF coil 제작에 대한 기초연구

The substantial improvement of the signal-to-noise ratio (SNR) can be achieved with small-size samples or low-field MRI system by high-temperature superconducting(HTS) RF coil. The typical HTS RF coil made of HTS thin film is expensive and is limited the coil geometry to planar surface coil. In this study, commercial Bi-2223 HTS tapes was used as RF coil for a 0.35T permanent MRI system. It has advantages of both much lower cost and easier fabrication over HTS thin film coil. SNR gain of the image obtained from the HTS RF coil over a conventional Cu RF coil at room temperature was about 2.4-fold and 1.9-fold using the spin echo pulse sequence and gradient echo pulse sequence respectively.

Development of A SQUID-Based Low-Field NMR/MRI System

Abstract In the case of most physiological tissues, longitudinal relaxation time (T1) depends on the applied magnetic field strength, and the T1-contrast between different types of tissues can be enhanced at low magnetic fields. Prepolarized SQUID-based MRI could be used to perform T1-weighted imaging at low fields. We aimed to develop a low-field MRI system that can provide clear T1-weighted images. The system consists of five sets of coils for generating magnetic fields required for generating the NMR/MRI signal and a first-order SQUID gradiometer for detecting free induction decay (FID) signals from a sample. We improved the polarizing-field-driving circuit to increase the field, and developed the SQUID-driving electronics to reduce the artifacts in the increased polarizing field. Using this system, we measured T1 of water–oil mixture sample. Clear T1 spectrum of each ingredient of the sample was obtained by CONTIN analysis.

SQUID-Based Low Field MRI System for Small Animals

Low field MRI and MEG are based on the ability of SQUID sensors to detect femtotesla magnetic fields. We are now developing a low field MRI system which can be integrated with the MEG system for small animals. The fast switching of the polarizing field inherently induces large transient signals in low field MRI experiments. We have optimized our system to minimize its influence and to improve the SNR of the MRI signal, by modifying a SQUID gradiometer and FLL (fluxed-locked loop) circuit and by reducing the back electromotive force of the polarizing coil. Our system includes five sets of magnetic field and gradient coils. Employing the “target field method” for the design of shielded planar coils, we achieved the measurement field homogeneity of 0.5% over 40 mm DSV (diameter sphere volume). We observed 4 pT NMR signal from 30 ml water by applying the polarizing field of 4.4 mT and the measurement field of 47 μT (1.9 kHz). The magnetic field noise spectral density is 9 fT/√Hz.

The design of a body RF coil for low-field open MRI using pseudo electric dipole radiation and simulated annealing

The paper presents a new body RF coil design scheme for a low-field open MRI system. The RF coil is composed of four rectangular loops which are made of wide copper strips located near the surfaces of the bottom and top pole faces of the permanent magnet. The body RF coil has been designed by using the pseudo electric dipole radiation (PEDPR) method with the Metropolis algorithm. In the calculation of the RF fields via the finite difference time domain (FDTD) method, the computational time increases as the RF frequency becomes lower. Moreover, the computational process using the FDTD method takes a very long time when the RF coil is optimized. The optimization requires varying the configuration of the RF coil system and performing successive calculations of field strength and field homogeneity. When we perform these successive calculations, the computational time can be reduced by using the PEDPR method, where the segmented current elements of the RF coil are treated as pseudo electric dipole radiation sources. Because the RF coil is made of wide strips, the variation of the current density on the strip has been considered in the B 1-field calculation. For each configuration of the RF coil system, the current distribution is calculated via circuit analysis, where each copper strip is considered as a parallel combination of current element lines. The preliminary field calculation study by the FDTD method verifies both the circuit analysis method for the current distribution and the PEDPR method for the radiation field strength. The optimization of the RF coil configuration is performed by the Simulated Annealing (SA) process using the Metropolis algorithm. Simulations have been performed for a 10 MHz RF frequency. The optimized RF coil has four rectangular loops of 37 cm × 100 cm with 6.5 cm wide strips which are separated vertically 49 cm and horizontally center-to-center 63 cm. In the 25 cm diameter of spherical volume (DSV), the design results show a good field inhomogeneity of the B 1-field below 0.49 dB (5.8%).

Retrospektive Analyse der 5-Jahres-Ergebnisse von MRT-gesteuerten Biopsien im Niederfeldsystem

The purpose of this study was to evaluate the safety and clinical value of MR-guided biopsies in an open 0.2 T low-field system. A total of 322 patients with suspicious lesions of different body regions were biopsied in a low-field MRI system (0.2 T, Concerto, Siemens). The procedures were guided using T 1-weighted Flash sequences (TR/TE = 100/9; 70 degrees). The lesions were repeatedly biopsied using the coaxial technique with a 15-gauge (diameter 2 mm) puncture needle. Complications and biopsy findings were analyzed retrospectively. In all cases the biopsy procedures were successfully performed with MR guidance. In 298 patients diagnosis was able to be confirmed on the basis of the probes. The clinical follow-up showed that in 24 patients the lesions were missed by MR-guided biopsy. From this a sensitivity of 86%, a specificity of 87% and an accuracy of 93% were calculated. In two patients major complications were observed (morbidity rate 0.6 %). MR-guided biopsy can be performed safely and precisely in a low-field MR system and are a supplement to US or CT-guided biopsies.

IRM opératoire et chirurgie de l’épilepsie

Intraoperative imaging, in particular intraoperative MRI, is a developing area in neurosurgery and its role is currently being evaluated. Its role in epilepsy surgery has not been defined yet and its use has been limited. In our experience with a compact and mobile low-field intraoperative MRI system, a few epilepsy surgeries have been performed using this technique. As the integration of imaging and functional data plays an important role in the planning of epilepsy surgery, intraoperative verification of the surgical result may be highly valuable. Therefore, teams that have access to intraoperative MRI should be encouraged to use this technique prospectively to evaluate its current relevance in epilepsy surgery.

MR-guided breast biopsy and hook wire marking using a low-field (0.23 T) scanner with optical instrument tracking

The purpose of this study was to evaluate the technical feasibility of MR-guided percutaneous breast biopsy (LCNB) and breast hook wire marking in a low-field (0.23 T) MRI system with optical instrument tracking. MR-guided core biopsy and/or hook wire marking was performed on 13 lesions observable at MR imaging only. Seven breast LCNBs and 10 hook wire markings were performed under MR guidance on 11 patients. The diagnosis was confirmed by excision biopsy or mastectomy in 12 lesions and with histopathological and cytological diagnosis and 12-month clinical follow-up in one lesion. All lesions seen in the high-field scanner were also successfully identified and targeted in the low-field scanner. The following procedures were typically technically successful. There were difficulties due to unsatisfactory functioning of some core biopsy guns. Detailed description of low-field MR guidance and optical tracking in breast biopsies is provided. The procedure seems accurate and safe and provides means to obtain a histological diagnosis of a breast lesion only seen with MRI. The low-field biopsy system is comparable to the high-field MRI system. MR-compatible biopsy guns need to be improved.

MR-guided transgluteal biopsies with an open low-field system in patients with clinically suspected prostate cancer: technique and preliminary results.

The purpose of this study was to examine the feasibility and safety of MR-guided biopsies with a transgluteal approach in patients with uncertain or suspicious prostate lesions. Twenty-five patients with uncertain or suspicious focal prostate lesions detected by high-field MR imaging of the prostate gland using endorectal coil imaging were biopsied with a transgluteal approach in a low-field MRI system (0.2 T, Concerto, Siemens). The procedures were guided using T1-weighted FLASH sequences. The prostate gland was biopsied repeatedly with a coaxial technique through a 15-gauge pencil tip with a 16-gauge biopsy handy (median 3.8 samples per patient). Complications and biopsy findings were documented retrospectively. Using T1-weighted sequences biopsy procedures were performed successfully with MR guidance in all cases without any side effects or complications. The median intervention time was 11.3 min. Pathological findings revealed ten cases of hyperplasia or atrophy, three cases of prostatitis, ten cases of carcinoma and two cases of normal tissue. The clinical follow-up showed that in two patients prostate cancer was missed at MR-guided biopsy. Transgluteal MR-guided biopsy of the prostate gland is a safe and promising approach for histological clarification of uncertain or suspicious lesions.

MRI-guided periradicular nerve root infiltration therapy in low-field (0.23-T) MRI system using optical instrument tracking

The purpose of this study was to evaluate the feasibility of the MRI-guided periradicular nerve root infiltration therapy. Sixty-seven nerve root infiltrations under MRI guidance were done for 61 patients suffering from lumbosacral radicular pain. Informed consent was acquired from all patients. A 0.23-T open-MRI scanner with interventional tools (Outlook Proview, Philips Medical Systems, MR Technologies, Finland) was used. A surface coil was used in all cases. Nerve root infiltration was performed with MRI-compatible 20-G needle (Chiba type MReye, Cook, Bloomington, Ind.; or Manan type, MD Tech, Florida). The evaluation of clinical outcome was achieved with 6 months of clinical follow-up and questionnaire. The effect of nerve root infiltration to the radicular pain was graded: 1=good to excellent, i.e., no pain or not disturbing pain allowing normal physical activity at 3 months from the procedure; 2=temporary, i.e., temporary relief of pain; 3=no relief of pain; and 4=worsening of pain. As an adjunct to MRI-guided positioning of the needle the correct needle localization by the nerve root was confirmed with saline injection to nerve root channel and single-shot fast spin echo (SSFSE) imaging. The MRI guidance allowed adequate needle positioning in all but 1 case (98.5%). This failure was caused by degeneration-induced changes in anatomy. Of patients, 51.5% had good to excellent effect with regard to radicular pain from the procedure, 22.7% had temporary relief, 21.2% had no effect, and in 4.5% the pain worsened. Our results show that MRI guidance is accurate and safe in performing nerve root infiltration at lumbosacral area. The results of radicular pain relief from nerve root infiltration are comparable to CT or fluoroscopy studies on the subject.

Magnetic Resonance Imaging (MRI) of fingers with a high spatial resolution using a low-field magnetic system.

For our low-field resistive MRI-System we have developed suitable rf-probes to be able to image fingers with a possibly high spatial resolution. To reduce the measurement time and to improve the signal/noiseratio, which is inherently bad in a low field system, we used a multi-echo sequence, accumulating up to six echo signals for each image. For a 256 x 256 pixel matrix we achieved with this technique a spatial resolution of 176 microns/pixel in measuring times between 4 and 5 minutes. In comparison to healthy persons also first MR-images of patients suffering on arthrosis at the fingers and small joint injures were acquired. Using a 512 x 512 pixel matrix and a field of view of 5 cm images with a resolution of 80 microns/pixel were recorded in a total measuring time of less than 10 minutes.

MRI-guided trephine biopsy and fine-needle aspiration in the diagnosis of bone lesions in low-field (0.23 T) MRI system using optical instrument tracking.

The purpose of this study was to evaluate the feasibility of MRI-guided bone biopsy with optical instrument tracking and evaluate advantage of combined fine-needle aspiration (FNA) with trephine biopsy. Twenty trephine bone biopsies and 13 FNAs were performed under MRI and CT guidance in 14 patients. Informed consent was obtained from all patients. The evaluation of diagnostic accuracy was achieved by comparing the histopathological and cytological diagnosis with current or final diagnosis made during 6-month clinical follow-up. A 0.23-T open MRI scanner with interventional tools (Outlook Proview, Marconi Medical Systems, Cleveland, Ohio) was used. A surface coil was used. For trephine biopsy MRI-compatible bone biopsy set (Daum medical, Schwerin, Germany) was used. The FNA was performed with MRI compatible 20-G needle (Cook, Bloomington, Ind.). The diagnostic accuracy of MRI-guided trephine biopsy was 95%. The FNA sample diagnosis concurred with the histological in 54%. Our results show that MRI guidance in bone biopsies is accurate and safe. It is comparable to CT-guided or open biopsy. The role of combined FNA with bone biopsies remains controversial.

Wireless application of an intravascular resonator within the Vena Cava for MRI

This paper describes an investigation of MR-active vena cava filter (VCF) implants. The active part of the VCF is a resonating circuit, which is tuned to the Lamor frequency of the MR scanner. The experiments were performed on a low field MRI system. The susceptibility artifact behaviour of different commercially available VCFs and their materials was investigated. Three MR-active VCF prototypes were built based on these results and evaluated. No lead connection between the resonating circuit of the implant and the MR system is necessary for reception of the MR signal. Localisation of the prototypes within the MR image was possible by their signal enhancement. The results provide a basis for further developments and investigations of MR-active VCF implants for high field MR scanners.

Sedation and monitoring of paediatric patients undergoing open low‐field MRI

The purpose of this study was to determine the need, effectiveness and safety of sedation and monitoring in infants and children in a paediatric open low‐field MRI system. Of 274 patients (median age 9 y) examined, only 74 children (median age 25 mo) needed sedation. Sedation was achieved by intravenous administration of midazolam (0.2 mg/kg) and etomidate (0.2 mg/kg). Mean total doses required were 0.28 and 0.27 mg/kg, respectively. With the exception of eight primarily ventilated patients, all children breathed spontaneously. 02 saturation, arterial blood pressure and ECG were monitored. The low resonance frequency of the MRI system required a specially designed high frequency (HF) shielding of the monitor system to avoid HF artifacts. The overall sedation rate was markedly lower (74/274 = 27%) compared to a control group previously examined in a closed high‐field MRI system (52/111 =47%). This was due to a significant lower need for sedation in patients aged up to 10 y (p≤ 0.0001) in the open MRI unit. General anaesthesia could be avoided in all patients. No significant movement artifacts occurred in any of the MRI examinations and no serious side effects were observed. Conclusions: MRI of children is easier in an open MRI system and with fewer sedations, as in closed high‐field systems. Sedation by a combination of midazolam and etomidate is highly effective and safe. Monitoring devices for high‐field systems may have to be modified for low‐field systems. An in‐hose paediatric MRI unit with an open and special paediatric design is of major advantage for imaging pediatric patients.

Thick film YBCO receive coils for very low field MRI

Superconducting receive coils can be used to improve the signal to noise ratio of a MRI system. The greatest improvement is seen in cases where it is the coil noise dominates the noise. This is the case for low field MRI where surface coils are used. The properties of large area (100 mm square) YBCO thick film coils designed for a compact very low field MRI system are reported. The properties are compared with copper and silver mimics.

MRI simulation of bronchogenic carcinomas using an open low-field MRI

The initial target volume for primary radiation therapy of lung cancer is usually determined with the aid of computed tomography. Due to the axial CT-scans the simulation of the RT-field is often difficult. MRI in its superior ability to demonstrate and characterize soft tissue and its possibilities of multiplanar imaging can be beneficial. As MRI is less available and more expensive the use of MRI in radiotherapy planning is still restricted. With the introduction of open low-field MRI-systems there is now a cost-saving alternative. The aim of this study was the clinical evaluation of the use of a new open low-field MRI in radiotherapy planning of bronchogenic cancer. Fifteen patients undergoing primary radiotherapy for lung cancer were studied using an open low-field MRI-system (Picker Outlook 0.23 Tesla). Conventional CT-assisted treatment planning was compared to a MRI-assisted procedure. Contours from coronary T1-weighted MRI-sections were superimposed onto the simulator radiograph using a subtrascope (MR-simulation). Open low-field MR-imaging using T1-weighted sequences resulted in excellent delineation of tumor masses from mediastinal fat, the airways and the vascular structures as well as the radial tumor infiltration into the vicinity of the lung (Figures 1a to 1c). This allowed an exact and reproducible transfer of tumor contours onto the simulator radiograph. The MR-simulation led to optimization in the field configuration in 5/15 patients (Figure 2). Open low-field MRI-systems can be very useful in treatment planning. They are less expensive and need less extensive rebuilding compared to high-field MRI-systems. In the radiotherapy planning of bronchogenic carcinoma the MR-simulation is reasonable and clinically practicable. One of the main advantages of open MRI-systems in comparison to CT and standard MRI-systems in radiotherapy planning is that there is a much greater variety of treatment positions.

Comparison of low field (0.2T) and high field (1.5T) MR imaging in the differentiation of torned from intact menisci

Purpose: To evaluate the usefulness of a low field MRI system (0.2T; Esaote, Biomedica) for the evaluation of meniscal tears with regard to anatomic site, and to compare the results with findings from a high field unit (1.5T; Siemens, Erlangen). Material and Methods: MRI was performed in 25 patients in a low field (0.2T; Esaote, Biomedica), and a high field (1.5T; Siemens, Erlangen) MRI unit. The images were analyzed for the presence or absence of meniscal tears and the confidence of decision making. Results were further analyzed for the number of identical and unidentical findings on both imaging modalities. In seven patients, arthroscopy was performed and the findings compared with the results from MR imaging. Statistical analysis was performed by χ 2-test, Wilcoxon test and Friedman analysis. Results: Qualitative evaluation of the level of confidence in decision making was significantly superior on high field strength images. When comparing the evaluations from both image modalities in 21 of 25 patients (84%), the diagnosis concerning the presence or absence of meniscal tears was identical. Conclusion: Although low field MR imaging might offer diagnostic potential concerning the presence or absence of meniscal tears, the level of confidence in decision making is significantly superior with high field strength imaging, probably reflecting the higher conspicuity of lesions from high field strength units.

Examination of the hand and wrist joints with a dedicated low-field MRI device

Purpose of this study was to evaluate the diagnostic value of a low field dedicated MRI system in hand and wrist imaging. All 308 exams of the hand and wrist, that were performed on a low-field dedicated MRI system (Artoscan, Esaote Biomedica, Italy) in our institution in 1996, and high-field MRI exams performed in addition as part of the diagnostic work-up, were evaluated and correlated to final operative (n = 64) and histologic (n = 12) reports. 90% of all low-field MRI scans stated a diagnosis according to clinical suspicion. In 62% the clinical question was answered, and in 26% additional pathologies were identified. An MR-diagnosis completely different from the clinical suspicion was stated in 2%. High field exams contributed additional information in 6 of 36 patients. In 3 patients a tumor was not shown completely in the limited field-of-view of the dedicated low-field MRI-system. Frequency-selective fat-suppression pulse sequences and a better spatial resolution were the reasons for the additional information obtained in the other three patients. Low-field dedicated MR-imaging is a valuable method in the extensive work-up of the hand and wrist. Osseous, ligamentous and tendinous pathologies are well depicted. Large or infiltrative tumors should be referred to a high-field system.

A tuned SQUID amplifier for MRI based on a DOIT flux locked loop

We have developed a 4.2 K, flux locked, tuned d.c. SQUID amplifier to improve the SNR of a low field MRI system operating at 0.01 T (425 kHz). The flux locked loop, based on the Direct Offset Integration Technique, has a noise level of 2.6 /spl mu//spl Phi//sub 0/ Hz/sup -1/2/ and a slew rate of 1.6/spl times/10/sup 6/ /spl Phi//sub 0/ s/sup -1/ at 425 kHz when used with a commercially obtained SQUID. The high intrinsic Q-factor of the MRI pick-up coil is damped by the action of the loop and by an additional feedback circuit to provide imaging bandwidths of up to 10 kHz. We have developed a special low noise liquid helium cryostat so that the final system has a magnetic field resolution of 0.08 fT Hz/sup -1/. This receiver was used in a small scale MRI system to image non-conducting test objects and the human arm. The images show significant improvements in SNR over those obtained with an equivalent room temperature receiver.