RPS 613 - MRI spinning for development and sustainability

Author Block: S. J. Riederer¹, R. Pabi¹, E. Borisch¹, A. Froemming¹, A. Kawashima², N. Takahashi¹; ¹Rochester, MN/US, ²Phoenix, AZ/US
Purpose: To study if the long echo trains of fast-spin-echo (FSE) prostate T2-weighted imaging (T2-WI) cause the optimum echo time (TE) for distinguishing normal peripheral zone (PZ) vs. malignancy to deviate from the optimum TE chosen based on standard T2 decay.
Methods or Background: All work was done at 3 Tesla. Experiments were done using a standard (NIST) phantom containing ten vials with known T2 relaxation times. Vials 5 (T2=133.3 msec) and 6 (96.9 msec) were analyzed, having values closest to literature-taken T2 measurements in normal PZ (125 to 150 msec) and PZ malignancy and normal transition zone (TZ) (75 to 105 msec). The phantom was imaged using conventional spin-echo (TE 97, 113, 129, 153 msec; scan time 20 min for each) and a clinical T2-WI sequence (echo-train-length 21; TE-EFF 104, 114, 135, 145 msec; scan time 2:30 for each). 30 consecutive subjects with suspected prostate cancer were imaged using the clinical T2-WI sequence at both TE-EFF 100 and TE-EFF 150. Results were compared visually for relative contrast of PZ to TZ and any suspected lesions.
Results or Findings: Contrast-to-noise ratio (CNR) between Vials 5 and 6 for conventional spin-echo peaked at TE=125 msec, consistent with theory. However, CNR between the vials for the long-ETL T2-WI sequence was highest at TE-EFF 145 msec, 10% higher vs. 104 msec. In 18 of 30 patient studies the TE-EFF 150 series had superior contrast vs. TE-EFF 100, inferior in 2/30, and equivalent in 10/30.
Conclusion: Fast-spin-echo acquisition in prostate T2-WI artifactually prolongs the apparent T2 relaxation, causing the optimum echo times for distinguishing normal from malignant tissue to be higher than that predicted assuming tabulated T2 values. TE-EFF 150 consistently provides improved contrast vs. TE-EFF 100 msec.
Limitations: Limited number of subjects
Funding for this study: This work was funded by NIH.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Informed consent was provided by all human subjects.

Author Block: I. S. Shah, P. C. P. Joshi, V. Jahanvi; Pune/IN
Purpose: To compare the quality and image acquisition time between conventional and Compressed SENSE sequences in brain magnetic resonance imaging (MRI) in paediatric population.
Methods or Background: Thirty children (below the age of 18 years) undergoing MRI brain were included in this study. In addition to the routine sequences, one Compressed SENSE sequence was added. 2D - T1, T2, and FLAIR axial sequences were acquired for brain using conventional and Compressed SENSE techniques. One of each sequence was acquired in 10 patients undergoing an MRI brain study on a 3T MRI using coil 32 channel coil for adults and pediatric 8ch head coil for neonates.

Two consultant radiologists (with 35 years and 5 years experience in radiology) independently scored the image quality using the 5-point Likert scale based on resolution, visualization of anatomical regions, grey-white matter differentiation, sharpness of the image and artefacts. The subjective criteria details for image quality as per the 5-point Likert scale were: non-diagnostic (1), poor (2), moderate (3), good (4) and excellent (5).
Results or Findings: The time reduction achieved with 2D T1 at 2 reduction factor were 60 seconds(24%), with 2D T2 at reduction factor of 2.2 66 seconds(47.83%) and with 2D FLAIR at reduction factor 2 66 seconds(40%). Inter-rater agreement for overall diagnostic confidence was rated higher for Compressed SENSE (k – 0.632) than conventional (k – 0.464). Nonsignificant statistical difference was found regarding image quality and image contrast ratio between both techniques.
Conclusion: Compressed SENSE has potential in reducing the image acquisition time without compromising the image quality and diagnostic confidence. Motion artefacts are also reduced with reduction in time with the use of Compressed SENSE sequence.
Limitations: A small sample size.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Institutional ethics committee approval was obtained

Author Block: L. Wei¹, A. Volk¹, S. Campana Tremblay², J. Poujol³, S. Ammari¹, G. Garcia¹, C. Balleyguier¹, N. Lassau¹, F. Bidault¹; ¹Villejuif/FR, ²La Ciotat/FR, ³Buc/FR
Purpose: The purpose was to compare the CEST effect at 1.5T and 3T on high resolution 2D images provided by a deep-learning algorithm. In addition, the effect of saturation offset number reduction combined with B0 Mapping on MTRasym values was investigated, in order to minimize acquisition time for clinical use.
Methods or Background: The phantom consisted of 12 tubes filled with BSA at pH 7 for 3 different physiological concentrations, at 37 °C. Acquisitions were performed on 1.5T (GE Artist) and 3T (GE Signa Premier) MR scanners. CEST data were acquired with 2D SSFSE using the AIR Recon Deep Learning option (ARDL) for image reconstruction, CW saturation was used, with 61 offsets. SNR was compared with previously acquired data without ARDL. B0 correction was performed by using the chemical shifts of Z-spectra minima, and by using the 2D B0 Mapping GRE sequence. MTRasym was compared for different offset numbers ranging from 6 to 61 using B0 Map GRE.
Results or Findings: 2D CEST SSFSE using ARDL had better SNR compared to the sequence without ARDL at 1.5T and 3T. For the two B0 correction methods, MTRasym values were similar and increased with protein concentration at 1.5T and 3T. For ΔB0<0.3ppm, MTRasym variations were small for all offset numbers (max 6%) . For ΔB0>0.5ppm, MTRasym were similar for 16 offsets or more.
Conclusion: This study provided a comparison of 2D CEST SSFSE at 1.5T and 3T on a phantom carried out during the same imaging session. The findings open up the prospect of high-resolution time efficient APTw-CEST clinical MRI at 1.5T.
Limitations: However, the results are preliminary, hence repeatability studies and proof of concept in patients will be considered.
Funding for this study: This material is based upon work supported by the ANRT with a CIFRE fellowship granted to Lecong Wei.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: Not applicable

Author Block: C. Birkl¹, M. Panzer¹, C. Kames², A. Rauscher², B. Glodny¹, E. R. R. Gizewski¹, H. Zoller¹; ¹Innsbruck/AT, ²Vancouver, BC/CA
Purpose: Aceruloplasminemia (ACP) is a rare autosomal recessive disorder characterized by progressive iron accumulation in multiple organs, including the brain, liver, and pancreas. Magnetic Resonance Imaging (MRI) is commonly used to detect iron overload, with Quantitative Susceptibility Mapping (QSM) emerging as a promising method for assessing brain iron levels. Despite its potential, QSM faces challenges such as susceptibility artifacts and a lack of standardization. This prospective study aimed to evaluate the performance of different QSM algorithms in measuring brain iron in patients with severe iron overload, compared to healthy controls.
Methods or Background: QSM images were acquired using a 3D multi-echo gradient echo sequence in three patients with ACP and three healthy controls. We evaluated six QSM algorithms: (I) Fast Nonlinear Susceptibility Inversion (FANSI), (II) Improved Sparse Linear Equation and Least-Squares (iLSQR), (III) Morphology-Enabled Dipole Inversion (MEDI), (IV) Streaking Artifact Reduction (STAR) QSM with Rapid Open-source Minimum Spanning Tree (ROMEO) phase unwrapping, (V) STAR QSM with Laplacian phase unwrapping, and (VI) Multi-Echo Rapid Two-Step (MERTS) QSM. Regional susceptibility values were analyzed in the caudate nucleus, putamen, globus pallidus, and thalamus.
Results or Findings: We observed significant variability in susceptibility values across the different algorithms for patients with ACP. Among the algorithms tested, only one showed consistently elevated susceptibility values in the globus pallidus of ACP patients compared to healthy controls. Many susceptibility maps showed signal dropouts in brain regions with extreme iron overload.
Conclusion: Our findings suggest that only a subset of QSM algorithms reliably reflect extreme brain iron deposition. Additionally, the study highlights that performing echo combination prior to phase unwrapping and background field removal may introduce artifacts, resulting in lower-than-expected susceptibility values due to signal dropouts.
Limitations: A limitation is the small sample size.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the local ethics committee (number 1270/2021)

Author Block: L. Widmer¹, S. Bhumiwat², F. Porões¹, J. M. M. Froehlich³, H. Thoeny¹; ¹Fribourg/CH, ²Phatum Wan/TH, ³Zurich/CH
Purpose: Gadopiclenol (Elucirem™) is a recent high-relaxivity macrocyclic gadolinium-based contrast agent (GBCA), with limited data on its detailed concentration dynamics, essential for optimizing its use in diverse clinical scenarios. This study compared signal intensity (SI) curves from three GBCAs across various MRI sequences, field strengths, coils and concentrations.
Methods or Background: Signal intensity of gadopiclenol, gadoteric acid and gadobutrol vials were measured across 18 MRI sequences on 1.5T and 3T machines and 8 concentrations ranging from 0 to 25 mmol/L in an experimental in-vitro setting. Relationship between SI and concentrations were compared in SE, FSE, GRE and IR sequences.
Results or Findings: Concentration had no linear correlation with the SI. At the same concentrations, gadopiclenol produced higher maximal SI than the other two contrast agents in half of the sequences (50%, 9/18). In most sequences (56%, 10/18), gadopiclenol had a left-shifted curve maximum, reflecting higher SI at lower concentrations. Results of identification of curves patterns by sequence type are still pending.
Conclusion: Signal intensity curve analysis helps optimize imaging and injection parameters, though in-vivo application requires considering vessel and tissue distribution. These findings suggest using reduced dose of high-relaxivity agents compared to conventional GBCAs, supporting sustainable radiology.
Limitations: Experimental
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: None

Author Block: D. Takenaka, H. Nagata, T. Ueda, M. Nomura, T. Yoshikawa, Y. Ozawa, Y. Ohno; Toyoake/JP
Purpose: Deep learning reconstruction (DLR) and zero fill interpolation (ZIP) technique have been clinically applied on routine clinical MRIs. Recently, deep learning-based spatial resolution improving algorithm (Precise IQ Engine: PIQE) is developed to transform MR data from low-spatial resolution data to high-spatial resolution data. The purpose of this study was to directly compare utilities of PIQE for scan time reduction and image quality improvement of MRIs as compared with DLR with and without ZIP techniques.
Methods or Background: 28 consecutive patients suspected with 17 brain tumors, 6 spinal diseases and 5 musculoskeletal diseases were prospectively scanned with conventional MR (224-382256-512matrix) and new MR protocols (160-192192-416matrix). Then, both MR protocol data were reconstructed by DLR with and without ZIP technique or PIQE techniques (total five MR data sets). Each standard protocol was determined as conventional MR protocol reconstructed by DLR without ZIP technique. To compare scan time reduction and image quality improvement among all protocols, mean examination time and signal-to-noise ratios (SNRs) were compared between standard protocol and others by Dunnett's test. To evaluate qualitative image quality improvement, overall image quality, artifact and diagnostic confidence level were assessed by 5-point scales and compared between standard protocol and others by Steel's multiple comparison test.
Results or Findings: Mean examination times of new MR protocols were significantly shorter than that of conventional protocols (p<0.05), although SNRs had no significant differences. As compared with standard protocol, overall image quality and artifact were significantly improved by conventional protocol reconstructed by DLR with ZIP and new protocol with PIQE (p<0.05).
Conclusion: PIQE is equal to or more useful for reduce examination time and image quality improvements as with DLR with and without ZIP technique.
Limitations: LImited study number and no diagnostic performance evaluation
Funding for this study: Canon Medical Systems Corporation
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Fujita Health University Hospital

Author Block: H-M. Klein; Burbach/DE
Purpose: Purpose:
Develop a concept for grid independent, power saving MRI operation using a permanent magnet, solar energy, and a generator supported battery system.
Methods or Background: We installed a 0,4 T MRI system with an open design permanent magnet. Regenerative energy is produced with a 29,8 kWp solar array. To achieve grid independency, we installed an ´island solution´ using a 22 kWh LiFePO4 battery. For longer periods of power outage, and insufficient solar energy, a specially designed, direct current (DC), high voltage diesel generator is used. This generator simulates the power profile of a solar array, and is connected to the solar power converter, feeding the battery.
Results or Findings: Annual energy uptake of the MRI was 7.022 kWh in 2023. RIS and PACS components consumed 4.959 kWh. Heating and air conditioning consumed 12.500 kWh. Total energy consumption of the practice was 26.801 kWh. Total energy production was 30.930 kWh. Energy balance was positive with 4.129 kWh. Battery and DC generator can provide power grid independent operation. Without grid and solar energy, the practice has an energy consumption rate of max. 1,9 l gasoil/hour.
Conclusion: Grid independent, sustainable MRI operation is possible using permanent magnet technology, solar energy production, battery storage and a specially designed power generator.
Limitations: Only very few high quality low field MRI with permanent magnet technology are available in the market.
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: Not applicable

Author Block: Y. Jung, R. Alizadeh, M. Corwin, L. Hacein-Bey, A. M. Hernandez; Sacramento/US
Purpose: To quantify image quality, potential cost savings, and greenhouse gas emission reductions in accelerated brain and prostate MRI exams using deep learning reconstruction (DLR) and phantom imaging.
Methods or Background: A data logger and current transformer sensor were installed upstream of the power distribution unit on three 3T MRI platforms (GE, Siemens, United Imaging) to measure power consumption at 1-second intervals. The ACR phantoms were scanned using the T2 FLAIR sequence from routine brain MRI and the T2-weighted sequence from routine prostate MRI protocols. Phantom scanning was performed using three scan times with different acceleration factors, and images were reconstructed using conventional inverse Fourier transform (IFT) and DLR at three strength levels: Low, Medium, and High. Signal-to-noise ratio (SNR) and low-contrast detectability (LCD) measurements were taken using the ACR phantom. Total energy consumption was recorded for each acquisition.
Results or Findings: Total energy consumption decreased monotonically with reduced scan time across all systems and protocols. SNR and LCD, averaged across all scan times, were generally higher for DLR compared to IFT, and SNR increased with increasing DLR strength. Reducing the scan time by ~ 5 minutes with medium-strength DLR resulted in a ~65% reduction in energy consumption compared to a non-accelerated acquisition, while maintaining comparable SNR and LCD. Extrapolating these savings to all T2 FLAIR and T2-weighted sequences performed annually at our institution would result in estimated total savings of 15,944 USD and 74.1 MTCO2e, equivalent to 16.6 gasoline-powered passenger vehicles driven for one year.
Conclusion: DLR-accelerated MRI exams provide substantial reductions in cost and greenhouse gas emissions without compromising image quality in phantom imaging experiments.
Limitations: Patient data would be required to assess the actual impact of DLR on image quality and energy savings in clinical practice.
Funding for this study: RSNA Emerging Issues Environmental Impact and Sustainability grant
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: None