An end-to-end deep learning method for reconstructing SMS-PI accelerated musculoskeletal MRI.

Motivation: Current clinically available DL image reconstruction techniques can reconstruct 6-fold accelerated knee MRI with excellent image quality. However, higher acceleration factors reduced image quality, requiring novel image reconstruction approaches. Goal(s): To evaluate a novel end-to-end DL image reconstruction method for 8-fold combined simultaneous multislice (SMS) and parallel imaging (PI) acceleration of TSE MRI. Approach: Twenty 8-fold accelerated knee MRI studies were acquired and reconstructed using a novel end-to-end DL-based reconstruction network that jointly handles SMS and Pl acceleration. Results: Image quality and detection rates of abnormalities were comparable to the clinical standard with lower acceleration factors. Impact: First clinical results for 8-fold accelerated knee MRI using a dedicated end-to-end DL reconstruction approach for combined SMS and PI acceleration demonstrate image quality comparable to slower standard-of-care MRI, holding promise for significantly faster clinical TSE MRI in the future.

Phase-sensitive inversion recovery (PSIR) is a powerful cardiac MRI method to assess myocardial viability, which can eliminate the background phase and preserve the sign of the desired magnetization during inversion recovery. To shorten the acquisition time of myocardial viability imaging by introducing both simultaneous multislice (SMS) and parallel imaging (PI) into PSIR without additional acquisitions for calibration data. Prospective study. A high-resolution phantom and three vials with doped solutions matching typical postcontrast T1 and T2 values of scar, healthy myocardium, and blood; 18 patients (six with known myocardial infarction) were included in this study. 3T/segmented fast spoiled gradient echo pulse sequence. Phantom and in vivo experiments were performed to compare the performance of conventional PSIR, SMS accelerated PSIR (SMS-PSIR, 2× acceleration), and SMS as well as PI accelerated PSIR (SMS + PI-PSIR, 4× acceleration). In phantom experiments, the error maps, local signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were calculated. In in vivo experiments, the image quality and artifact level of each study were qualitatively graded (by three radiologists). G-factor maps were calculated. The infarct size presented as a percentage of the left ventricle was measured (full-width half-maximum). Acquisition time of each study was recorded. One-way analysis of variance, Kruskal-Wallis test. In phantom experiments, SNR and CNR were well preserved for SMS-PSIR, while they dropped for SMS + PI-PSIR, as expected. In 15 subjects, the overall image quality scores were not significantly different among conventional PSIR (3.70 ± 1.06), SMS-PSIR (3.78 ± 0.99), and SMS + PI-PSIR (3.47 ± 0.94; P = 0.20). The artifact level scores were also comparable among conventional PSIR (3.67 ± 1.04), SMS-PSIR (3.77 ± 1.03), and SMS + PI-PSIR (3.45 ± 1.00; P = 0.22). SMS-PSIR achieved negligible g-factor noise amplification (1.04 ± 0.03) and SMS + PI-PSIR showed higher g-factors (2.83 ± 0.48). The infarct size was consistent among conventional PSIR (22.51 ± 25.05%) and SMS-PSIR (22.98 ± 26.19%), as well as SMS + PI-PSIR (22.93 ± 25.68%; P = 0.98). The acquisition time of two short-axis slices for SMS-PSIR (17.6 ± 1.7 sec, 16 heartbeats) and SMS + PI-PSIR (9.8 ± 1.9 sec, 8 heartbeats) was 30% and 17% of that for conventional PSIR (56.2 ± 8.5 sec, 32 heartbeats), respectively. SMS can be implemented in PSIR without additional reference scan. The image quality is comparable with conventional PSIR, while the acquisition time is much shorter. The proposed method is also compatible with PI to further reduce the scan time. 2 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2019;50:1964-1972.

Magnetic resonance imaging (MRI) is crucial for accurately diagnosing a wide spectrum of musculoskeletal conditions due to its superior soft tissue contrast resolution. However, the long acquisition times of traditional two-dimensional (2D) and three-dimensional (3D) fast and turbo spin-echo (TSE) pulse sequences can limit patient access and comfort. Recent technical advancements have introduced acceleration techniques that significantly reduce MRI times for musculoskeletal examinations. Key acceleration methods include parallel imaging (PI), simultaneous multi-slice acquisition (SMS), and compressed sensing (CS), enabling up to eightfold faster scans while maintaining image quality, resolution, and safety standards. These innovations now allow for 3- to 6-fold accelerated clinical musculoskeletal MRI exams, reducing scan times to 4 to 6min for joints and spine imaging. Evolving deep learning-based image reconstruction promises even faster scans without compromising quality. Current research indicates that combining acceleration techniques, deep learning image reconstruction, and superresolution algorithms will eventually facilitate tenfold accelerated musculoskeletal MRI in routine clinical practice. Such rapid MRI protocols can drastically reduce scan times by 80-90% compared to conventional methods. Implementing these rapid imaging protocols does impact workflow, indirect costs, and workload for MRI technologists and radiologists, which requires careful management. However, the shift from conventional to accelerated, deep learning-based MRI enhances the value of musculoskeletal MRI by improving patient access and comfort and promoting sustainable imaging practices. This article offers a comprehensive overview of the technical aspects, benefits, and challenges of modern accelerated musculoskeletal MRI, guiding radiologists and researchers in this evolving field.

BACKGROUND. Deep learning (DL) superresolution image reconstruction enables higher acceleration factors for combined parallel imaging-simultaneous multislice-accelerated knee MRI but requires performance validation against external reference standards. OBJECTIVE. The purpose of this study was to validate the clinical efficacy of six-fold-accelerated sub-5-minute 3-T knee MRI using combined threefold parallel imaging (PI) and twofold simultaneous multislice (SMS) acceleration and DL superresolution image reconstruction against arthroscopic surgery. METHODS. Consecutive adult patients with painful knee conditions who underwent sixfold PI-SMS-accelerated DL superresolution 3-T knee MRI and arthroscopic surgery between October 2022 and July 2023 were retrospectively included. Seven fellowship-trained musculoskeletal radiologists independently assessed the MRI studies for image-quality parameters; presence of artifacts; structural visibility (Likert scale: 1 [very bad/severe] to 5 [very good/absent]); and the presence of cruciate ligament tears, collateral ligament tears, meniscal tears, cartilage defects, and fractures. Statistical analyses included kappa-based interreader agreements and diagnostic performance testing. RESULTS. The final sample included 124 adult patients (mean age ± SD, 46 ± 17 years; 79 men, 45 women) who underwent knee MRI and arthroscopic surgery within a median of 28 days (range, 4-56 days). Overall image quality was good to very good (median, 4 [IQR, 4-5]) with very good interreader agreement (κ = 0.86). Motion artifacts were absent (median, 5 [IQR, 5-5]), and image noise was minimal (median, 4 [IQR, 4-5]). Visibility of anatomic structures was very good (median, 5 [IQR, 5-5]). Diagnostic performance for diagnosing arthroscopy-validated structural abnormalities was good to excellent (AUC ≥ 0.81) with at least good interreader agreement (κ ≥ 0.72). The sensitivity, specificity, accuracy, and AUC values were 100%, 99%, 99%, and 0.99 for anterior cruciate ligament tears; 100%, 100%, 100%, and 1.00 for posterior cruciate ligament tears; 90%, 95%, 94%, and 0.93 for medial meniscus tears; 76%, 97%, 90%, and 0.86 for lateral meniscus tears; and 85%, 88%, 88%, and 0.81 for articular cartilage defects, respectively. CONCLUSION. Sixfold PI-SMS-accelerated sub-5-minute DL superresolution 3-T knee MRI has excellent diagnostic performance for detecting internal derangement. CLINICAL IMPACT. Sixfold PI-SMS-accelerated PI-SMS DL superresolution 3-T knee MRI provides high efficiency through short scan times and high diagnostic performance.

To grow and sustain referrals for musculoskeletal magnetic resonance imaging (MRI), successful group practices have recruited radiologists with specialized interest and training in musculoskeletal imaging. Musculoskeletal radiologists are valuable because they can talk-the-talk with orthopedists. Skillful communication is important. All orthopedists are trained to interpret radiographic studies during their residency. Sports orthopedists consider themselves self-sufficient when it comes to MRI interpretation. They have the advantage of knowing patients' symptoms and findings on physical examination; however, orthopedists are impressed by added value. Therefore, to get respect—as they say on the playing field—radiologists must see more in the images than their orthopedic colleagues. Successful musculoskeletal radiologists understand the anatomy of joints, the biomechanics of trauma, the sequela of injury, and the rationale behind surgical treatments. It also helps if they participate in conferences, observe arthroscopic or open surgical procedures, and learn about cutting-edge operative techniques. MRI, with its unique soft-tissue contrast resolution and multiplanar capabilities, has become one of the most widely applied diagnostic tools in the assessment of musculoskeletal disorders. The development of MRI has coincided with a boom in health consciousness and physical fitness. New echelons of weekend warriors advance onto football fields, ski slopes, and golf courses to challenge themselves and their aging joints and spines. They are breaking in new running shoes, mountain bikes, and snowboards, while sometimes breaking tendons, ligaments, and cartilage. Over the past 2 decades, the radiologic literature has exploded with publications describing the profound diagnostic capabilities of MRI in evaluating sports-related injuries. Using the latest imaging techniques, surface coils, and software, MRI has matched the capabilities of orthopedists to diagnose intra-articular disorders while maintaining the advantage in assessing myotendinous and occult osseous injuries. This two-part series of Topics in Magnetic Resonance Imaging is devoted to sports medicine. Authors will share current advanced information about MRI acquisition and interpretation (Part I appears in Topics in Magnetic Resonance Imaging, volume 14, number 1; Part II appears in Topics in Magnetic Resonance Imaging, volume 14, number 2). In the opening article of Part II, Chris Beaulieu and Garry Gold describe the elements of interactive MR imaging and interventional MR-guided procedures. Carl Winalski and Krishanu Gupta address the MRI of articular cartilage before and after surgical repair procedures. They describe the nuances of lesion detection, including the rationale supporting protocol design and pulse parameter selection. Jenny Bencardino, Ara Kassarjian and William Palmer review practical information about the hip, providing an overview of MR applications in sports-related osseous injury and myotendinous injury, as well as MR arthrographic applications in the evaluation of acetabular labral tear. In the knee article, William Palmer takes a biomechanical approach to explain the functional relationships between stabilizing structures. William Morrison highlights and illustrates the spectrum of osseous, tendon, and ligament abnormalities seen in athletes, and he comments on pertinent technical developments and anatomic subtleties. Diego Jaramillo and Rachel Oeppen focus on the young athlete and the specific injuries that predominate in the immature skeleton. This second issue concludes with the article by Leon Rybak and Martin Torriani, who explore the role of MRI in the assessment of trauma-related injuries to the myotendinous unit. We are indebted to all authors for their extraordinary contributions of time and expertise, to Scott Atlas for his invitation to compile this issue, and to Andrea Allison-Williams and Helen Powers for their invaluable editorial assistance.

Simultaneous multi-slice (SMS) bSSFP imaging enables stress myocardial perfusion imaging with high spatial resolution and increased spatial coverage. Standard parallel imaging techniques (e.g., TGRAPPA) can be used for image reconstruction but result in high noise level. Alternatively, iterative reconstruction techniques based on temporal regularization (ITER) improve image quality but are associated with reduced temporal signal fidelity and long computation time limiting their online use. The aim is to develop an image reconstruction technique for SMS-bSSFP myocardial perfusion imaging combining parallel imaging and image-based denoising using a novel noise map estimation network (NoiseMapNet), which preserves both sharpness and temporal signal profiles and that has low computational cost. The proposed reconstruction of SMS images consists of a standard temporal parallel imaging reconstruction (TGRAPPA) with motion correction (MOCO) followed by image denoising using NoiseMapNet. NoiseMapNet is a deep learning network based on a 2D Unet architecture and aims to predict a noise map from an input noisy image, which is then subtracted from the noisy image to generate the denoised image. This approach was evaluated in 17 patients who underwent stress perfusion imaging using a SMS-bSSFP sequence. Images were reconstructed with (a) TGRAPPA with MOCO (thereafter referred to as TGRAPPA), (b) iterative reconstruction with integrated motion compensation (ITER), and (c) proposed NoiseMapNet-based reconstruction. Normalized mean squared error (NMSE) with respect to TGRAPPA, myocardial sharpness, image quality, perceived SNR (pSNR), and number of diagnostic segments were evaluated. NMSE of NoiseMapNet was lower than using ITER for both myocardium (0.045 ± 0.021 vs. 0.172 ± 0.041, p < 0.001) and left ventricular blood pool (0.025 ± 0.014 vs. 0.069 ± 0.020, p < 0.001). There were no significant differences between all methods for myocardial sharpness (p = 0.77) and number of diagnostic segments (p = 0.36). ITER led to higher image quality than NoiseMapNet/TGRAPPA (2.7 ± 0.4 vs. 1.8 ± 0.4/1.3 ± 0.6, p < 0.001) and higher pSNR than NoiseMapNet/TGRAPPA (3.0 ± 0.0 vs. 2.0 ± 0.0/1.3 ± 0.6, p < 0.001). Importantly, NoiseMapNet yielded higher pSNR (p < 0.001) and image quality (p < 0.008) than TGRAPPA. Computation time of NoiseMapNet was only 20s for one entire dataset. NoiseMapNet-based reconstruction enables fast SMS image reconstruction for stress myocardial perfusion imaging while preserving sharpness and temporal signal profiles.

Magnetic resonance imaging (MRI) is the keystone of modern musculoskeletal imaging; however, long pulse sequence acquisition times may restrict patient tolerability and access. Advances in MRI scanners, coil technology, and innovative pulse sequence acceleration methods enable 4-fold turbo spin echo pulse sequence acceleration in clinical practice; however, at this speed, conventional image reconstruction approaches the signal-to-noise limits of temporal, spatial, and contrast resolution. Novel deep learning image reconstruction methods can minimize signal-to-noise interdependencies to better advantage than conventional image reconstruction, leading to unparalleled gains in image speed and quality when combined with parallel imaging and simultaneous multislice acquisition. The enormous potential of deep learning-based image reconstruction promises to facilitate the 10-fold acceleration of the turbo spin echo pulse sequence, equating to a total acquisition time of 2-3 minutes for entire MRI examinations of joints without sacrificing spatial resolution or image quality. Current investigations aim for a better understanding of stability and failure modes of image reconstruction networks, validation of network reconstruction performance with external data sets, determination of diagnostic performances with independent reference standards, establishing generalizability to other centers, scanners, field strengths, coils, and anatomy, and building publicly available benchmark data sets to compare methods and foster innovation and collaboration between the clinical and image processing community. In this article, we review basic concepts of deep learning-based acquisition and image reconstruction techniques for accelerating and improving the quality of musculoskeletal MRI, commercially available and developing deep learning-based MRI solutions, superresolution, denoising, generative adversarial networks, and combined strategies for deep learning-driven ultra-fast superresolution musculoskeletal MRI. This article aims to equip radiologists and imaging scientists with the necessary practical knowledge and enthusiasm to meet this exciting new era of musculoskeletal MRI.

Readout-segmented echo-planar imaging (rs-EPI) with simultaneous multislice (SMS) technology has been successfully applied to tumor research in many organs, but no feasibility study in rectal cancer has been reported, and the optimal acceleration of SMS with rs-EPI in rectal cancer has not been well determined yet. To investigate the feasibility of SMS rs-EPI of rectal cancer with different acceleration factors (AFs) and its influence on image quality, acquisition time and apparent diffusion coefficients (ADCs) in comparison to conventional sequences. All patients underwent rs-EPI and SMS rs-EPI with AFs of 2 and 3 (2 × SMS rs-EPI and 3 × SMS rs-EPI, respectively) using a 3T scanner. Acquisition times of the three rs-EPI sequences were measured. Image qualitative parameters (5-point Likert scale), signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), geometric distortion, and apparent diffusion coefficient (ADC) values of the three sequences were compared. A total of eighty-three patients were enrolled in our study. rs-EPI and 2 × SMS rs-EPI offered equivalently high overall image quality with a scan time reduction to nearly half (rs-EPI: 137 s, 2 × SM rs-EPI: 60 s). 3 × SMS rs-EPI showed significantly poorer image quality (p < 0.05). ADC values were significantly lower in 3 × SMS rs-EPI compared to rs-EPI in rectal tumors and normal tissue (tumor tissue: rs-EPI 1.19 ± 0.21 × 10-3 mm2/s, 3 × SMS rs-EPI 1.10 ± 0.26 × 10-3 mm2/s, p < 0.001; normal tissue: rs-EPI 1.68 ± 0.13 × 10-3 mm2/s, 3 × SMS rs-EPI 1.54 ± 0.20 × 10-3 mm2/s, p < 0.001). SMS rs-EPI using an AF of 2 is feasible for rectal MRI resulting in substantial reductions in acquisition time while maintaining diagnostic image quality and similar ADC values to those of rs-EPI when the slice distance and number of shots are the same among three rs-EPI sequences.

To prospectively document musculoskeletal magnetic resonance imaging (MRI) use and how it affects diagnosis, playing status, and treatment of Division I university athletes. We hypothesized that MRI often has little or no effect on the diagnosis or treatment plan. Cross-sectional study. Division I university sports medicine program. Division I university varsity athletes. Data were collected of musculoskeletal MRI use in varsity student athletes for 2 full academic years from 2010 to 2012 at a National Collegiate Athletic Association Division I institution. Timing of the injury, first physician visit, and MRI and pre- and post-MRI diagnosis, playing status, and treatment (surgical vs nonsurgical). Eighty-six MRIs were obtained during the 2 years studied. Average age was 19.9 (18-23) years. Forty-five percent of injuries occurred during competition season, 34% occurred preseason, and 21% occurred postseason. There was a change in diagnosis in 13 athletes (15.1%, 1 led to surgery performed after completion of the season), and there was a change in participation status in 8 athletes (9.3%, 5 increased and 3 decreased). Treatment plan changed in 1 athlete (1.2%). No athlete required surgery immediately after an MRI that was not already being planned. Every athlete treated nonsurgically pre-MRI was able to finish their season. Magnetic resonance imaging was obtained in 14% of athletes and did not demonstrate a clear benefit over history, examination, and radiographs. Magnetic resonance imaging did change diagnosis in 15% of cases, though it did not appreciably change the playing status or treatment plan.

3 - Magnetic Resonance Imaging of the Knee: Conventional and Novel Techniques

Diffusion tensor imaging (DTI) of the brain provides essential information on the white matter integrity and structural connectivity. However, it suffers from a low signal-to-noise ratio (SNR) and requires a long scan time to achieve high spatial and/or diffusion resolution and wide brain coverage. With recent advances in parallel and simultaneous multislice (multiband) imaging, the SNR efficiency has been improved by reducing the repetition time (TR ). However, due to the limited number of RF coil channels available on preclinical MRI scanners, simultaneous multislice acquisition has not been practical. In this study, we demonstrate the ability of multiband DTI to acquire high-resolution data of the mouse brain with 84 slices covering the whole brain in 0.2 mm isotropic resolution without a coil array at 9.4 T. Hadamard-encoding four-band pulses were used to acquire four slices simultaneously, with the reduction in the TR maximizing the SNR efficiency. To overcome shot-to-shot phase variations, Hadamard decoding with a self-calibrated phase was developed. Compared with single-band DTI acquired with the same scan time, the multiband DTI leads to significantly increased SNR by 40% in the white matter. This SNR gain resulted in reduced variations in fractional anisotropy, mean diffusivity, and eigenvector orientation. Furthermore, the cerebrospinal fluid signal was attenuated, leading to reduced free-water contamination. Without the need for a high-density coil array or parallel imaging, this technique enables highly efficient preclinical DTI that will facilitate connectome studies.

Effectiveness of simultaneous multislice accelerated readout-segmented echo planar imaging for the assessment of rectal cancer

Application of artificial intelligence to imaging interpretations in the musculoskeletal area: Where are we? Where are we going?

PurposeDuchenne muscular dystrophy (DMD) is the most common and severe form of muscular dystrophy. Current diagnostic tests like genetic testing, needle electromyography, and muscle biopsy are either not easily available or invasive, and they are impractical for assessing disease progression and treatment outcomes. Therefore, there is a need for a non-invasive and accurate investigative modality for DMD. In recent years, musculoskeletal magnetic resonance imaging (MRI-MSK) along with fractional anisotropy (FA) and diffusion tensor imaging (DTI) have become major non-invasive tools.Material and methodsT1-weighted MRI-MSK and FA measures of DTI of 78 DMD patients were retrospectively studied to identify the distinct pattern of muscle involvement and fatty infiltration as age and/or disease progresses. Correlation analysis was performed between MRI-MSK grade score vs. age, muscle strength, and Vignos scale. Spearman’s rank correlation coefficient was used.ResultsAs age increased, the MRI grade score and Vignos score increased. There was a statistically significant high positive correlation between MRI-MSK grade score and age, and low positive correlation with Vignos scores. With increasing age, the muscle strength on manual muscle testing (MMT) and FA value decreased. There was high negative correlation with muscle strength on MMT and low positive correlation between FA values and MMT score.ConclusionsOn T1-weighted MRI, a distinct pattern, extent, and distribution of lower limb muscle involvement can be seen. MRI-MSK grade score worsens with progressing age, reducing strength, and increasing functional impairment. FA alone may not be an accurate marker in assessing progression of DMD. MRI-MSK and other DTI measures should be further explored as diagnostic and prognostic tools for DMD.

This study aims to evaluate the utility of simultaneous multislice (SMS) acceleration for routine magnetic resonance neurography (MRN) at 3 T. Patients with multiple sclerosis underwent MRN of the sciatic nerve consisting of a standard fat-saturated T2-weighted turbo spin echo (TSE) sequence using integrated parallel acquisition technique (PAT2) acceleration and 2 T2 TSE sequences using a combination of PAT-SMS acceleration (1) to reduce scan time (PAT2-SMS2; SMS-TSE FAST ) and (2) for time neutral increase of in-plane resolution (PAT1-SMS2; SMS-TSE HR ). Acquisition times were 5:29 minutes for the standard T2 TSE, 3:12 minutes for the SMS-TSE FAST , and 5:24 minutes for the SMS-TSE HR . Six qualitative imaging parameters were analyzed by 2 blinded readers using a 5-point Likert scale and T2 nerve lesions were quantified, respectively. Qualitative and quantitative image parameters were compared, and both interrater and intrarater reproducibility were statistically assessed. In addition, signal-to-noise ratio/contrast-to-noise ratio (CNR) was obtained in healthy controls using the exact same imaging protocol. A total of 15 patients with MS (mean age ± standard deviation, 38.1 ± 11 years) and 10 healthy controls (mean age, 29.1 ± 7 years) were enrolled in this study. CNR analysis was highly reliable (intraclass correlation coefficient, 0.755-0.948) and revealed a significant CNR decrease for the sciatic nerve for both SMS protocols compared with standard T2 TSE (SMS-TSE FAST /SMS-TSE HR , -39%/-55%; P ≤ 0.01). Intrarater and interrater reliability of qualitative image review was good to excellent (κ: 0.672-0.971/0.617-0.883). Compared with the standard T2 TSE sequence, both SMS methods were shown to be superior in reducing pulsatile flow artifacts ( P < 0.01). Ratings for muscle border sharpness, detailed muscle structures, nerve border sharpness, and nerve fascicular structure did not differ significantly between the standard T2 TSE and the SMS-TSE FAST ( P > 0.05) and were significantly better for the SMS-TSE HR than for standard T2 TSE ( P < 0.001). Muscle signal homogeneity was mildly inferior for both SMS-TSE FAST ( P > 0.05) and SMS-TSE HR ( P < 0.001). A significantly higher number of T2 nerve lesions were detected by SMS-TSE HR ( P ≤ 0.01) compared with the standard T2 TSE and SMS-TSE FAST , whereas no significant difference was observed between the standard T2 TSE and SMS-TSE FAST . Implementation of SMS offers either to substantially reduce acquisition time by over 40% without significantly impeding image quality compared with the standard T2 TSE or to increase in-plane resolution for a high-resolution approach and improved depiction of T2 nerve lesions while keeping acquisition times constant. This addresses the specific needs of MRN by providing different imaging approaches for 2D clinical MRN.