Faculty Bibliography

BackgroundRapid real-time magnetic resonance (MR) sequences enable dynamic articular kinematic assessment. The abduction-external rotation (ABER) position has long been used to characterize glenohumeral pathology.PurposeTo evaluate a dynamic gradient recall echo (GRE) sequence for ABER-positioned glenohumeral joint kinematic assessment correlating with subjective instability and clinical apprehension testing.Material and MethodsSymptomatic patients were scanned using a routine MR arthrogram protocol supplemented by an additional "dynamic ABER" GRE technique acquired with the arm abducted and then internally-externally rotated in real time. Dynamic motion of the humeral head between the extremes of motion in the abducted and externally rotated positions was evaluated. The cohort was followed for 3 years.ResultsA total of 15 dynamic ABER studies in 15 different patients were evaluated by three readers (right: n=9; left: n=6), with a mean age of 30 years (range=19-45 years). Good accuracy of the humeral head excursion between the abducted and externally-internally rotated positions (AUC=0.88) was observed as a test for positively detecting instability. An association was detected between clinical instability and mean humeral head excursion as measured by all three readers (P = 0.026), although no association between positive apprehension testing and mean humeral head excursion was detected. There was a trend towards surgery-naïve patients with higher mean humeral head excursion subsequently undergoing surgical management (P=0.088), although this did not reach statistical significance.ConclusionCorrelation between subjective instability and humeral head translation demonstrated on a dynamic ABER sequence added to MR shoulder arthrograms was observed but without association with clinical apprehension testing.

Rheumatoid arthritis is a common chronic inflammatory arthritis, primarily characterized by proliferative synovitis of the small joints of the appendicular skeleton. Rapidly evolving treatment regimens for rheumatoid arthritis patients, including conventional and biologic disease-modifying antirheumatic drugs, promise improved quality of life and prognosis. These treatments necessitate earlier clinical detection of rheumatoid arthritis, a goal that has forever changed the role of imaging in this pursuit. This review discusses the dominant imaging modalities in state-of-the-art rheumatoid arthritis management: radiography, magnetic resonance imaging, and ultrasound. It focuses on hand and wrist evaluation, with key relevant findings detected by each modality and recent evidence supporting them. Emerging techniques are also described in this discussion to understand likely future radiologic contributions to rheumatoid arthritis diagnosis and management.

Ultrasound (US) is a valuable tool in the evaluation of arthritis both for diagnosis and treatment response. Pertinent findings such as joint effusions, synovitis, bursitis, bone erosions, tenosynovitis, and enthesitis can all be readily evaluated sonographically. In this article, we describe specific considerations in the US evaluation of rheumatoid arthritis, psoriatic arthritis, juvenile idiopathic arthritis, crystalline arthritis (gout, calcium pyrophosphate deposition disease, and hydroxyapatite deposition disease), septic arthritis, and osteoarthritis with attention to certain differentiating features. The potential role of US in the diagnosis of specific arthritides is discussed, together with an overview of newer technologies and future directions.

OBJECTIVE:To evaluate the Neuropathy Score-Reporting and Data System (NS-RADS) MRI grading system in conjunction with electrodiagnostic (EDx) testing for radial neuropathy at the elbow. MATERIALS AND METHODS/METHODS:Patients presenting between 2010 and 2023 with suspected radial neuropathy who underwent both EDx testing in the form of electromyography and nerve conduction studies and MRI within a 12-month period were evaluated. Three blinded radiologists used the NS-RADS grading system to evaluate nerve entrapment (E grades), muscle denervation (M grades) proximally within the supinator/extensor carpi radialis brevis (ECRB), and more distally within the forearm extensor muscles. These grades and the presence of lateral epicondylitis were then correlated with EDx abnormalities. RESULTS:Forty-nine participants were included. Inter-reader reliability for M grades in the forearm extensor muscles was good (ICC = 0.90 [95% CI = 0.83 - 0.94], p < 0.001), as was reliability for the supinator/ECRB muscles (ICC = 0.91 [95% CI = 0.86-0.95], p < 0.001). Inter-reader reliability for E grades was moderate (ICC = 0.83 [95% CI = 0.69-0.90], p < 0.001). Patients with positive EDx studies had a significantly different distribution of M grades for the forearm extensors and supinator/ECRB than those with negative studies (all p values < 0.001). However, overall consensus reads showed no significant difference in the distribution of E grades between patients with positive and negative EDx studies. CONCLUSION/CONCLUSIONS:Muscle grading strongly correlated with EDx positivity, with a high level of inter-reader agreement for muscle denervation-related alterations. Nerve grading, however, did not show a statistical correlation.

Background Conventional chemical shift selective (CHESS) fat suppression may fail in distal extremity MRI due to sensitivity to field inhomogeneities. Purpose To develop a patient-specific fat-suppression method for distal extremity 3-T MRI by exploiting the spectral heterogeneity adaptive radiofrequency pulse (SHARP) technique and to compare it to fat suppression with CHESS. Materials and Methods SHARP uses the routinely acquired frequency spectrum at MRI calibration to adapt the frequency range and time-bandwidth product of the fat-suppression pulse. In this prospective study, fat suppression by SHARP was assessed by numerical simulations, phantom experiments, and imaging in 15 asymptomatic participants who underwent ankle, foot, and hand (in superman and hand-by-the-side positions) MRI using SHARP, CHESS, and reference standard (short-tau inversion recovery or Dixon) techniques. Three readers ranked the MRI scans from 1 (best) to 3 (worst) regarding fat-suppression homogeneity. The added value of SHARP was defined as the difference between the proportions of images where SHARP outranked CHESS and where CHESS outranked SHARP. Friedman, Wilcoxon signed rank, and χ² tests were used to compare in vivo data. Results At numerical simulations, SHARP showed 0% water and 62%-70% fat suppression, whereas CHESS showed 2% water and 57% fat suppression. Phantom data demonstrated lower fat-suppression inhomogeneity indexes with Dixon (1.0%) and SHARP (2.4%) compared with CHESS (10.7%). In 15 participants (mean age, 38.5 years ± 12.8 [SD]; six female participants), mean ranking by readers of fat homogeneity in the reference technique (ankle, foot, hand in superman position, and hand-by-the-side position: 1.02, 1.02, 1.03, and 1.06, respectively) was higher than those with SHARP (1.39, 1.46, 1.50, and 1.66, respectively), which were higher than those with CHESS (1.64, 1.80, 1.61, and 1.80, respectively) (all P < .001). The added value of SHARP was highest for images in the foot (389 of 1158; 33.6%; P < .001 vs other joints), followed by the ankle (247 of 971 [25%]; P < .001 vs both hand positions), and lowest for hand-by-the-side and hand in superman positions (158 of 1223; [13%] and 133 of 1193 [11%], respectively; P = .18). Conclusion SHARP provided more homogeneous fat suppression than CHESS. © RSNA, 2024 Supplemental material is available for this article.

OBJECTIVES/OBJECTIVE:To evaluate a dynamic half-Fourier acquired single turbo spin echo (HASTE) sequence following anterior cervical discectomy and fusion (ACDF) at the junctional level for adjacent segment degeneration comparing dynamic listhesis to radiographs and assessing dynamic cord contact and deformity during flexion-extension METHODS: Patients with ACDF referred for cervical spine MRI underwent a kinematic flexion-extension sagittal 2D HASTE sequence in addition to routine sequences. Images were independently reviewed by three radiologists for static/dynamic listhesis, and compared to flexion-extension radiographs. Blinded assessment of the HASTE sequence was performed for cord contact/deformity between neutral, flexion, and extension, to evaluate concordance between readers and inter-modality agreement. Inter-reader agreement for dynamic listhesis and impingement grade and inter-modality agreement for dynamic listhesis on MRI and radiographs was assessed using the kappa coefficient and percentage concordance. RESULTS:A total of 28 patients, mean age 60.2 years, were included. Mean HASTE acquisition time was 42 s. 14.3% demonstrated high grade dynamic stenosis (> grade 4) at the adjacent segment. There was substantial agreement for dynamic cord impingement with 70.2% concordance (kappa = 0.62). Concordance across readers for dynamic listhesis using HASTE was 81.0% (68/84) (kappa = 0.16) compared with 71.4% (60/84) (kappa = 0.40) for radiographs. Inter-modality agreement between flexion-extension radiographs and MRI assessment for dynamic listhesis across the readers was moderate (kappa = 0.41; 95% confidence interval: 0.16 to 0.67). CONCLUSIONS:A sagittal flexion-extension HASTE cine sequence provides substantial agreement between readers for dynamic cord deformity and moderate agreement between radiographs and MRI for dynamic listhesis. CLINICAL RELEVANCE STATEMENT/CONCLUSIONS:Degeneration of the adjacent segment with instability and myelopathy is one of the most common causes of pain and neurological deterioration requiring re-operation following cervical fusion surgery. KEY POINTS/CONCLUSIONS:• A real-time kinematic 2D sagittal HASTE flexion-extension sequence can be used to assess for dynamic listhesis, cervical cord, contact and deformity. • The additional kinematic cine sequence was well tolerated and the mean acquisition time for the 2D HASTE sequence was 42 s (range 31-44 s). • A sagittal flexion-extension HASTE cine sequence provides substantial agreement between readers for dynamic cord deformity and moderate agreement between radiographs and MRI for dynamic listhesis.

Ultrasound visualization affords proceduralists versatile and accurate guidance for a variety of percutaneous, minimally invasive procedures in the musculoskeletal system including joint (intra-articular) injections or aspirations, intra-bursal injections, peritendinous, and perineural injections. A variety of percutaneous procedures are traditionally performed blindly, but may be more easily or more accurately performed with the real-time assistance of ultrasound guidance. Other procedures are only possible utilizing image-guidance, due to the required precision of the injection because of delicate local anatomy or depth of the injection; ultrasound is a safe, portable, and widespread modality that can be used to assist the proceduralist in localizing the needle tip in such cases, to ensure safe and accurate delivery of the medication, most frequently a solution of steroid and anesthetic. This review aims to provide a foundational approach to ultrasound-guided procedures in the musculoskeletal system, offering tips and tricks that can be employed in many different procedures including intra-articular, juxta-articular, and perineural injections for a multitude of clinical scenarios. Technical considerations regarding ultrasound transducer selection, sonographic technique, as well as common indications, contraindications, and complications of these procedures, are presented. Additionally, a variety of pharmacologic considerations for proceduralists contemplating ultrasound-guided injections are discussed.

BACKGROUND:Detection of rotator cuff tears, a common cause of shoulder disability, can be time-consuming and subject to reader variability. Deep learning (DL) has the potential to increase radiologist accuracy and consistency. PURPOSE:The aim of this study was to develop a prototype DL model for detection and classification of rotator cuff tears on shoulder magnetic resonance imaging into no tear, partial-thickness tear, or full-thickness tear. MATERIALS AND METHODS:This Health Insurance Portability and Accountability Act-compliant, institutional review board-approved study included a total of 11,925 noncontrast shoulder magnetic resonance imaging scans from 2 institutions, with 11,405 for development and 520 dedicated for final testing. A DL ensemble algorithm was developed that used 4 series as input from each examination: fluid-sensitive sequences in 3 planes and a sagittal oblique T1-weighted sequence. Radiology reports served as ground truth for training with categories of no tear, partial tear, or full-thickness tear. A multireader study was conducted for the test set ground truth, which was determined by the majority vote of 3 readers per case. The ensemble comprised 4 parallel 3D ResNet50 convolutional neural network architectures trained via transfer learning and then adapted to the targeted domain. The final tear-type prediction was determined as the class with the highest probability, after averaging the class probabilities of the 4 individual models. RESULTS:The AUC overall for supraspinatus, infraspinatus, and subscapularis tendon tears was 0.93, 0.89, and 0.90, respectively. The model performed best for full-thickness supraspinatus, infraspinatus, and subscapularis tears with AUCs of 0.98, 0.99, and 0.95, respectively. Multisequence input demonstrated higher AUCs than single-sequence input for infraspinatus and subscapularis tendon tears, whereas coronal oblique fluid-sensitive and multisequence input showed similar AUCs for supraspinatus tendon tears. Model accuracy for tear types and overall accuracy were similar to that of the clinical readers. CONCLUSIONS:Deep learning diagnosis of rotator cuff tears is feasible with excellent diagnostic performance, particularly for full-thickness tears, with model accuracy similar to subspecialty-trained musculoskeletal radiologists.

OBJECTIVE:To evaluate imaging features of soft tissue metastases, technical factors associated with diagnostic yield of image-guided biopsy, and clinical impact of biopsy results on patient outcomes. MATERIALS AND METHODS/METHODS:A total of 1605 image-guided soft tissue biopsies were retrospectively identified from December 2010 to December 2020. Included lesions were histologically proven musculoskeletal soft tissue metastases. Lesions were excluded if intraabdominal, intrathoracic, retroperitoneal, associated with osseous lesions or surgical scar implants or arising from skin or lymph nodes. Image guidance modality, needle size, number of cores, and lesion location, size, and depth from skin were recorded. Patient demographics, malignancy history, biopsy-driven changes in management, and survival rate after biopsy were collected. RESULTS:Forty-six patients met the inclusion criteria with a biopsy diagnostic yield of 44/46 (95.7%). Metastases were most commonly located truncal (82.6%, p < 0.001) and intramuscular (78.3%, p < 0.001). A total of 37/46 (80.4%) biopsies were US-guided. And 9/46 (19.6%) were CT-guided. There was no significant difference in the number of cores or mean needle gauge between diagnostic and nondiagnostic biopsies. At time of review, 23 (50%) patients were deceased, with a mean survival of 13.5 months after biopsy. The majority (71.7%) of patients had a known primary malignancy at time of biopsy, most commonly lung (24.2%) and breast (24.2%). Overall survival showed no association with anatomic location (p > 0.83) or tissue type (p > 0.34). The most common biopsy-driven outcome was initiation of chemotherapy, immunotherapy, and/or radiotherapy (52.2%). CONCLUSION/CONCLUSIONS:Image-guided biopsy for soft tissue metastases has high diagnostic yield and commonly influences clinical management. Metastases were most commonly intramuscular in the trunk and are associated with poor prognosis.