Clinical Applications, Challenges, and Recommendations for Artificial Intelligence in Musculoskeletal and Soft-Tissue Ultrasound: AJR Expert Panel Narrative Review.

Artificial intelligence (AI) is increasingly used in clinical practice for musculoskeletal imaging tasks, such as disease diagnosis and image reconstruction. AI applications in musculoskeletal imaging have focused primarily on radiography, CT, and MRI. Although musculoskeletal ultrasound stands to benefit from AI in similar ways, such applications have been relatively underdeveloped. In comparison with other modalities, ultrasound has unique advantages and disadvantages that must be considered in AI algorithm development and clinical translation. Challenges in developing AI for musculoskeletal ultrasound involve both clinical aspects of image acquisition and practical limitations in image processing and annotation. Solutions from other radiology subspecialties (e.g., crowdsourced annotations coordinated by professional societies), along with use cases (most commonly rotator cuff tendon tears and palpable soft-tissue masses), can be applied to musculoskeletal ultrasound to help develop AI. To facilitate creation of high-quality imaging datasets for AI model development, technologists and radiologists should focus on increasing uniformity in musculoskeletal ultrasound performance and increasing annotations of images for specific anatomic regions. This Expert Panel Narrative Review summarizes available evidence regarding AI's potential utility in musculoskeletal ultrasound and challenges facing its development. Recommendations for future AI advancement and clinical translation in musculoskeletal ultrasound are discussed.

Diagnostic and Interventional US of the Wrist and Hand: Quadrant-based Approach.

Diagnostic and interventional US of the wrist and hand can be challenging due to the small size and superficial location of structures and various disorders that need to be considered. A quadrant-based approach (volar, ulnar, dorsal, and radial) provides a rational method for performing a focused examination and joint positioning during both diagnostic imaging and intervention. Volar wrist disorders primarily involve the median nerve and the digital flexor system comprised of the flexor tendons and pulleys. The ulnar nerve and extensor carpi ulnaris tendon are chiefly responsible for ulnar-sided wrist pain. The differential diagnosis for dorsal-sided symptoms typically involves the extensor tendon compartments and includes distal intersection syndrome, extensor pollicis longus tear, and digital extensor apparatus injury. The soft-tissue ganglion is the most common abnormality in the dorsal wrist, typically associated with wrist ligaments or joint capsule. Radial-sided pain may be secondary to de Quervain tenosynovitis and must be differentiated from the more proximal intersection syndrome. US is an important tool for assessing the ulnar collateral ligament of the first metacarpophalangeal joint of the thumb and differentiating between displaced and nondisplaced tears, thereby influencing management. Despite the complexity of the anatomy and potential pathologic features within the wrist, a focused quadrant-based examination can permit the sonologist to focus on the structures of relevance. In conjunction with a systematic approach, this can aid in precise and efficient diagnostic scanning and intervention of the wrist and hand. ©RSNA, 2023 Supplemental material is available for this article. Quiz questions for this article are available through the Online Learning Center.

Dynamic musculoskeletal ultrasound: slipping rib, muscle hernia, snapping hip, and peroneal tendon pathology.

Dynamic musculoskeletal ultrasound is an important diagnostic tool that allows the practitioner to observe soft tissue structures over a range of motion and identify pathology not diagnosed on other modalities. Familiarity with this modality allows health care practitioners to appropriately refer patients for this type of examination. This article will review several indications for dynamic ultrasound imaging, including slipping rib, muscle hernia, snapping hip, and peroneal tendon pathology. The examination technique and expected findings for common pathology in each location are discussed.

Ultrasonography of Superficial Soft-Tissue Masses: Society of Radiologists in Ultrasound Consensus Conference Statement.

The Society of Radiologists in Ultrasound convened a panel of specialists from radiology, orthopedic surgery, and pathology to arrive at a consensus regarding the management of superficial soft-tissue masses imaged with US. The recommendations in this statement are based on analysis of current literature and common practice strategies. This statement reviews and illustrates the US features of common superficial soft-tissue lesions that may manifest as a soft-tissue mass and suggests guidelines for subsequent management.

Imaging Characteristics of the Proximal Lateral Collateral Ligament of the Knee: Findings on Ultrasound and MRI With Histologic Correlation.

OBJECTIVES: Determine prevalence of increased signal intensity of the lateral collateral ligament (LCL) of the knee on MRI and decreased echogenicity on ultrasound, and compare with cadaveric histologic evaluation. METHODS: After IRB approval of this prospective study with informed consent, patients having knee MRI were additionally evaluated with ultrasound. Signal intensities of LCL on MRI (low, intermediate, high), echogenicity at ultrasound (hyperechoic, hypoechoic, anechoic), and extent of findings were assessed. Descriptive statistics, Wilcoxon signed ranked test, and intraclass correlation coefficient (ICC) were calculated. Two cadaveric knees were imaged with MRI and ultrasound, including histologic LCL evaluation. RESULTS: Seventy-three subjects were included (39 males, 34 females; mean age 48 ± 14 years) with 77 knee examinations. On MRI, low, intermediate, and high signals were present in 21% (16/77), 75% (58/77), and 4% (3/77), respectively. On ultrasound, echogenicity was assessed as hyperechoic, hypoechoic, and anechoic in 62% (48/77), 38% (29/77), and 0% (0/77), respectively. Mean length of increased signal was 8.6 mm (±4.9) on MRI, and 6.5 mm (±4.8) on ultrasound. The ICC showed a good to excellent intermodality reliability (0.735-0.899) without statistically significant difference for interreader measurements (P = .163-.795). Histology evaluation showed transition of ligament fibers to fibrocartilage at its insertion with increased connective tissue mucin corresponding to MRI and ultrasound findings. CONCLUSIONS: Increased signal intensity of the proximal LCL on ultrasound and MRI is common and corresponds to normal connective tissue mucin.

Medial Meniscal Extrusion Evaluation With Weight-Bearing Ultrasound: Correlation With MR Imaging Findings and Reported Symptoms: Correlation With MR Imaging Findings and Reported Symptoms.

OBJECTIVES: To compare medial meniscal extrusion on weight-bearing ultrasound (US) with supine US and magnetic resonance (MR) imaging correlating with meniscal pathology and reported symptoms. METHODS: IRB approved study with informed consent. Patients obtaining routine knee MR imaging for suspected knee pathology were prospectively evaluated with supine and weight-bearing US of the medial meniscus. Meniscal extrusion was measured independently by two fellowship-trained musculoskeletal radiologists. Correlation was made to presence or absence of meniscal degeneration or tear on MR imaging, as well as reported symptoms. Statistical significance was calculated via intraclass correlation coefficient (ICC) and analysis of variance (ANOVA). RESULTS: Ninety-nine knees from 95 subjects (50 males, 45 females; mean age 45 ± 15 years) were included. Mean medial meniscal extrusion measured at US for a normal meniscus (n = 36) was 0.8 mm when supine, increasing to 1.6 mm on weight-bearing. Mean meniscal extrusion in subjects with mucoid degeneration (n = 20) and those with meniscal tears (n = 43) was 1.6 mm, increasing to 2.3 mm with weight bearing. Inter-reader reliability showed ICC values of 0.853 to 0.940. There was a significant difference in medial meniscal extrusion comparing subjects with a normal medial meniscus at magnetic resonance imaging (MRI) and subjects with either meniscal degeneration or tear. There was no significant difference in degree of meniscal extrusion between subjects with meniscal degeneration or tear. There was trend of worsening symptoms and increasing functional limitations moving from normal meniscus to meniscal degeneration to meniscal tear. CONCLUSIONS: A normal meniscus shows lesser mobility between supine and upright position, than a pathologic meniscus. Both mucoid degeneration and meniscal tear demonstrate extrusion in the supine position, which increases with weight-bearing position.

Recommended Musculoskeletal and Sports Ultrasound Terminology: A Delphi-Based Consensus Statement.

OBJECTIVES: The current lack of agreement regarding standardized terminology in musculoskeletal and sports ultrasound presents challenges in education, clinical practice, and research. This consensus was developed to provide a reference to improve clarity and consistency in communication. METHODS: A multidisciplinary expert panel was convened consisting of 18 members representing multiple specialty societies identified as key stakeholders in musculoskeletal and sports ultrasound. A Delphi process was used to reach consensus which was defined as group level agreement >80%. RESULTS: Content was organized into seven general topics including: 1) General Definitions, 2) Equipment and Transducer Manipulation, 3) Anatomic and Descriptive Terminology, 4) Pathology, 5) Procedural Terminology, 6) Image Labeling, and 7) Documentation. Terms and definitions which reached consensus agreement are presented herein. CONCLUSIONS: The historic use of multiple similar terms in the absence of precise definitions has led to confusion when conveying information between colleagues, patients, and third-party payers. This multidisciplinary expert consensus addresses multiple areas of variability in diagnostic ultrasound imaging and ultrasound-guided procedures related to musculoskeletal and sports medicine.

Recommended musculoskeletal and sports ultrasound terminology: a Delphi-based consensus statement.

The current lack of agreement regarding standardised terminology in musculoskeletal and sports ultrasound presents challenges in education, clinical practice and research. This consensus was developed to provide a reference to improve clarity and consistency in communication. A multidisciplinary expert panel was convened consisting of 18 members representing multiple specialty societies identified as key stakeholders in musculoskeletal and sports ultrasound. A Delphi process was used to reach consensus, which was defined as group level agreement of >80%. Content was organised into seven general topics including: (1) general definitions, (2) equipment and transducer manipulation, (3) anatomical and descriptive terminology, (4) pathology, (5) procedural terminology, (6) image labelling and (7) documentation. Terms and definitions which reached consensus agreement are presented herein. The historic use of multiple similar terms in the absence of precise definitions has led to confusion when conveying information between colleagues, patients and third-party payers. This multidisciplinary expert consensus addresses multiple areas of variability in diagnostic ultrasound imaging and ultrasound-guided procedures related to musculoskeletal and sports medicine.

MR imaging of the quadriceps femoris tendon: distal tear characterization and clinical significance of rupture types.

OBJECTIVE: To characterize quadriceps femoris tendon tears on magnetic resonance (MR) imaging regarding tear extent, location, and presence of bony avulsion. MATERIALS AND METHODS: IRB approval was obtained and informed consent was waived for this retrospective case series. Electronic medical records from all patients in our hospital system were searched for keywords: knee MR imaging, and quadriceps tendon rupture or tear. MRI studies were randomized and independently evaluated by two fellowship-trained musculoskeletal radiologists. MR imaging was used to characterize each individual quadriceps tendon as having tendinosis, tear (location, partial versus complete, size, and retraction distance), and bony avulsion. Knee radiographs were reviewed for presence or absence of bony avulsion. Descriptive statistics and inter-reader reliability (Cohen's Kappa and Wilcoxon-signed-rank test) were calculated. RESULTS: Fifty-two patients with 53 quadriceps tears were evaluated (45 males, 7 females; mean age: 51 ± 13 years). The vastus intermedius (VI) tendon more often incurred a partial rather than a complete tear (39.6% vs. 37.7%), while the rectus femoris (RF), vastus medialis (VM), and vastus lateralis (VL) incurred complete tears more commonly (64.2-66%). Subjects with bony avulsion on radiographs had higher-grade tears of the RF, VM, and VL tears (p = 0.020-0.043) but not the VI. Most tendons tore at or immediately proximal to the patella (84.8-93.6%). Gaps in retracted torn tendons measured between 2.3 and 2.7 cm. Inter-reader reliability was substantial to almost perfect (κ = 0.624-0.953). CONCLUSION: Quadriceps femoris tendon tears most commonly involve the RF or VL/VM layers usually in proximity to the patella. A bony avulsion correlates with a more extensive tear. KEY POINTS: • Quadriceps femoris tendon tears most commonly involve the rectus femoris or vastus lateralis/vastus medialis layers. • A rupture of the quadriceps femoris tendon usually occurs in proximity to the patella. • A bony avulsion of the patella correlates with a more extensive tear of the superficial and middle layers of the quadriceps tendon.

Computed Tomography Evidence of Psoas Muscle Atrophy Without Concomitant Tendon Wasting in Early Sepsis.

BACKGROUND: Morphomic studies have demonstrated a correlation between sarcopenia and clinical outcomes in septic patients. However, tendon morphomics has not yet been studied in this context. The purpose of the present study was to evaluate tendon morphology in septic patients through analytic morphomics. We hypothesized that morphomic analyses would reveal concomitant muscle and tendon wasting in sepsis patients. The results of this study may help to implement different rehabilitation modalities for critically ill patients. MATERIALS AND METHODS: The volume and fat content of bilateral psoas muscles and tendons were measured on abdominal computed tomography scans of 25 ICU septic and 25 control trauma patients admitted to the University of Michigan between 2011 and 2012. Univariate and multivariate analyses were performed to determine the relationship between psoas muscle and tendon morphometric data, and the association with clinical variables such as smoking and comorbidities. RESULTS: Average psoas muscle volume was 12.21 ± 5.6 cm3 for control patients and 9.318 ± 3.3 cm3 in septic patients (P = 0.0023). The average psoas muscle/fat ratio for septic patients was 0.0288 ± 0.071 cm3, compared with 0.0107 ± 0.008 cm3 in the control group (P = 0.075). Average tendon volume in the septic population (0.508 ± 0.191 cm3) was not different than the control cohort (0.493 ± 0.182 cm3) (P = 0.692). CONCLUSIONS: Our results demonstrate significantly smaller psoas muscle volume in septic patients than in age-, gender-, and BMI-matched trauma patients but no demonstrable change in tendon morphology between patient groups. These findings begin to define the boundaries of clinical application within the field of morphomics.

Normal and Injured Ankle Ligaments on Ultrasonography With Magnetic Resonance Imaging Correlation.

Ultrasonography (US) has been increasingly used in the evaluation of ankle ligamentous injuries given its advantages as a dynamic, efficient, noninvasive, and cost-effective imaging method. Understanding the anatomy of the ankle ligaments is critical for correct diagnosis and treatment. This pictorial essay describes and illustrates the US scanning technique and potential pitfalls in evaluating the ankle ligaments and also provides an overview of the US appearance of normal and injured ankle ligaments with magnetic resonance imaging correlation. Highlighted structures include the lateral complex, medial/deltoid complex, spring (calcaneonavicular) ligament complex, and syndesmosis.

Ultrasound Features of Palmar Fibromatosis or Dupuytren Contracture.

OBJECTIVES: To retrospectively characterize the ultrasound appearance of palmar fibromatosis in patients with a surgical or clinical diagnosis of palmar fibromatosis. METHODS: A search of ultrasound reports from 2005 to 2015 and a subsequent search of medical records were performed to identify patients with a surgical or clinical diagnosis of palmar fibromatosis. The ultrasound images were retrospectively reviewed to record the lesion location, size, echogenicity, compressibility, hyperemia, and calcification. RESULTS: A total of 36 patients were identified (average age, 60 years; 61% male), yielding a total of 55 palmar fibromatosis lesions, of which 2%, 7%, 29%, 36%, 20%, and 5% were located at the first, second, third, fourth, and fifth digits and between the fourth and fifth digits, respectively. The lesions were located directly superficial to the flexor tendons in 93% with their epicenters at the distal metacarpal in 89%. Average lesion dimensions were 13.1 mm in length, 6.8 mm in width, and 2.5 mm in depth. On ultrasound images, the lesions were characteristically hypoechoic (98%) and noncompressible (95%). Atypical features included calcification (2%), compressibility (5%), hyperemia on color Doppler images (6%), epicenters at the metacarpophalangeal joint (7%) or proximal phalanx (4%), and location superficial but lateral to the flexor tendons (7%). CONCLUSIONS: Palmar fibromatosis most commonly appears hypoechoic and is located directly superficial to the flexor tendons with an epicenter at the distal metacarpal, most commonly the fourth digit. However, the epicenter location may be at the distal metacarpal and proximal phalanx of other digits, adjacent to the flexor tendons, with possible hyperemia and calcification.

Sonography of Wrist Ganglion Cysts: Which Location Is Most Common?

OBJECTIVES: The literature states that wrist ganglion cysts are most commonly dorsal; however, our experience suggests a volar location is more common. The purpose of this study was to identify the locations of ganglion cysts of the wrist as imaged with sonography. METHODS: After Institutional Review Board approval with informed consent waived, retrospective review of wrist sonography reports from January to April 2016 was completed. Only patients who had a comprehensive evaluation that included the dorsal and volar wrist were included. Ultrasound images were reviewed to characterize wrist ganglion cysts. RESULTS: The study group consisted of 98 subjects (78% female, 22% male) (median age, 51 years; range 13-79) with 124 wrist ganglion cysts, where 69% (86 of 124) were volar and 31% (38 of 124) were dorsal. Ganglion cysts were located between the radial artery and flexor carpi radialis in 63% (78 of 124), followed by a dorsal location superficial to the scapholunate ligament in 20% (25 of 124), other dorsal locations in 11% (13 of 124), and other volar locations in 6% (8 of 124). With dorsal ganglion cysts, the scapholunate ligament when imaged appeared normal in 91% (20 of 22). Ganglion cysts were multilocular/multilobular in all subjects. CONCLUSIONS: The most common location for wrist ganglion cysts is in the volar wrist, particularly between the radial artery and flexor carpi radialis tendon, appearing multilocular or multilobular with a mean largest dimension of 1.0 cm. Given the propensity of ganglion cysts to occur at this site, attention to this area when imaging the wrist with sonography or magnetic resonance imaging should be considered.

VIDEO: Ultrasound of the Elbow: Scanning Techniques.

OBJECTIVE: When imaging the elbow with ultrasound, it is important to follow a routine that includes a checklist of anatomic structures. CONCLUSION: This video will show ultrasound scanning and anatomy of the elbow divided into four quadrants, beginning anterior, then moving medial, lateral, and posterior.

Multimodality Imaging Review of Normal Appearance and Complications of the Postoperative Rotator Cuff.

OBJECTIVE: The objective of this article is to provide a multimodality imaging review of the imaging features of the postoperative rotator cuff, including a discussion of the most common complications. CONCLUSION: Both MRI and ultrasound can be used to evaluate the repaired rotator cuff and potential complications. The appearance of the rotator cuff within the first 6 months after repair is heterogeneous and may be difficult to interpret but normalizes over time.

Imaging appearance of well-differentiated liposarcomas with myxoid stroma.

OBJECTIVE: Describe the imaging appearance of well-differentiated liposarcoma with myxoid stroma (WDLMS) and correlate with histopathology. MATERIALS AND METHODS: A keyword search of the institution medical records was performed from 1 January 2000 to 30 June 2017. The histopathology slides of cases identified in this fashion were then reviewed by a pathologist. Additional cases were prospectively collected from extramural referrals and tumor boards. Diagnostic imaging studies of pathologically proven cases of WDLMS were then reviewed in consensus and correlated with pathology. RESULTS: Ten cases of pathologically proven WDLMS were identified (7 men, 3 women, ages 26-81). Tumor location included the retroperitoneum (n = 5), thigh (n = 4), and the shin (n = 1). Nine patients had macroscopic fat on imaging. The nonlipomatous components had a variable appearance, including septal, nodular, and lacelike patterns. Two cases included two distinct areas that were predominantly myxoid or lipomatous ("bi-morphic"). One tumor had no macroscopic fat on imaging. On CT, the nonlipomatous nodular components were hypodense/had hypodense areas. On MRI, the nodular components had intermediate/bright T2W signal. Interval nonlipomatous nodular growth was identified in 3 cases. CONCLUSION: WDLMS may present on imaging as a mass with variable morphology and amounts of nonlipomatous components. Histopathological diagnosis of WDLMS is challenging and imaging correlation may be helpful, as this tumor may have ≥50% fatty volume, may have a myxoid nodular component or bi-morphic appearance, or may be located in the retroperitoneum, features that are unusual for myxoid liposarcoma. WDLMS with a nodular component cannot be distinguished from dedifferentiated liposarcoma based on imaging alone.

Ultrasound of the Knee: Common Pathology Excluding Extensor Mechanism.

Ultrasound can be used to diagnose many types of pathology and guide various diagnostic or therapeutic procedures. The most common applications for musculoskeletal ultrasound of the knee include evaluation of the joint recesses, the extensor mechanism, and bursae including Baker's cyst. However, a role exists for ultrasound in the evaluation of cartilage, other tendons, ligaments, peripheral nerves, osseous structures, and vasculature. This article reviews common pathologies of these structures as seen with ultrasound. In addition, various ultrasound-guided procedures pertaining to these pathologies are introduced. The importance of a comprehensive evaluation is also emphasized as a method to maximize information gained from the ultrasound examination.

Ultrasound evaluation of bursae: anatomy and pathological appearances.

A bursa is an extra-articular sac that may communicate with a joint and functions to decrease friction between tendons and either bone or skin. Bursae can be classified as native and non-native (adventitious) bursae. The native bursae are lined with a synovial membrane and occur at predictable anatomical sites; knowledge of these normal structures can help distinguish them from other pathological entities. An adventitious bursa can form at sites of friction rather than at predictable anatomical sites, but otherwise have imaging features similar to native bursae. Bursal distention can occur from many pathological processes, most commonly resulting from chronic overuse injury. When imaging bursal pathological conditions, there is often an overlap of imaging findings, regardless of the cause. In general, ultrasound of a distended bursa reveals a fluid collection with either simple anechoic or more complex hypoechoic fluid. Bursal distention is characteristically unilocular and compressible, unlike other structures such as ganglion cysts, which are usually multilocular and non-compressible. This article reviews the anatomical locations of common bursae and shows pathological examples using ultrasound. Knowledge of typical locations and imaging appearances of bursae can aid in narrowing the differential diagnosis and guiding further management and treatment decisions.

Ultrasound of the Gruberi Bursa With Cadaveric and MRI Correlation.

OBJECTIVE: The purpose of this study was to describe the appearance and ultrasound characteristics of the Gruberi bursa using a cadaveric model and retrospective ultrasound imaging review. MATERIALS AND METHODS: For the cadaveric study, ultrasound of the dorsolateral ankle of a foot-ankle specimen was performed and was followed by injection of latex between the extensor digitorum longus (EDL) tendons and the talus and dissection. For the ultrasound imaging review, the radiology database was searched for ultrasound studies performed from September 15, 2000, through April 1, 2015, to identify subjects with a dorsolateral foot or ankle fluid collection detected on ultrasound. Images were retrospectively reviewed to characterize the location and size of the fluid collection, assess for the number of locules, and evaluate the compressibility of the fluid collection. It was determined whether the ultrasound findings were significantly different from chance: CI and p values were obtained from performing a test for one proportion. RESULTS: Dissection of a cadaveric specimen revealed latex within a well-defined region between the EDL tendons and the dorsolateral talus; this location is consistent with a Gruberi bursa. For the image review, the imaging examinations of 162 subjects (age range, 16-88 years; 31 male subjects and 131 female subjects) were reviewed. On the ultrasound images, a fluid collection with its epicenter between the dorsolateral talus and EDL was found in 93% of ankles. Of the fluid collections identified on ultrasound, 98% were unilocular and 94% were anechoic. Of these fluid collections, 133 were assessed for compressibility, and 89% were compressible. The positive findings for a Gruberi bursa that were different from chance (p < 0.0001) were a fluid collection being located between the EDL tendons and the dorsolateral talus and being unilocular, anechoic, and compressible. CONCLUSION: The Gruberi bursa characteristically is located between the EDL and the talus; on ultrasound, the Gruberi bursa is most commonly unilocular, anechoic, and compressible.

US of the Knee: Scanning Techniques, Pitfalls, and Pathologic Conditions.

Pain and other disorders of the knee are a common presenting complaint in the ambulatory setting. Although the cornerstones of imaging evaluation of the knee are radiographs and magnetic resonance (MR) imaging, ultrasonography (US) is less expensive than MR imaging, easily available, and of comparable accuracy in the evaluation of certain pathologic conditions of the knee. The benefits of US include portability, low cost, high spatial resolution, dynamic imaging, and ability to guide percutaneous interventions when indicated. US also allows direct patient contact, facilitating immediate clinical correlation and the ability to compare with the contralateral knee. US evaluation of the knee can be targeted to a specific region on the basis of the complaint or be a comprehensive review. For comprehensive evaluation, the knee is divided into anterior, medial, lateral, and posterior compartments for structured evaluation of the tendons, ligaments, joint space, osseous structures, as well as peripheral nerves and vasculature. US is particularly well suited for evaluating injuries of the quadriceps and patellar tendons, injuries of the medial and lateral collateral ligaments, joint effusions, and fluid collections around the knee. There is additional utility in evaluation of the distal hamstrings tendons, the iliotibial tract, the superficial patellar cortex, the common peroneal nerve, the popliteal vessels, and juxta-articular cystic collections including Baker cyst. In-depth appreciation of relevant sonographic anatomy, common pathologic conditions, knowledge of important pitfalls, and mastery of US technique will allow one to effectively use this powerful bedside tool for the evaluation of a wide variety of knee disorders. ©RSNA, 2016.