XR remains the most readily available imaging modality for detecting and monitoring structural damage and growth abnormalities, but given the current emphasis on early intervention, the detection of pre-erosive joint changes has become a priority.

However, in cases of clinical uncertainty, XR plays a crucial role, excluding other differential diagnostic options such as traumatic or orthopedic diseases (fractures, osteochondral lesions), tumors, and infectious causes (osteomyelitis) [4, 6].

In 2018, the Task Force of the French Societies of Rheumatology, Radiology, and Paediatric Rheumatology, focusing on XR, attempted for the first time to provide pragmatic guidelines for daily practice specific for each non-systemic JIA subtype and for situations of particular interest [43], summarized in Table 2.

Interestingly, Weiss et al. [44] recently provided the first consensus-derived radiographic definition of sacroiliitis in skeletally immature adolescents as a criterion for classifying axial disease in juvenile spondyloarthritis when MRI is not available. Nevertheless, the use of XR in the diagnosis of sacroiliitis is discouraged.

In recent decades, new radiological scoring systems have been developed and adult radiological scores have been adapted for use in JIA. Their application in non-controlled JIA clinical trials has demonstrated that standardized assessment of radiological progression is feasible. This has led to the suggestion that semiquantitative measurement of radiological damage should also be considered when evaluating treatment efficacy in JIA [45,46,47].

In advanced stages of disease, XR allows visualization of late complications (erosions, ankylosis, subluxation or joint malalignment, enlarged epiphysis, premature growth plate fusion leading to limb length inequality, spinal deformities, muscle atrophy) (Fig. 3) [48,49,50,51].

Anteroposterior radiographs of both hands in a 19-year-old young woman with a long history of aggressive seronegative polyarticular juvenile idiopathic arthritis, onset at the age of 7 years and with poor adherence to treatment. Images show mild bilateral periarticular osteoporosis (long arrows), growth arrest lines in the radial metaphysis and bone-in-bone appearance of carpal bones (short arrows), joint space narrowing of carpo-metacarpal joints and intercarpal bones, more pronounced on the right (arrowheads), and mild soft tissue edema around the right ulna (thin arrow)

In addition, XR has a historical role in assessing bone maturity and detecting bone age delay or progression, which in JIA may also help to distinguish where disease control is suboptimal or whether other factors are influencing growth retardation [52].

In summary, XR is a useful method for differential diagnosis in doubtful cases, evaluation of structural and morphological changes before diagnosis (indication based on clinical examination), and evaluation and monitoring of joint destruction and growth disorders.

Musculoskeletal US represents an easily accessible, clinically relevant routine examination in children with JIA.

Before US can be established as a valuable imaging modality, two significant challenges must be overcome: understanding age-related normal findings and standardizing the US protocol for different joints.

Numerous papers, including the 2018 OMERACT US in Paediatrics Working Group, have described the standardization of US examination for different joints, including physiological intra-articular vascularization, patient and joint position, and transducer placement for each examination approach [53,54,55,56,57,58].

Although these reports provide information to help differentiate between normal and pathological findings of joints in children, they currently serve as baseline information. Assessment of changes in US characteristics within an individual subject over time could potentially be more informative than a simple comparison with a cutoff value [56]. Another issue is the minimum and optimal set of joints that should be scanned for routine musculoskeletal US surveillance. Scanning all accessible joints is not feasible in routine practice, and different studies have tested different numbers of joints.

The reduced model with ten joints by Collado et al. [59] showed higher responsiveness to changes than the evaluation of a larger number of joints. Overall, these results suggest that an US assessment which focuses on a reduced number of joints and includes the sites that are most commonly affected in JIA may satisfactorily provide information about the overall burden of disease activity [59].

Joints most suitable for musculoskeletal US examination are the ankle, knee, hip, wrist, and small joints of the hands and feet (Figs. 4 and 5) [18]. Apart from the anatomical consideration and the challenge of clinical assessment of these joints, Magni-Manzoni et al. [60] have demonstrated a higher incidence of subclinical synovitis in the wrists, proximal interphalangeal (PIP) joints, subtalar joints, and ankles.

Power Doppler ultrasound of left talonavicular joint in sagittal projection in a 19-year-old young woman with a long history of aggressive seronegative polyarticular juvenile idiopathic arthritis, onset at the age of 7 years and with poor adherence to treatment (same patient as in Fig. 3) shows thickened and chronically altered synovium without a joint effusion or significant hyperemia (arrow)

Ankle ultrasound of a 2-year-old girl with juvenile idiopathic arthritis (HLA B27 negative, anti-nuclear antibodies positive). a, b Sagittal projection of right (a) and left (b) ankle with effusion in the right side (arrow); (c) Sagittal power Doppler study of right ankle with effusion (long arrow), and thickened and hyperemic synovium (short arrow)—signs of synovitis

Although the role of US in assessing the axial skeleton remains limited, a recent study by Falsetti et al. [61] suggested the potential role of power Doppler ultrasound with spectral wave analysis as a screening method in children with suspected juvenile spondyloarthritis. They found higher power Doppler US scores at the SIJs in patients with a confirmed diagnosis of juvenile spondyloarthritis based on MRI diagnosis [61].

The most important clinical contribution of US is the identification and differentiation of synovitis, tenosynovitis, bursitis, and enthesitis.

In several studies, US has been shown to be superior to physical examination in the diagnosis of synovitis [26, 28, 37, 60, 62,63,64,65,66,67,68]. Nevertheless, it remains questionable whether physical examination or US can be more accurate in detecting joint inflammation. Indeed, confidence and competence in musculoskeletal examinations may be low, particularly in pediatrics. A systematic literature review on the assessment of synovitis in JIA by Collado et al. [69] on the assessment of synovitis in JIA highlighted key issues, such as small sample size, lack of MRI comparison, technical difficulties, and lack of a control score.

Enthesitis is the main feature of the JIA subgroup of enthesitis-related arthritis. Clinical recognition of enthesitis in children is challenging due to the particular distribution of fat, which can obscure anatomical landmarks, and the often-inadequate cooperation of very young children.

A number of studies confirm a higher sensitivity of musculoskeletal US compared to clinical examination in detecting enthesitis [70,71,72,73]. On the other hand, a recent systematic review indicated that the existing evidence suggests that there is no standardized US definition of enthesitis in children, and that discriminant validity has not been demonstrated [74].

The additional role of US is to monitor the response to treatment and the disease course. In this context, the importance of power Doppler US in addition to conventional US and the use of a standardized US scoring system seems crucial.

US-guided procedures are the next important application of US in JIA. They allow precise localization of inflammation and accurate needle placement in clinically difficult-to-assess or hard-to-reach sites, such as wrists, TMJs, hip, small joints of hands and feet, ankles, and tendons. This maximizes treatment efficacy while minimizing local side effects, such as subcutaneous atrophy or localized skin hypopigmentation [75,76,77,78,79]. Due to limited and conflicting data, certain critical aspects require further investigation, particularly for specific sites such as the TMJs [80].

In summary, musculoskeletal US is useful in daily practice to assess the presence and degree of inflammation in areas that are more difficult to assess clinically, such as the wrist, ankle, and foot joints. It can be used to guide intra-articular injections. Ideally, it should be performed by an experienced radiologist or rheumatologist.

MRI is the most promising imaging technique for assessing the presence and extent of inflammation (synovial hypertrophy, joint effusion, soft tissue swelling), bone marrow changes, and cartilage status (Figs. 6 and 7). Furthermore, MRI can serve as a diagnostic tool for certain intra-articular pathologies that mimic JIA [81, 82].

Magnetic resonance imaging of the right ankle in a 2-year-old girl with juvenile idiopathic arthritis (HLA B27 negative, anti-nuclear antibodies positive, same patient as in Fig. 5). a Sagittal proton-density-weighted fat suppressed image shows marked synovial proliferation and effusion of the anterior and posterior recesses of the ankle joint (long arrows) and around the extensor tendons (short arrow)—signs of synovitis and tenosynovitis. b Sagittal postcontrast T1-weighted fat-suppressed sequence shows marked synovial proliferation and enhancement of the anterior recess of the ankle joint (long arrow) and around the tarsal bones (short arrow)—signs of active synovitis. c Axial postcontrast T1-weighted fat-suppressed sequence shows marked synovial proliferation and enhancement of the anterior and posterior recesses of the ankle joint (long arrows) and around the flexor tendons (short arrow)—signs of active synovitis and tenosynovitis

Magnetic resonance imaging of the right wrist in a 14-year-old boy with juvenile idiopathic arthritis with enthesitis (HLA B27 positive). Coronal (a) and axial (b) proton-density-weighted fat-suppressed images show marked synovial proliferation (long arrows), reactive edema of the carpal bones (short arrows), and a small joint effusion (arrowhead in b)—signs of synovitis

As with all imaging modalities, the main criticisms of using MRI include the challenge of differentiating between pathological and physiological changes in bone marrow depending on the age and sex of the patient [83,84,85,86,87,88,89,90,91]. There is also a need for standardization, quantification, and validation of scoring systems to rigorously and consistently assess joint changes, in both cross-sectional and longitudinal studies. It is worth noting that efforts are currently underway to achieve this standardization [92].

The importance of MRI for monitoring inflammation and response to treatment has been confirmed. In 2015, the EULAR-PRES task force [6] and the European Society of Musculoskeletal Radiology (ESSR) Arthritis Subcommittee [93] published the indications for performing MRI for diagnosis, monitoring, and prediction, as well as MRI protocols for the most commonly affected joints in JIA; these have recently been updated by the ESSR and the European Society of Paediatric Radiology musculoskeletal imaging taskforce [19].

Emerging issues in MRI surveillance of JIA patients relate to the presence of subclinical synovitis as a predictor of disease flare-up (Fig. 8) [12]. Bone marrow edema is a questionable predictor of an unfavorable outcome, as it can be found in more than 50% of healthy children, as shown by the Norwegian group [87].

Coronal magnetic resonance imaging of the hips in a 15-year-old girl with psoriatic arthritis (HLA B27 negative, ANA positive). a T2 turbo inversion recovery image shows an effusion in the left hip joint (arrow); bone and cartilage are normal in appearance. b Postcontrast T1-weighted fat-suppressed image shows synovial enhancement on the left (arrow)—a sign of active synovitis

MRI has demonstrated greater sensitivity than US and CR in detecting bone erosions, even in the early stages of the disease [94,95,96]. In contrast, there are conflicting data regarding the detection of cartilage erosions, probably due to the lack of cartilage-specific sequences in MRI protocols for JIA [36].

Specifically, standardized MRI protocols and semi-quantitative classification systems have been developed to assess inflammation and osteochondral changes in the large and small joints of JIA patients. These are currently undergoing validation, including assessing their correlation with clinical disease activity [97,98,99,100,101,102]. The ability to identify a “target joint” that reflects the global burden of disease activity may be an optimal target [46, 97].

MRI is highly valuable for difficult-to-access joints like the TMJ and the axial skeleton, which are commonly affected in JIA. Early detection is essential to prevent functional issues, including mandibular condyle growth inhibition and micrognathia (Fig. 9) [98, 102,103,104]. Recently, a consensus MRI protocol for the examination of the TMJ has been developed by Inarejos Clemente et al. [20], describing the degree of normal and pathological findings using the currently available MRI scoring systems developed for JIA.

Postcontrast T1-weighted fat-suppressed magnetic resonance images of the temporomandibular joints in a 14-year-old girl with oligoarticular juvenile idiopathic arthritis (ANA positive). a Axial image shows synovial thickening and enhancement on the right (arrow)—a sign of active synovitis. b Sagittal image shows a flattened mandibular head (long arrow) and synovial thickening and enhancement on the right (short arrow)—signs of chronic bone changes and active synovitis

Contrast-enhanced MRI is the preferred method for identifying cervical spine involvement, a significant prognostic factor for JIA progression [21, 22]. It has demonstrated higher sensitivity than clinical examination, although cervical arthritis is often clinically silent [23].

MRI is valuable for monitoring disease progression, response to treatment, and evaluation of late changes and complications, including atlantoaxial instability, dens deformity, joint ankylosis, and spinal cord compression [105].

MRI is also the method of choice for the assessment of the SIJ.

Nevertheless, the ASAS criteria, commonly used in adults to evaluate both active inflammatory and structural lesions, may present challenges applied to children [42].

The OMERACT expert working group together with OMERACT-JAMRIS-SIJ is developing and evaluating a preliminary pediatric consensus scoring system of SIJMRI. This system assesses inflammation and structural changes in the SIJ of children, including erosion, sclerosis, fat lesion, and ankylosis considering growing bone and active bone marrow [10].

In addition, pelvic MRI in juvenile ankylosing spondylitis is also valuable for assessing enthesopathy at the tendon and fascial attachment sites and for coxofemoral joint involvement, which is often associated with sacroiliitis [13].

Whole-body MRI is a promising tool for detecting and monitoring inflammation involving the peripheral joints, the axial joints, and the entheses in rheumatological diseases such as spondyloarthropathies [106,107,108]. On the other hand, there are no clear guidelines for the standardized detection, interpretation, and quantification of JIA on whole-body MRI. Moreover, MRI