The ideal musculoskeletal imaging modality for a clinical rheumatologist needs to fulfil a wide range of capabilities and exacting standards. Of paramount importance is the need to obtain precise, detailed pathoanatomical images—lacking artefacts—in any given plane of any given joint of every patient, all in real time so as to provide functional information under the examiner’s immediate visual control. The end result should be to produce additional, different and more valuable information superior to that gathered from clinical examination alone. Furthermore, the imaging process should be easily and accurately repeatable without causing any harm to the patient, in order to gain reliable information and for systematic comparison of sequential images following therapy. Cost-effectiveness, a short examination time, instant access and portability are also important factors when evaluating different imaging methods. Last, but not least, all rheumatologists appreciate the value of imaging guidance when performing technically difficult interventional procedures at the bedside or in the outpatient clinic.

In addition to all of these criteria the ideal musculoskeletal imaging modality would need to automatically combine morphological and functional musculoskeletal images in real time. Morphological imaging in musculoskeletal medicine first began with anatomical drawing, followed by the use of microscopy and photography. Permeating the human skin under in vivo conditions was a radical leap forward, when Wilhelm Conrad Roentgen was awarded the first Nobel Prize in Physics for pioneering X-ray imaging in 1895. The X-ray continues to be an everyday tool in rheumatological practice, though in terms of structural delineation of musculoskeletal tissues the rheumatologist now has a wider armamentarium to call upon, including ultrasonography, computerised tomography and magnetic resonance imaging.

Functional imaging of musculoskeletal tissues started with the invention of the motion picture, which allowed the development of cineradiography. This was followed by radionuclide imaging based on the works of George de Hevesy, who was awarded the Nobel prize in 1943 for using the first radioactive substances as indicators. The subsequent development of real-time ultrasonography, varied ultrasound and laser Doppler methods, infrared thermography, videoscope, cineradiography and functional magnetic resonance imaging, single photon emission computed tomography and positron emission tomography have considerably broadened the horizon of functional imaging in rheumatology. Rheumatologists today now take for granted the ability to visualise the processes of musculoskeletal inflammation that previously could only be implied from morphological imaging.

Over the last decade, real-time ultrasound has emerged as one of the leading contenders to be the ideal musculoskeletal imaging modality, capable of combining both morphological and functional imaging in musculoskeletal soft tissues. Increasing numbers of rheumatologists have been using grey-scale, colour and power Doppler ultrasonography not just as a research tool, but also—especially in many European countries—in daily rheumatological practice. It is now included in the rheumatology curriculum of many different European countries and the European League Against Rheumatism (EULAR) has promoted basic, intermediate, and advanced ultrasound courses in addition to a number of projects under the auspices of the EULAR Working Party on Musculoskeletal Imaging aimed at standardising both ultrasound training and practice.

High-resolution grey-scale ultrasonography improves our ability to detect tiny, hidden erosions and minute amounts of fluid and soft tissue changes in synovial joints at the earliest stages of disease.1 2 The resolution of grey-scale ultrasonography is now under 300 μm producing one of the highest levels of definition of musculoskeletal soft tissue morphology.3 In combination with real-time imaging advances in colour and power Doppler imaging may allow functional depiction and assessment of inflamed joints and vasculitides. Initially, power Doppler was considered far more superior in sensitivity with respect to the detection of slow flow in soft tissues and also promised fewer artefacts, than the forerunner, colour Doppler. However, only one study has tested different ultrasound machines to establish the variability of the threshold level of detecting slow flow in musculoskeletal conditions4 and only one study has compared power Doppler findings with a physically different type of flow measurement in patients with rheumatoid arthritis.5 Our knowledge of Doppler imaging of normal joints is also limited to a single ultrasound study of hands and wrists, which lacked direct anatomical or physiological correlates of joint perfusion as a control.6 While most ultrasound novices are extremely keen to move all too rapidly from grey-scale imaging to power or colour Doppler imaging the subsequent interpretation of a power Doppler scan of a rheumatoid metacarpophalangeal joint may prove perplexing.

These bad experiences with power Doppler are often due to the frequent presence of a number of technical, systematic or observer-related artefacts that are presented in detail by Dr Torp-Pedersen and Dr Terslev in an excellent picturesque overview in this issue (page 143).7 This paper will help trainees set up their machines for their first examinations, to learn the initial steps of Doppler imaging and the “knobology” of blood flow depiction via ultrasound while their mentors guide them through this blossoming new world of musculoskeletal imaging. However, even with a clear awareness of artefacts it is hard to convince a sceptical rheumatologist to accept an image that depicts real flow and artefacts at the same time. All ultrasonographers would be happier if we could exclude all artefacts from the ultrasound scan, though this is not the case even with grey-scale ultrasonography. It is also possible to make mistakes using ultrasonography, by failing to notice an existing, detectable change or by misjudging an artefact for an existing, detectable change (as highlighted by Drs Torp-Pedersen and Terslev). It is also quite problematic to evaluate an image with confluent signals covering large parts of the joint or to be certain that minor Doppler signals represent true blood flow. Practising ultrasonographers already know that the puncture of a joint that has a large confluent power Doppler signal occupying more than 50% of the joint space will not produce an aspiration of pure blood. In the pursuit of superior grey scale and Doppler imaging today’s ultrasonographer is using ultrasound equipment with higher power outputs than ever before for longer examination times on a wider range of joints. It is more important than ever to balance the pursuit of technical excellence with patient safety by practising the “As Low As Reasonably Achievable” (ALARA) principle8 of the American Institute of Ultrasound in Medicine (AIUM) and the European Federation of Societies for Ultrasound in Medicine and Biology (EFSUMB).9 10

Three- and four-dimensional power Doppler—and in the future vascular endothelial growth factor driven contrast agents—may produce more accurate details of vascular branches and networks and provide more convincing evidence of joint inflammation. With the necessary resolution, the combined use of grey-scale and power Doppler could be advantageous in determining whether an erosion is active is burned out or which parts of a pannus are inflamed and which parts are fibrotic. It can also be used detect changes in intra-articular perfusion induced by local cryotherapy, various tumour necrosis factor blockers as well as the local administration of steroid.11^–15 We can only hope that we will reach that pinnacle in the near future, where colour and power Doppler images (functional images) perfectly correspond to real-time grey-scale images (morphological images). In this way different ultrasound methods will supplement each other similarly to that seen in the combination of different imaging modalities (eg, positron emission tomography/computed tomography) that produce fusion images.16