In the sporting setting, selection of the most appropriate imaging modality may be greatly influenced by accuracy concerns, workflow and costs (please see also  “MRI of Muscle Injuries”). If US is performed sideline at the sporting event within the first few hours after the injury using portable machines, the risk of missing muscle tears should be taken into account because fresh hemorrhage (within 1-day after trauma) looks echogenic and may be quite similar to normal muscle. This problem may be compounded by two factors: (i) the image quality of portable US equipments may be suboptimal at the sideline of the sporting event to disclose minimal echotextural changes or does not have enough penetration to recognize deep-seated injuries; (ii) the examiner in charge of the event and the team is more likely to have learnt US as an adjunct to clinical work and to be less skilled to recognize the injury in a difficult operating setting (Allen and Wilson 2007). In the few hours after trauma, signs of a significant injury may be, in fact, subtle placing great demand on machine quality and operator experience. On the other hand, MR imaging is more reliable and easy to perform, but there may be problems of access and the cost may not be sustained by minor sports clubs systematically. Given some reservations, a wised use of US with selection of an appropriate timing of examination later than 1 day from trauma, when the hematoma starts to liquefy and collect becoming increasingly hypoechoic and conspicuous, might reduce the number of MR imaging examinations (Allen and Wilson 2007). This seems particularly true in the instance of mild trauma and outside the context of the highest professional levels.

When trauma occur in a professional sporting setting, the coaching staff of the club typically faces pressure to ensure a quick return of the athlete to play in order to avoid too many lost games, but a premature return to competition in an incompletely assessed and rehabilitated athlete has a high risk of reinjury and may result in a longer time removed from activity (Koulouris et al. 2007). Under these circumstances, US and MR imaging are extensively used, sometimes redundantly and in combination, to better prognosticate lay-off, healing time, mid-term outcome, as well as to guide rehabilitation and assess risk of late complication such as posttraumatic fibrosis, herniation and ossification (Connell et al. 2004; Gibbs et al. 2004; Schneider-Kolsky et al. 2006; Koulouris et al. 2007; Slavotinek 2010). Outside the high-performance context, muscle injuries often occur in the general population engaged in amateur athletic and non-athletic trivial activities. In these cases, imaging is basically directed to focus on more essential information, such as confirming the injury, defining its location and severity without the need to speculate on the exact time of return to play (Peetrons 2002). The need for serial follow-up examinations is also less. Given these considerations, one can imagine that the choice of the most appropriate imaging modality to perform is not only based on accuracy concerns but it may change depending on non-clinical factors such as level of the team, league requirements and budget (Allen and Wilson 2007). In the hamstrings, the percentage cross-sectional area (or volume) of injured muscle has been found to be a promising parameter to predict outcome, somewhat reflecting the proportion of myofibrils that have been disrupted at the level of measurement. An association between longer recovery time (>6 weeks) and involved muscle cross-sectional area >55 % has also been found (Slavotinek et al. 2002; Slavotinek 2010; Ekstrand et al. 2012). However, the strongest correlation with prognosis is given by the craniocaudal extension of muscle abnormality, as an estimate of the number of muscle units separated from the aponeurosis at the time of injury (Connell et al. 2004). Despite noticeably improvement in US technology with introduction and refinement of 3D systems and wide-FOV algorithms, size parameters are measured more simply and reproducibly on MR imaging rather than on US. In addition, the extent of the injury appears consistently smaller on US than on MR imaging, both in longitudinal and cross-sectional views, due to a lower sensitivity of US to depict edematous changes (Connell et al. 2004). Similarly, abnormalities tend to resolve sooner on US than on MR imaging during the healing process, thus making US less accurate in predicting the exact time to return to full competition.

US seems to have some advantages over MR imaging to depict posttraumatic scars (Fig. 2). They typically appear as a linear or stellate hyperechoic areas attached to the affected aponeurosis or the epimysium and cause distortion of the surrounding pennate pattern (Koh and McNally 2007). Local abnormal contraction of the muscle fibers can be revealed around the scar. From the biomechanical point of view, the local amount and severity of architectural distortion in a scarred muscle can be nicely estimated with US due to its dynamic capability. Scar conspicuity is also greater with US (Allen and Wilson 2007). On MR imaging, fresh scar may be possibly confused with an acute tear/retear due to a high signal on fluid-sensitive sequences, whereas old scars may be missed or underestimated in their extension as they are associated with low signal intensity in all sequences. Differently, a fresh retear on an old scar is more conspicuous on MR imaging due to its higher sensitivity in edema detection.