Received: 25 February 2002Revised: 13 June 2002Accepted: 26 June 2002Published online: 27 August 2002© ISS 2002

Abstract Objective: The purpose ofthis study was to review the normaland sonographic (US) anatomy ofthe central aponeurosis of the rectusfemoris muscle, describe the sono-graphic appearance of its tears andcorrelate it with the MR findings.Design and patients: The rectus fem-oris internal architecture was evalu-ated by cadaveric dissection. To cor-relate the sonographic normal find-ings with cadaveric data, axial sec-tions were compared with the corre-sponding US images. The normal invivo sonographic appearance of rec-tus femoris was assessed in 20 heal-thy subjects (40 thighs). To evaluatethe US findings in central aponeuro-sis tears we performed a retrospec-tive review of 17 examinations of 17 patients suffering from acute inju-ries. Follow-up examinations wereavailable in five patients. Sono-graphic findings were correlatedwith MR findings in eight patients.Results: Anatomical dissection ofthe rectus femoris confirmed thepresence of the central aponeurosis,a sagittally oriented fibrous band lo-cated within the proximal two-thirdsof the muscle belly. In vitro USshowed the central aponeurosis as acurvilinear hyperechoic structurewhose shape correlated well with thecadaveric data, and in vivo US dem-onstrated it in all healthy subjects. In the retrospective analysis of thepatient group, we classified the le-sions into three groups according to

the size at sonography: group 1(n=9), hyperechoic band surroundingan intact central aponeurosis; group2 (n=7), mixed hypo- and hyper-echoic bands surrounding an intactcentral aponeurosis but associatedwith globular enlargement of the rec-tus femoris; group 3 (n=1), completediscontinuity of the rectus femoris.These findings can reflect small par-tial tears, larger partial tears andcomplete musculotendonous junctiontears respectively. In patients evalu-ated with both techniques, MRI con-firmed the US findings. In five pa-tients follow-up studies showed anirregular hyperechoic ill-defined areacentred on the central aponeurosisthat was compatible with a centralscar. Conclusions: Sonography candemonstrate the normal internalanatomy of the rectus femoris andpost-traumatic changes at the myo-tendinous junction of the centralaponeurosis. Sonographic data corre-late well with MR findings, and thelow cost and wide availability of sonography make it the first-linetechnique in the evaluation of inju-ries of the rectus femoris.

Keywords Ultrasound · Quadricepsmuscle · Trauma

Skeletal Radiol (2002) 31:581–586DOI 10.1007/s00256-002-0559-z A RT I C L E

Stefano BianchiCarlo MartinoliNyali Peiris WaserMaria Pia Bianchi-ZamoraniEgisto FedericiJean Fasel

Central aponeurosis tears of the rectus femoris: sonographic findings

S. Bianchi (✉) · N.P. WaserDépartement de Radiologie, Division de Radiodiagnostic et de Radiologie Interventionelle, Hôpital cantonal Universitaire de Genève, Geneva, Switzerlande-mail: Stefano.Bianchi@hcuge.chFax: +41-022-3727047

S. BianchiGEL – Groupement des Echographistes de l’ Appareil Locomoteur (http://www.gelonline.org)

C. MartinoliCattedra di Radiologia “R”, DICMI-Università di Genova, Largo Rosanna Benzi 8, 16132 Genoa, Italy

M.P. Bianchi-ZamoraniHôpital cantonal Universitaire de Genève, Geneva, Switzerland

E. FedericiServizio di Radiologia, Ospedale di Lavagna, Italy

J. FaselDépartement de Morphologie, Centre medical Universitaire de Genève, Geneva, Switzerland

Introduction

Rectus femoris tears are common in sport activity [1]and can involve the proximal or distal parts of the mus-cle. Distal lesions affect the caudal myotendinous junc-tion where the muscle fibres can be seen inserting into aflat posterior tendon, which gives origin to the superfici-al layer of the quadriceps tendon. Distal tears can easilybe diagnosed by clinical means as a palpable lump due tothe proximally retracted muscle belly. Proximal intrasub-stance strains have recently been described and related topreviously unrecognised rectus femoris muscle internalarchitecture [2, 3]. They involve the central aponeurosis,a distinct, sagittally oriented, proximal aponeurosis lo-cated in the mid-muscle belly substance [3]. Due to theirdeep location and to the peculiar internal architecture ofthe rectus femoris, central aponeurosis tears are not asso-ciated with muscle retraction and therefore are difficultto diagnose at physical examination. Moreover, becauseof the associated muscle swelling, they can mimic softtissue tumors [4]. Sporadic cases [5] and a single series[3] have reported the MRI findings of central aponeuro-sis myotendinous lesions. In this article we present andcorrelate the cadaveric and sonographic (US) anatomy ofthe central aponeurosis and we review the US features oflesions that, to the best of our knowledge, have not pre-viously been described, and correlate them with the MRfindings.

Materials and methods

The normal architecture of the rectus femoris was studied by care-ful dissection of the right and left rectus femoris muscles of anembalmed cadaver. To correlate the sonographic internal appear-ance with the normal anatomy, axial sonograms were obtained atthe cranial, middle and distal third of two thighs of a second ca-daver and compared with the anatomical axial sections obtained atthe same levels. To determine whether in vivo US can detect thecentral aponeurosis and assess its appearance 40 thighs were thenexamined in 20 healthy volunteers (14 men, 6 women; mean age31 years). Reference scans of the normal sonographic anatomywere obtained both at rest and during isometric muscle activation.Axial and sagittal sonograms were obtained at the middle third ofthe muscle with the subject lying supine, the knee extended andthe quadriceps relaxed, and during isometric contraction of themuscle (20° hip flexion obtained with the knee extended). The ap-pearance, shape and maximal thickness of the central aponeurosiswere noted. All examinations were performed with the sameequipment (HDI 5000 ATL, Bothell, Wash.) working with lineararray transducers (frequency band 12–5 MHz).

We retrospectively reviewed the US images of 17 patients withcentral aponeurosis lesions (16 men, 1 woman; mean age26 years) who were referred over a 4-year period for acute anteriorthigh pain related to a recent sport injury. All patients reported asudden sharp pain localised in the proximal anterior aspect of thethigh during a sprint or a kicking motion during the course of afootball match. At physical examination the rectus femoris musclewas exquisitely tender to deep palpation in the proximal portion.The anterior thigh was swollen in eight patients. The occurrenceof the more common distal myotendinous tear was excluded on

the basis of the physical examination (absence of a distal lump andmore cranial location of pain). No history of systemic disorders orcorticosteroid therapy was found. Sonographic examinations wereperformed using three systems (AU-4 Idea, Esaote, Genoa, Italy;HDI-3000 and HDI-5000 ATL, Bothell, Wash.) working with lin-ear array transducers (frequency band 12–5 MHz, 10–13 MHz).Sagittal and axial sonograms were obtained over the anterior as-pect of the thighs with the patient supine and the knee slightlyflexed by a small pillow placed under the popliteal space. Nostandoff pad was deployed. Dynamic ultrasound included scan-ning during moderate quadriceps muscle contraction. The contra-lateral side was also examined to correlate the pathological find-ings with the normal appearance in the same subject. Color Dopp-ler images where obtained in five cases.

MRI was performed on eight patients using a 2 T unit (Pres-tige, Elscint, Haifa, Israel). The examination was undertaken withthe patient supine and the knee flexed at approximately 20°. Apelvic coil was utilized. Both thighs were examined to correlatethe pathological and normal side. Scout sequences were obtainedin the coronal plane to identify the optimal axial plane to evaluatethe affected area. Slice thickness was 7–10 mm without an inter-slice gap, FOV was 40 cm and two averages were obtained. Theimaging protocol consisted of proton density and T2-weightedFSE sequences (2900 /20–128) and IRFSE sequences (2600/130/90) obtained in the axial plane. In addition T2-weighted FSEsequences (3000/128) were obtained in the sagittal plane. BothMRI and sonography were obtained within 10 days of the trauma(mean 7 days). Five patients had follow-up US examination (meanfollow-up 30 days).

Results

Our anatomical study confirmed the previously de-scribed architecture of the rectus femoris (Fig. 1) [3].Correlation of axial cadaveric sections with correspond-ing in vitro sonograms demonstrated in all six sectionsthat the central aponeurosis can easily be detected as acomma-shaped, sharply delineated hyperechoic band,oriented on the sagittal plane and located in the superfi-cial portion of the cranial two-thirds of the muscle. Thesonographic findings of the central aponeurosis (size,thickness and appearance) correlated well with data fromanatomical cross-sections (Fig. 2).

In the 20 healthy volunteers US detected the centralaponeurosis in all 40 thighs. Images obtained in the axialplane were the most useful in showing the aponeurosiswhereas sagittal sonograms could not identify it satisfac-torily. The mean thickness was 1.5 mm. Its sonographicappearance was similar to that observed in vitro althoughin vivo images provided better delineation between theaponeurosis and the muscle fibres (Fig. 3). The poorerquality of the US image was probably related to partiallysis of the muscle fibres in cadavers. Axial dynamic ex-amination obtained with progressive distal movement ofthe transducer disclosed the fibroadipose septa as hyper-echoic linear structures that converge and insert on boththe lateral and medial surfaces of the central aponeuro-sis. Examination obtained during isometric muscle acti-vation showed a more globular appearance of the rectusfemoris together with decreased echogenicity, due to

582

muscle shortening and increased size of the hypoechoicmuscle fibres. Sonography could also detect a straight-ened appearance of the central aponeurosis, probably re-lated to simultaneous contraction of muscle fibres insert-ing into its lateral and medial aspects.

In the retrospective analysis of our patients threegroups of US patterns emerged. In the first group (9 pa-tients) the muscle maintained its normal oval shape andsize comparable to the unaffected contralateral side. So-nography showed an irregular hyperechoic area insidethe muscle surrounding the central aponeurosis whichwas suspected to represent blood infiltration of the peri-aponeurotic tissues (Fig. 4). In these cases the centralaponeurosis was partially masked by hyperechoic bloodand barely detectable. In a second group (7 patients) themuscle appeared swollen and showed a more globular

583

Fig. 1A, B Macroscopic appearance of a dissected rectus femoris muscle (right distal side, left proximal side). A Anterior face, B posterior face. SA superficial aponeurosis, arrows central aponeurosis, DT distal tendon

Fig. 2 A Axial cadaveric section obtained at rectus femoris middle third. B Corresponding in vitro US image. RF rectus femoris muscle,arrows central aponeurosis, VI vastus intermedius muscle

Fig. 3 Axial US image obtained at the rectus femoris middle thirdin a normal subject. RF rectus femoris muscle, arrows central apo-neurosis, VI vastus intermedius muscle

appearance with a larger central area made of both hypo-echoic and hyperechoic zones, probably related to bloodinfiltration and haematoma (Fig. 5). In some patients abull’s eye appearance was found, made by a hypoechoicarea encircling the central aponeurosis and surroundedby a hyperechoic circular zone. This appearance wasthought to represent a haematoma surrounded by periph-eral blood infiltration. Pressure applied over the injuredmuscle showed a partial collapse of the hypoechoic area.It must be noted that in both groups the outer portion ofthe muscle appeared normal. Colour Doppler sonogramsshowed peripheral flow signals related to surroundinghyperaemia, but in our experience added little to thestandard examination. In the third group (1 patient) acomplete disruption of the insertion of the muscle fibresin the central aponeurosis was evident. In this patient thecentral aponeurosis appeared as a hyperechoic continu-ous structure crossing an anechoic area, suspected to re-present a haematoma, interposed between the retractedmuscle ends (Fig. 6). In all patients of our series the cen-tral aponeurosis appeared uninterrupted, thus confirmingthe fact that the basic lesion of the rectus femoris intra-

substance tear is a myotendinous disruption rather than atrue tendon lesion. In five patients (two in group 1 andthree in group 2) the follow-up examinations (mean fol-low-up 45 days) showed a localised disruption of the nor-mal architecture of the area surrounding the central apo-neurosis which appeared hyperechoic and poorly delin-eated, probably due to fibrous tissue replacement (Fig. 7).

Axial MR images of the normal contralateral thighsshowed the unaffected central aponeurosis as a linear hy-pointense structure surrounded by normal muscle tissue.The overall appearance of the cadaveric, US and MR im-ages correlated well. In group 1 lesions T2-weighted, FST2-weighted and IRFSE MR sequences showed an areaof high signal intensity surrounding the unaffected cen-tral aponeurosis, suspected to represent acute inflamma-tory infiltration. Typically, as in the sonograms, the outerportion of the muscle was normal. In group 2 lesions alocalised, more hyperintense area probably related to he-matoma was evident. The patient presenting at US with acomplete tear (group 3) refused MRI evaluation.

Because of limited sport requirements our patientswere treated with medical therapy and physiotherapy. A

584

Fig. 4A, B Group 1 lesion. A Axial US image. B Corre-sponding FS T2-weighted MRimage. In A note the central,poorly delimitated, rounded hy-perechoic area masking thecentral aponeurosis suspectedto represent a haemorrhagic in-filtration (arrows). The periph-eral portion (PP) of the rectusfemoris muscle is normal. In Bnote the rounded hyperintensearea (arrows) centred on theuninvolved, normal-appearingcentral aponeurosis (small arrow). Here again the periph-eral portion (PP) of the rectusfemoris muscle is unaffected

Fig. 5A, B Group 2 lesion. A Axial US image. B Corre-sponding T2-weighted MR image. In A note the roundedheterogeneous, hypo-hyper-echoic area suspected to repres-ent a haematoma and tissue in-filtration (arrows). The centralaponeurosis appears as a curvi-linear hyperechoic structure(small arrows). In B note thecorresponding irregular hyper-intense area (arrows) centredon the uninvolved, normal-ap-pearing central aponeurosis(small arrows). In both imagesnote the globular appearance ofthe rectus femoris and the unaf-fected peripheral portion (PP)

good outcome was noted in the patients who had sono-graphic follow-up (5 patients).

Discussion

Rectus femoris is the most superficial muscle belly ofthe quadriceps muscle and originates from the region ofthe antero-inferior iliac spine, crossing the hip to inserton the upper pole of the patella. Its main actions are kneeextension and hip flexion.

The direct and the indirect tendon compose the proxi-mal tendon of the muscle. The direct tendon originatesfrom the anterior inferior iliac spine and continues intoan aponeurosis which covers the anterior aspect of theproximal muscle (superficial aponeurosis). The indirecttendon originates from the superior acetabular ridge andcontinues in the sagittal-oriented central aponeurosis that

is located inside the proximal muscle belly. From bothaponeuroses, muscle fibres run distally to insert into theposterior distal tendon. The fibres which originate fromthe superficial aponeurosis are more peripheral and giverise to a unipennate muscle whereas those from the later-al and medial aspect of the central aponeurosis are locat-ed in a central position and form a bipennate muscle.The rectus femoris therefore has a peculiar “musclewithin a muscle” appearance due to the presence of thetwo distinct muscles: the peripheral unipennate whichsurrounds the central bipennate. It must be stressed thatthis peculiar internal architecture was unrecognised priorto the report by Hughes et al. [3] in 1995 and not de-scribed in the main anatomical textbooks [6, 7, 8] or inarticles dedicated to the clinical anatomy of commonlyinjured muscles [9].

The rectus femoris is frequently affected by strainsbecause, like the medial head of the gastrocnemius, itcrosses two joints and has a high percentage of type IImuscle fibres, well suited to rapid forceful activity.Moreover the possibility of a strain is increased by thefact that it contracts while it is passively stretched. Labo-ratory studies have demonstrated that the myotendinousjunction is the most commonly affected site for partial orcomplete muscle injury [10]. The more commonly re-cognised rectus femoris injury is located at the distalmuscle-tendon junction, occurs in sports in which kick-ing or sprinting are required, and is usually diagnosed onbasis of the clinical findings. The recently described in-trasubstance tear is related to a specific internal architec-ture of the muscle and, because of its deep location, ismore difficult to diagnose clinically. On the other hand,the early, accurate location and assessment of the size ofthe rupture can be helpful in assessing the time to returnto competition.

In our study ultrasound was able to show the normalinternal architecture of the rectus femoris in a cadavericspecimen as well as in normal subjects. In the patientgroup sonography showed the typical appearance of dif-ferent degrees of muscle tears [11]. As previously report-

585

Fig. 6A, B Group 3 lesion. A Sagittal US image. B AxialUS image.A complete tear ofthe rectus femoris is evident.The muscle gap is filled by ananechoic collection suspectedto represent a haematoma(Hem). Note the rounded butcontinuous central aponeurosis(arrowheads) and the retractedrectus femoris

Fig. 7 Group 2 lesion. Follow-up appearance. Axial US image ofa group 2 tear obtained 2 months after the injury. The normal in-ternal architecture of the rectus femoris is no longer apparent. Ahyperechoic, spiculated area located in its central portion is evi-dent, suspected to represent scar tissue involving the area aroundthe central aponeurosis (arrow). VI vastus intermedius muscle

ed, in smaller lesions a localised irregular hyperechoicarea reflects the haemorrhagic infiltration of the sur-rounding muscle [12]. In group 1 lesions, which usuallypresented minor clinical findings, a high degree of suspi-cion and a careful examination technique must be uti-lized to detect subtle changes of the echogenicity whichotherwise may be easily overlooked. Local pressure withthe probe while scanning the entire muscle is very help-ful in localising the lesion because of the painful reactionof the patient when the pressure is applied directly. Al-though small tears carry a good prognosis, their early de-tection is important since sport activity can avoided andthe development of larger rupture prevented. In group 2lesions a localised, irregular hypoechoic area is probablyrelated to a haematoma and the surrounding hyperechoichalo may be related to haemorrhagic infiltration of themuscle. In the case of complete myotendinous disruptiona huge fluid collection interposed between the retractedmuscle ends was evident and the central aponeurosis hada rounded appearance on axial sonograms, probably dueto the absence of the action of muscle fibre tension. Al-though larger tears are also generally treated conserva-tively, this topic is debated and some authors suggestsurgical treatment in those involved in sports at a high

level. In this situation sonography is of value in the as-sessment of the size of the strain, which can affect thetherapeutic decision, and in excluding a soft tissue tumorin patients presenting with huge local edema. In the fol-low-up of the patients sonographic findings of irregularperiaponeurotic tissue seemed to correlate with the de-scribed rectus femoris internal fibrosis [13], although wecan only speculate on this point since our study has nohistological correlation.

In summary, central aponeurotic lesions are recentlyrecognised entities which involve the proximal myoten-dinous junction and are difficult to diagnose clinically,and imaging is valuable in assessing the size of the tearand in excluding a local tumor. Sonography shows thenormal architecture in cadavers as well as in normal sub-jects. It demonstrated tears retrospectively in all 17 pa-tients evaluated retrospectively and correlated well withthe MR appearance. Three types of appearances wereseen in this study. We suggest the use of sonography asthe first imaging modality in the diagnosis and follow-upof central aponeurosis lesions because of its low cost,noninvasiveness and availability. If US is inconclusive,MRI will be the subsequent imaging modality for furtherevaluation.

586

References

1. Garrett WE Jr. Muscle strain injuries.Am J Sports Med 1996; 24:S2–8.

2. Hasselman CT, Best TM, Hughes C4th, Martinez S, Garrett WE Jr. An ex-planation for various rectus femorisstrain injuries using previously unde-scribed muscle architecture. AmJ Sports Med 1995; 23:493–499.

3. Hughes C 4th, Hasselman CT,Best TM, Martinez S, Garrett WE Jr.Incomplete, intrasubstance strain inju-ries of the rectus femoris muscle. Am J Sports Med 1995; 23:500–506.

4. Temple HT, Kuklo TR, Sweet DE, Gibbons CL, Murphey MD. Rectusfemoris muscle tear appearing as apseudotumor. Am J Sports Med 1998;26:544–548.

5. Greco A, McNamara MT, Escher RM,Trifilio G, Parienti J. Spin-echo andSTIR MR imaging of sports-relatedmuscle injuries at 1.5 T. J Comput Assist Tomogr 1991; 15:994–999.

6. Williams PL, Bannister L, Berry MM,et al.: Gray’s anatomy. Edinburgh:Churchill Livingstone, 1995:872–873.

7. Hollinshead WH, Rosse C. Textbookof anatomy, 4th edn. New York: Harperand Row, 1985:376–377.

8. Rouvière H. Anatomie humaine, descriptive, topographique et fonction-nelle. Paris: Masson, 1974:354.

9. Chammout MO, Skinner HB. The clin-ical anatomy of commonly injuredmuscle bellies. J Trauma 1986;26:549–552.

10. Garrett WE Jr. Muscle strain injuries:clinical and basic aspects. Med SciSports Exerc 1990; 22:436–443.

11. Takebayashi S, Takasawa H, Banzai Y,et al. Sonographic findings in musclestrain injury: clinical and MR imagingcorrelation. J Ultrasound Med 1995;14:899–905.

12. Bianchi S, Martinoli C, Abdelwahab IF,Damiani S, Derchi LE. Sonographicevaluation of tears of the gastrocne-mius medial head (tennis leg). J Ultrasound Med 1998; 17:157–162.

13. Rask MR, Lattig GJ. Traumatic fibrosisof the rectus femoris muscle. Report offive cases and treatment. JAMA 1972;17:268–269.