quantitative ultrasonography of facial muscles in patients with chronic facial palsy

QUANTITATIVE ULTRASONOGRAPHY OF FACIAL MUSCLES INPATIENTS WITH CHRONIC FACIAL PALSYGERD FABIAN VOLK, MD,1,2 MARTIN POHLMANN, MD,1 MAIK SAUER, MD,1 MIRA FINKENSIEPER, MD,1,2 and

ORLANDO GUNTINAS-LICHIUS, MD1,2

1 Department of Otorhinolaryngology, Jena University Hospital, Lessingstrasse 2, D-07740 Jena, Germany2 Facial Nerve Center, Jena University Hospital, Jena, Germany

Accepted 23 December 2013

ABSTRACT: Introduction: In this study we introduce quantitativefacial muscle ultrasound as a diagnostic tool for patients withchronic unilateral facial palsy. Methods: Muscle area, thickness,and echo intensity of 6 facial muscles (frontalis, orbicularis oculi,orbicularis oris, depressor anguli oris, depressor labii inferioris,and mentalis) and of 2 chewing muscles (temporalis and mass-eter, as controls) were measured in 20 patients with chronicfacial palsy. Results: Aside from 1, all facial muscles were signifi-cantly smaller on the paralyzed side. With exception of frontalisand orbicularis oculi muscles, all other facial muscles showedsignificantly higher echo intensity on the affected side. Musclesize and echo intensity of the chewing muscles showed no side-to-side asymmetry. Conclusions: Quantitative ultrasound of facialmuscles helps to better characterize their status in patients withchronic facial palsy in the phase of denervation and duringregeneration.

Muscle Nerve 50: 358–365, 2014

A severe facial nerve lesion, such as after temporalbone trauma or tumor surgery, leads to Walleriandegeneration of the peripheral facial nerve distalto the lesion site and consequent atrophy of themimetic (facial) musculature. If the facial nerve isregenerating and the muscles are reinnervated, themimic musculature often does not completelyrecover functionally. In case of misdirected reinner-vation and defective healing, the muscles canremain weakened or are characterized by synkinesisand hyperkinesis. Treatment decisions for (a) facialnerve reconstruction after long-term denervation or(b) adjuvant surgical and non-surgical treatment ofdefective healing after misdirected spontaneousregeneration or after facial nerve reconstruction sur-gery is based mainly on the clinical observation thatfunctional results are worse if >2 years elapse afterdenervation.1,2 In addition, during muscle denerva-tion, the degree of muscle atrophy can be roughlyestimated by analysis of the needle electromyo-graphic (EMG) insertional activity. Furthermore,multichannel EMG can be used to quantitativelycharacterize involuntary synkinetic activity inpatients with chronic facial palsy.1,2

In recent reports, quantitative muscle ultrasoundhas been used to measure facial muscle size andecho intensity in healthy subjects.3,4 In addition, thisnew imaging technique was applied as an examplein a patient with myotonic dystrophy type 1.4

In this study we report muscle cross-sectionalarea, thickness, and echo intensity in patients withchronic peripheral facial palsy in different stagesof denervation and reinnervation. The primarygoal was to determine whether quantitative muscleultrasound can detect differences in facial musclesize and echo intensity between paralyzed andhealthy sides in patients with chronic facial palsy.

METHODS

Patients. Twenty patients with chronic facial palsywere recruited between June and September 2012.Chronic palsy was defined as a peripheral facialpalsy with a duration of >3 months. Furthermore,an initial EMG study had to show signs of denerva-tion; that is, pathological spontaneous activity andreduced voluntary activity in at least 2 mimicmuscles on the affected side.5,6 The last EMG study(<4 weeks) before ultrasonographic examinationwas also evaluated for signs of reinnervation (poly-phasic motor unit potentials and synkinetic activa-tion). The study was approved by the local ethicscommittee, and informed consent was obtainedfrom all participants. Furthermore, consent wasobtained for use of patients’ images. Age, gender,body weight, height, body mass index (BMI), andhandedness of all subjects were recorded. Patients’charts were evaluated for the origin of facial palsy,types of reconstructive surgery, and facial nerve grad-ing using the Stennert index.7 The Stennert index atrest (range from 0 to 4) was evaluated for associa-tions with the ultrasonographic examinations. AStennert index at rest of 0 represents normal facialtone, and an increasing loss of facial tone is gradedup to 4; that is, complete loss of resting tone. Asonly ultrasonographic evaluations at rest were per-formed, the result of the Stennert index concerningfacial movements (range from 0 to 6; 0 5 normal to6 5 no motion) were not further explored.

Ultrasonographic Examination of Facial

Muscles. Each patient was examined once. Theultrasonography examinations were performed by

Abbreviations: EMG, electromyographyKey words: facial muscles; facial nerve; facial palsy; mimetic muscles;quantitative sonography; reconstructive surgeryCorrespondence to: O. Guntinas-Lichius; e-mail: [email protected]

VC 2013 Wiley Periodicals, Inc.Published online 28 December 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/mus.24154

358 Facial Muscle Sonography MUSCLE & NERVE September 2014

2 of the investigators (M.P. and M.S.). Examinerswere blinded to the patients’ baseline characteristics.All measurements were performed using a diagnos-tic ultrasound system (HD11 XE; Philips, Amster-dam, The Netherlands). Two different linear-arraytransducers were used: a 3–12-MHZ transducer (L12-3; Philips) and a 7–15-MHZ transducer (L15-7io; Phi-lips). As described recently,3 1 probe was preferredfor some muscles and another for other muscles.For muscles located on more curved portions of theface, such as the orbicularis oculi and orbicularisoris muscles, the smaller probe (L15-7io) was prefer-able due to its smaller footprint. Otherwise, thelarger transducer (L12-3) was used (see Table 1 inthe study by Volk et al.3). The equipment settingswere fixed, with mechanical index 1.5 (for L12-3)and 1.4 (for L15-7io), thermal index 0.6 (for both),and gain 75 dB, using constant time gain compensa-tion. All equipment settings were kept constant dur-ing the measurements. All measurements wereperformed with transmission gel, but without anyother coupling media. The transducer was alwaysplaced perpendicular to the skin surface.

Ultrasonography scans were made of the follow-ing 6 mimic muscles on both sides: frontalis; orbi-cularis oculi; depressor anguli oris; depressor labiiinferioris; orbicularis oris; and mentalis. As shownin previous studies in healthy subjects, other facialmuscles cannot always be identified in all subjectsor with sufficient reliability. Two additionalmuscles were analyzed: the temporalis and mass-eter muscles. This was done to establish referencedata for neighboring muscles that are not inner-vated by the facial nerve as bilateral referencemuscles. The identification of the muscles and themeasurement procedure of each muscle have beendescribed in detail recently.3

Quantitative Measurements. All ultrasound imageswere stored in the DICOM format. Three separatemeasurements were made for each muscle at rest.Muscle cross-sectional area, thickness, and gray-value analysis were estimated using quantificationsoftware (QLAB; Philips). To determine the cross-sectional area (in mm2) of each muscle on bothsides of the face, the muscle borders were delim-ited manually on each image as regions of interest(ROIs) using the freehand function of QLAB (Fig.1). Maximal muscle thickness was determinedorthogonal to the muscle fiber orientation. Thetemporalis, masseter, orbicularis oris, orbicularisoculi, and frontalis muscles were too large to visu-alize the complete muscle area on a single image.Hence, the maximal muscle diameter orthogonalto the muscle fiber orientation was measured (inmm). As in the recent pilot study,3 only the musclecross-sectional area but not the maximal musclethickness was estimated for the depressor anguli oris,depressor labii inferioris, and mentalis muscles. Theretest reliability of the thickness measurements forthese 3 muscles seems to be insufficient (Volk et al.,unpublished data). The ratio between the affectedside and the contralateral healthy side was calculated(in percent) as follows: ratio 5 [(palsy side/healthyside) � 100] 2 100. Using gray-scale analysis, the echointensity of each muscle was quantified.4 The meangray value of the ROI from each muscle was calcu-lated. The results of the 3 separate measurementswere averaged for each muscle to minimizevariation.4,8

Statistics. All statistical analyses were performedusing IBM SPSS, version 21.0. If not otherwise indi-cated, data are presented as mean 6 standard devi-ation (SD). The non-parametric Wilcoxon test was

Table 1. Muscle thickness and area at rest.

MusclePalsy side

[mean (SD)]Healthy side[mean (SD)]

Ratiopalsy/healthy

side* P†

Frontalis, thickness (mm) 3.00 (0.85) 3.15 (0.78) 24.6% 0.147Frontalis, area (mm2) 68.29 (18.82) 71.95 (20.65) 22.0% 0.502Orbicularis oculi, thickness (mm) 0.67 (0.31) 0.82 (0.23) 211.0% 0.040Orbicularis oculi, area (mm2) 5.56 (4.33) 7.92 (4.08) 225.2% 0.012Orbicularis oris, thickness (mm) 2.57 (0.39) 2.80 (0.61) 25.8% 0.037Orbicularis oris, area (mm2) 30.38 (6.75) 35.34 (9.94) 27.6% 0.028Depressor anguli oris, area (mm2) 17.43 (10.76) 29.17 (9.82) 236.2% 0.002Depressor labii inferioris, area (mm2) 14.81 (6.96) 21.60 (5.33) 233.6% <0.0001Mentalis, area (mm2) 25.60 (10.71) 34.56 (9.48) 223.2% 0.004Temporalis, thickness (mm) 18.85 (2.12) 19.39 (2.04) 22.6% 0.086Temporalis, area (mm2) 384.48 (63.23) 395.88 (58.47) 21.7% 0.550Masseter, thickness (mm) 10.96 (2.78) 10.86 (1.96) 0.1% 0.936Masseter, area (mm2) 243.38 (84.42) 233.29 (56.07) 2.7% 0.411

*Ratio (%) 5 [(palsy side / healthy side) 100] – 100.†Statistically significant P-values (P< 0.05) in bold.

Facial Muscle Sonography MUSCLE & NERVE September 2014 359

applied to analyze differences between the affectedfacial side and the healthy contralateral side. Thenon-parametric Mann–Whitney U-test was used toanalyze the influence of clinical parameters on themuscle measurements. Nominal P-values of 2-tailedtests are reported. The significance level was set atP< 0.05. Informed written consent was obtainedfor publication of patients’ images.

RESULTS

Patients’ Characteristics. The study sample con-sisted of 16 women and 4 men. The average agewas 57 6 18 years (median 61, range 26–85). BMIwas 27 6 6 kg/m2 (median 24, range 19–37). Allpatients, except 1, were right-handed. The etiologyof facial palsy was: postsurgical (n 5 14); idiopathic(n 5 3); Ramsey–Hunt syndrome (n 5 1); trauma(n 5 1); and preoperative malignant parotid tumorinfiltration (n 5 1). The average duration of thepalsy since onset was 51.3 6 59 months (range10.0–102.4). At the time of the ultrasonographicexamination, 10 patients had already undergonefacial nerve reconstruction via facial nerve interpo-sitional graft or hypoglossal–facial jump nervesuture. In these 10 patients, reconstructive surgerywas performed 11.8 6 7.3 months (median 3–21.8)before ultrasonographic assessment. EMG stillshowed evidence for denervation of the mimicmuscles in 6 patients (patient examples in Fig. 2)and reinnervation in 14 patients (patient examplesin Fig. 3). 5 patients showed synkinetic activity.The Stennert index at rest was 2.4 6 1.2 (median2.5, range 0–4). Stennert index during facial move-ments was 4.5 6 1.9 (median 6, range 1–6).

Muscle Thickness and Areas at Rest. Muscle thick-ness and area of the facial and chewing muscles

for the affected side and the contralateral healthyside are listed in Table 1. Significantly smaller sizeson the palsy side were measured for thickness andarea of the orbicularis oculi, orbicularis oris,depressor labii inferioris, depressor anguli oris,and mentalis muscles (all P< 0.05). No side differ-ences for affected and normal side were seen forthe frontalis muscle and the 2 chewing muscles.When focusing on the affected side only, dener-vated mimic muscles showed significantly smallersize than reinnervated muscles for orbicularisoculi thickness (P 5 0.003), orbicularis oculi area(P 5 0.026), and depressor anguli oris area(P 5 0.012). Only the depressor anguli oris showeda larger area in the case of synkinetic activity com-pared to patients without synkinesis (P 5 0.019). Ahigher Stennert index at rest (Stennert index 3and 4; i.e., less facial tonus) was associated with asmaller depressor anguli oris area (P 5 0.004),a thinner orbicularis oculi muscle (P 5 0.003), anda smaller orbicularis oris area (P 5 0.043), all onthe affected side, compared to patients with betterresting tone (Stennert index 1 and 2). All men-tioned parameters with influence on muscle meas-urements of the affected side had no significantinfluence on the healthy side (all P> 0.05). Takinginto account that a series of 1 patient per time-point was analyzed, it seemed from visual inspec-tion that, for most of the mimic muscles, the mus-cle thickness and area decreased on the paralyzedside in relation to the healthy side and remainedlower over the time course of denervation (Fig. 4A,top). After reinnervation, muscle thickness andarea seemed to increase again and even appearedto be higher finally on the paralyzed side than onthe healthy side (Fig. 4A, bottom). The 2 chewingmuscles did not show a significant effect over the

FIGURE 1. Example of quantitative facial muscle measurement after post-processing. The depressor anguli oris (DAO) and the

depressor labii inferior (DLI) on the healthy left side are shown. (A) Original image; (B) decreased brightness; (C) increased contrast;

and (D) final manual delimitation of the region of interest (ROI), using the freehand function of the software to determine the cross-

sectional areas. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]


time course of denervation (and reinnervation) ofthe facial nerve.

Muscle Echo Intensity at Rest. Muscle echo inten-sity of the facial and chewing muscles for theaffected side and the contralateral healthy side are

listed in Table 2. Significantly higher echo inten-sity was measured for all mimic muscles (P< 0.05),except for the orbicularis oculi and frontalismuscles (P 5 0.212 and P 5 0.108, respectively).The 2 chewing muscles did not show significantside differences. When focusing on the affected

FIGURE 2. Two faces at rest with chronic facial palsy (denervated facial muscles on the left side). (A, B) Patient 1 denervation for 30

months, Stennert index 4/6. (C, D) Patient 2 denervation for 43 months, Stennert index 4/6. (A, C) Normal imaging. (B, D) Image with

superimposed muscle drawing and ratio values of muscle size for frontalis, orbicularis oculi, orbicularis oris, depressor anguli oris,

depressor labii inferioris, and mentalis. Negative values: muscle size smaller on palsy side in black; positive values: muscle size larger

on palsy side in white. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]


side only, innervation status, synkinetic activity,and Stennert index at rest had no significant asso-ciation with echo intensity (all P> 0.05). Echointensity seemed to increase on the paralyzed sidefor some mimic muscles during denervation

(Fig. 4B, top) and appeared equivalent for bothsides during reinnervation (Fig. 4B, bottom). Echointensity of the 2 chewing muscles did not changesignificantly during the time course of denervationand reinnervation.

FIGURE 3. Two faces at rest with chronic facial palsy with reinnervated facial muscles and defective healing on the right side (A, B)

and on the left side (C, D) characterized by severe hyperkinesis of the orbicularis oris muscle. (A, B) Patient 1 reinnervation after 37

months, Stennert index 1/1. (C, D) Patient 2 reinnervation after 69 months, Stennert index 2/2. (A, C) Normal imaging. (B, D) Imaging

with superimposed muscle drawing and ratio values of muscle size for frontalis, orbicularis oculi, orbicularis oris, depressor anguli oris,

depressor labii inferioris, and mentalis; negative values 5 muscle size smaller on palsy side in black; positive values 5 muscle size

larger on palsy side in white. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]


DISCUSSION

Using quantitative ultrasonography, we haveshown that mimic muscle size is decreased afterfacial nerve denervation and that muscle sizeincreases again after reinnervation. In accordancewith the EMG data, we found that synkineticallyreinnervated muscles, especially with clinical andelectrophysiological signs of hyperkinesis, arecharacterized by an increase in muscle size com-pared to the healthy side. Previously, mimetic

muscle atrophy has only been demonstrated qualita-tively in selected muscles by magnetic resonanceimaging.9 Furthermore, we found that the echointensity of the facial muscles increased duringdenervation and was restored to equivalent valuescompared with the normal side during reinner-vation. Recently, it was shown that this methodcould detect changes of mimic muscle size andecho intensity in a patient with myotonic dystrophytype 1.4

Muscle atrophy is a secondary effect of denerva-tion induced by nerve transection.10 The denerva-tion and reinnervation process of human mimicmuscles has not been studied in detail. Mostknowledge we have is from human limb muscles.Here, histology typically shows no degenerativechanges in the first 2 months after a nerve lesion,but considerably reduced muscle fiber diameter 4months to 1 year after denervation.11,12 Takinginto account that we only had single-point meas-urements but no time course of quantitative ultra-sound, the plots over the time since onset of thefacial palsy (cf. Fig. 4) suggest that significantmimic muscle atrophy starts about 3–7 monthsafter facial denervation. In accordance with clinicalexperience, facial reanimation is even possibleafter denervation times of >2 years and can stilllead to recovery of mimic muscle size.1 Furtherstudies should clarify whether quantitative facialmuscle ultrasound has the potential to assess facialmuscle atrophy (and thereby the probability of suc-cess of facial nerve reconstruction surgery afterlong-term denervation) better than clinical exami-nation and EMG alone.

Routine clinical grading of facial palsy is typi-cally performed by subjective grading scales likethe House–Brackmann scale or the Sunnybrookfacial grading system.13,14 The latter, like the Sten-nert index used in this study, has the advantagethat the face is also evaluated at rest and in casesof chronic facial palsy for synkinesis.7 Nevertheless,all these subjective and rater-dependent gradingsystems have the disadvantage of insufficient intra-and interrater reliability.15 Therefore, there hasbeen debate about computer-generated objectiveevaluation techniques for years, but no image anal-ysis system has been accepted clinically.16 In thisstudy the Stennert index at rest showed a signifi-cant correlation with muscle size at rest. The intra-and interrater reliability using quantitative facialmuscle ultrasound for patients with facial palsy hasto be clarified in a further study. At that time,quantitative facial muscle ultrasound may offer thepotential for further development into an objectiveand regional facial grading system.

Muscle atrophy induced by denervation isaccompanied by infiltration of interstitial fat and

FIGURE 4. Facial muscle thickness and area (A) and echo

intensity (B) in relation to days since onset of the facial palsy

for the 6 mimic muscles during facial nerve denervation (top

graph) and during reinnervation (bottom graft). Each patient

was examined only once. Each time-point is from a different

patient. For better visualization of the time course of denerva-

tion and regeneration, data are plotted as lines. (A) Blue 5 fron-

talis thickness; green 5 frontalis area; beige 5 orbicularis oculi

thickness; purple 5 orbicularis oculi area; yellow 5 orbicularis

oris thickness; red 5 orbicularis oris area; turquoise 5 depressor

anguli oris area; gray 5 depressor labii inferior; light blue 5

mentalis. (B) Blue 5 frontalis, green 5 orbicularis oculi, bei-

ge 5 orbicularis oris; purple 5 depressor anguli oris, yellow 5 de-

pressor labii inferior; red 5 mentalis.


fibrous tissue.11 Increasing amounts of fibrous tis-sue seem to be correlated with higher ultrasoundecho intensity.4,17 Echo intensity is associated nega-tively with muscle strength, especially in theelderly.18 These findings are consistent with ourresults on echo intensity in mimic muscles duringdenervation (increase of echo intensity) and rein-nervation (restoration of baseline values) in patientswith chronic facial palsy. It would be worthwhile tocorrelate echo intensity with muscle strength in afuture study during reinnervation in facial palsypatients.

A problem still to be solved is the optimal refer-ence for quantitative ultrasound measurements ofunilaterally paralyzed facial muscles. In this studywe used the contralateral healthy side as a refer-ence. Therefore, the proportional results couldhave been influenced by changes on the healthyside. Overuse of muscles on the healthy side isoften seen in clinical practice and may result inmuscle enlargement that could lead to overestima-tion of the relative atrophy on the paralyzed side.The alternative is to reference all data to normalreference data of healthy adults.3 Looking at therange of facial muscle sizes in healthy adults andthe potential influence of factors like age, gender,height, and weight, it is first mandatory to establishreference data in large cohorts with different agegroups and a balanced gender ratio. To define thebest reference system will be even more importantfor the echo intensity data. In this study the meanecho intensity values of the facial muscles variedon the healthy side between 94 und 169. In con-trast, Alfen et al. measured mean values of between6 and 50.4 Echo intensity is very dependent on thepreset of the ultrasound machine and selection ofthe ROI. Therefore, direct comparison of data isnot possible. A good solution for the future wouldbe to report echo intensity data in reference to theecho intensity of a muscle that is not innervated bythe facial nerve. This is the reason we included thetemporalis and masseter in the present study. Both

chewing muscles are direct neighbors of the facialmuscles; that is, they are easy to access in the sameexamination. In addition, the results show thatmuscle size and echo intensity of both chewingmuscles seem not to be influenced by the facialpalsy, and thus both muscles would be suitable asreference muscles.

In conclusion, we recommend adding quantitativefacial ultrasonography to the available clinical testsfor patients with chronic facial palsy, as it is antici-pated to be of substantial value both in diagnosis andmonitoring of neuromuscular disorders of the face.The minimally invasive nature, easy access, cost-effectiveness, and reproducibility of ultrasonographymake it an ideal tool that offers many advantagesover computed tomography, magnetic resonanceimaging, or even biopsies for obtaining informationabout mimic muscle vitality and function.

REFERENCES

1. Guntinas-Lichius O, Streppel M, Stennert E. Postoperative functionalevaluation of different reanimation techniques for facial nerve repair.Am J Surg 2006;191:61–67.

2. Volk GF, Pantel M, Guntinas-Lichius O. Modern concepts in facialnerve reconstruction. Head Face Med 2011;6:25.

3. Volk GF, Wystub N, Pohlmann M, Finkensieper M, Chalmers HJ,Guntinas-Lichius O. Quantitative ultrasonography of facial muscles.Muscle Nerve 2013;47:878–883.

4. Alfen NV, Gilhuis HJ, Keijzers JP, Pillen S, van Dijk JP. Quantitativefacial muscle ultrasound: feasibility and reproducibility. Muscle Nerve2013;48:375–380.

5. Grosheva M, Wittekindt C, Guntinas-Lichius O. Prognostic value ofelectroneurography and electromyography in facial palsy. Laryngo-scope 2008;118:394–397.

6. Volk GF, Klingner C, Finkensieper M, Witte OW, Guntinas-LichiusO. Prognostication of recovery time after acute peripheral facialpalsy: a prospective cohort study. BMJ Open 2013;3:pii: e003007. doi:10.1136/bmjopen-2013-003007.

7. Stennert E, Limberg CH, Frentrup KP. [An index for paresis anddefective healing—an easily applied method for objectively determin-ing therapeutic results in facial paresis (author’s transl.)]. HNO1977;25:238–245.

8. Scholten RR, Pillen S, Verrips A, Zwarts MJ. Quantitative ultrasonog-raphy of skeletal muscles in children: normal values. Muscle Nerve2003;27:693–698.

9. Kaylie DM, Wax MK, Weissman JL. Preoperative facial muscle imag-ing predicts final facial function after facial nerve grafting. Am J Neu-roradiol 2003;24:326–330.

10. Gutmann E. The denervated nerve. Prague: Czechoslovak Academyof Sciences; 1962.

11. Bowden RE, Gutmann E. Denervation and re-innervation of humanvoluntary muscle. Brain 1944;67:273–313.

Table 2. Muscle echo intensity at rest.

MusclePalsy side

[mean (SD)]Healthy side[mean (SD)]

Ratiopalsy/healthy

side* P†

Frontalis 169.83 (12.98) 167.57 (13.59) 1.5% 0.108Orbicularis oculi 174.29 (13.67) 169.04 (13.85) 3.6% 0.212Orbicularis oris 113.41 (20.88) 94.41 (16.61) 22.7% 0.002Depressor anguli oris 141.39 (21.45) 120.57 (21.24) 20.7% 0.001Depressor labii inferioris 157.64 (24.11) 142.72 (26.86) 12.2% 0.004Mentalis 128.24 (23.24) 115.57 (27.75) 17.4% 0.008Temporalis 141.48 (14.75) 141.35 (14.39) 0.4% 0.502Masseter 120.63 (23.91) 120.34 (25.07) 2.5% 0.852

*Ratio (%) 5 [(palsy side / healthy side) 100] – 100.†Statistically significant P-values (P< 0.05) in bold.


12. Bowden RE, Gutmann E. Clinical value of muscle biopsies. Lancet1945;2:768–771.

13. House JW, Brackmann DE. Facial nerve grading system. OtolaryngolHead Neck Surg 1985;93:146–147.

14. Ross BG, Fradet G, Nedzelski JM. Development of a sensitive clinicalfacial grading system. Otolaryngol Head Neck Surg 1996;114:380–386.

15. Vrabec JT, Backous DD, Djalilian HR, Gidley PW, Leonetti JP, MarzoSJ, et al. Facial Nerve Grading System 2.0. Otolaryngol Head NeckSurg 2009;140:445–450.

16. Nakata S, Sato Y, Gunaratne P, Suzuki Y, Sugiura S, Nakashima T.Quantification of facial motion for objective evaluation using a high-speed three-dimensional face measurement system—a pilot study.Otol Neurotol 2006;27:1023–1029.

17. Pillen S, Tak RO, Zwarts MJ, Lammens MM, Verrijp KN, Arts IM,et al. Skeletal muscle ultrasound: correlation between fibrous tissueand echo intensity. Ultrasound Med Biol 2009;35:443–446.

18. Watanabe Y, Yamada Y, Fukumoto Y, Ishihara T, Yokoyama K, YoshidaT, et al. Echo intensity obtained from ultrasonography images reflect-ing muscle strength in elderly men. Clin Interv Aging 2013;8:993–998.


quantitative ultrasonography of facial muscles in patients with chronic facial palsy

Documents