Quantitative Muscle Ultrasound: A Non‐Invasive Biomarker for Monitoring Duchenne Muscular Dystrophy

ABSTRACT

Introduction/Aims

Quantitative muscle ultrasound (QMUS) shows promise as a non‐invasive biomarker for monitoring functional status in Duchenne muscular dystrophy (DMD). We evaluated the correlation between QMUS in various muscles and functional capacity assessments. Additionally, we compared the utility of composite QMUS scores as a comprehensive measure of disease progression to that of single‐muscle analyses.

Methods

We analyzed baseline data from an ongoing prospective study involving 23 patients with DMD (aged 5–18 years; mean 10.7; SD 4.2). High‐resolution ultrasound scans of muscle groups across the upper and lower limbs, head and neck, and diaphragm were assessed for muscle echogenicity using mean grayscale values (MGV). Functional capacity assessments included the 6‐min walk test (6MWT), North Star Ambulatory Assessment (NSAA), Performance of Upper Limb (PUL) module, timed function tests, quantitative muscle strength tests, and respiratory muscle strength tests.

Results

The composite MGV for upper and lower limbs demonstrated moderate to strong negative correlations with the 6MWT and NSAA scores (correlation coefficient [CC]: −0.44 to −0.74; p < 0.01) and strong to very strong positive correlations with rise‐from‐floor, 4‐stair climb, and 10‐m walk/run times (CC: 0.65 to 0.87; p < 0.01). For upper limb function, the MGV of the flexor carpi radialis muscle was negatively correlated with PUL scores (CC = −0.35, p = 0.02). However, no significant correlations were observed between QMUS parameters and muscle strength test (p > 0.05).

Discussion

QMUS offers a non‐invasive method for assessing muscle changes and tracking disease progression in DMD.

Trial Registration: ClinicalTrials.gov identifier: NCT05249361

Abbreviations

6MWT

6‐min walk test

CC

correlation coefficient

DMD

Duchenne muscular dystrophy

MEP

maximal expiratory pressure

MGV

mean grayscale value

MGV‐arm

mean grayscale value of upper extremity muscles (average of the mean grayscale value of deltoid, biceps brachii, and flexor carpi radialis muscles)

MGV‐leg

mean grayscale value of lower extremity muscles (average of the mean grayscale value of rectus femoris, tibialis anterior, and gastrocnemius muscles)

MGV‐neck

mean grayscale value of head and neck muscles (average of the mean grayscale value of masseter, sternocleidomastoid, and geniohyoid muscles)

MIP

maximal inspiratory pressure

MRI

magnetic resonance imaging

NSAA

North Star Ambulatory Assessment

PUL

Performance of Upper Limb module

QMUS

quantitative muscle ultrasound

1. Introduction

Advances in the treatment of Duchenne muscular dystrophy (DMD) have increased the need for biomarkers capable of assessing disease progression and treatment efficacy across all stages of the illness. Clinical trials for DMD frequently utilize functional assessments, such as the 6‐min walk test (6MWT), North Star Ambulatory Assessment (NSAA), Performance of Upper Limb module for DMD (PUL), quantitative muscle strength, and timed function tests [1, 2, 3, 4, 5, 6]. While these measurement tools provide valuable insights into motor function, they are less applicable for non‐ambulatory patients in the advanced stages of DMD, thus limiting the inclusion criteria for clinical trials [5, 6]. Neurodevelopmental comorbidities such as intellectual disability, attention deficit hyperactivity disorder, and autism spectrum disorder, affecting approximately 15%–32% of patients [7, 8], can impact functional measure outcomes, which rely on patient involvement and compliance. Moreover, 20%–45% of patients with DMD experience fractures [9, 10, 11], which can hinder participation in physical function tests. A biomarker for DMD that does not depend on patient cooperation and that is suitable across all disease severities would be of value.

Quantitative muscle ultrasound (QMUS) has emerged as a promising, non‐invasive tool for objectively assessing muscle changes in DMD [12, 13, 14, 15, 16]. Disease progression in DMD is characterized by increased fat and connective tissue infiltration in affected muscles [17], altering acoustic properties and resulting in increased echogenicity of more severely affected muscles during an ultrasound scan [12, 13, 14]. The degree of echo intensity observed in a muscle ultrasound scan can indicate the pathological changes caused by DMD [12, 13, 14, 15, 16]. QMUS using grayscale level analysis has been demonstrated to be a reliable method for assessing muscle involvement [18].

Prior studies have explored the correlations between muscle echogenicity and ambulatory function in patients with DMD [19, 20, 21]. However, inconsistencies in statistical significance across these studies underscore the need for further investigation to clarify the clinical relevance of QMUS in this population [19, 20, 21]. Evidence regarding the correlation between QMUS parameters and upper limb function remains limited [22], and the associations between echogenicity in muscles beyond the appendicular system and motor function have not been well characterized. To address these gaps, we aimed to identify the relationships between the QMUS parameters across different muscles and comprehensive physical function tests (including assessments of upper limb, lower limb, and respiratory function) in both ambulatory and non‐ambulatory patients with DMD.

2. Methods

2.1. Study Design

We conducted this cross‐sectional study as part of an ongoing prospective, observational, longitudinal study investigating correlations between functional capacity and functional capability in patients with DMD (ClinicalTrials.gov: NCT05249361). The study protocol was approved by the Institutional Review Board of the Samsung Medical Center (approval number: 2022‐01‐162). Patients who were genetically diagnosed with DMD and visited the outpatient clinic of the Department of Physical Medicine and Rehabilitation at the Samsung Medical Center were deemed eligible to participate. Written informed consent was obtained from the patients and their parents or guardians. After obtaining informed consent, we included patients aged 5–18 years who were diagnosed with DMD by a genetic test. The exclusion criteria were: (1) requiring daytime ventilator assistance or using invasive mechanical ventilation through tracheostomy; (2) a history of peripheral nerve injury; (3) major surgery within the last 12 weeks or expected major surgery during the research period; (4) central nervous system disorders; and (5) cognitive impairments that would preclude adherence to study procedures. We performed a sub‐analysis of the baseline data obtained from May 2023 to July 2024.

2.2. Data Collection

2.2.1. Clinical Assessment

Demographic and clinical data were collected, including age, sex, anthropometric measurements, DMD genotype, glucocorticoid treatment history, and comorbidities.

2.2.2. Muscle Ultrasound Acquisition and Analysis

We performed high‐resolution ultrasound scans using a Samsung Medison RS85 Prestige ultrasound system (Samsung Medison Co. Ltd., Republic of Korea) with a linear transducer (2–14 MHz). We obtained transverse ultrasound images from the dominant side of the deltoid, biceps brachii, flexor carpi radialis, rectus femoris, tibialis anterior, gastrocnemius (medial head), masseter, sternocleidomastoid, geniohyoid, and diaphragm muscles. Target muscles were selected based on the literature of muscles affected in DMD [13, 23] and the practicality of obtaining the images in an outpatient setting. A standardized patient position was used for each muscle (Table S1). In patients with joint contractures, the ultrasound examination was conducted with the participants positioned as close to the standardized position as possible. One experienced physiatrist (Y. J. I.) obtained all ultrasound images using a preset with consistent imaging parameters, including gain (52 units on a 0–100 scale), dynamic range (54 dB), and transducer frequency (“Gen.” mode). The transducer was placed perpendicular to the muscle with adequate acoustic gel application and minimal compression pressure. The imaging depth was set to the minimum level (2.5–4.0 cm) required to adequately visualize both the target muscle and reference structures (bone or fascia) necessary for echo intensity assessment [24]. The ultrasound focus point was positioned at the target muscle level to optimize visualization of intramuscular architecture [25]. As the transducer angle influences echo intensity, it was carefully adjusted until the optimal echo intensity of the underlying bone or fascia was achieved. For the evaluation of the diaphragm muscle, we employed the method described by van Doorn et al. [26] Two distinct images were obtained for each muscle. We evaluated the echogenicity of the scanned muscles using both qualitative and quantitative methods.

For the qualitative assessment, we employed the Heckmatt scale, a four‐point grading system used to evaluate muscle echo intensity [27]. In muscles lacking underlying bone structures (including sternocleidomastoid, geniohyoid, and diaphragm), we applied a modified Heckmatt scale utilizing intramuscular septa as reference structures [28].

We conducted the quantitative analysis of echo intensity using ImageJ software (version 1.54 by Wayne Rasband, National Institute of Health, USA) [23]. Within each muscle image, a region of interest was manually drawn, excluding surrounding muscle fascia. We calculated the mean grayscale value (MGV), ranging from 0 to 255, for each region of interest. To ensure reliability, we obtained two separate measurements for each muscle and used the average for analysis. Additionally, composite scores were developed to provide a regional assessment of musculatures. We derived the MGV of the upper extremity muscles (MGV‐arm) from the average of the MGV of the deltoid, biceps brachii, and flexor carpi radialis muscles. Similarly, we defined the MGV of the lower extremity muscles (MGV‐leg) as the average of the MGV from the rectus femoris, tibialis anterior, and gastrocnemius (medial head) muscles. The MGV of the head and neck muscles (MGV‐neck) was calculated from the mean MGV of the masseter, sternocleidomastoid, and geniohyoid muscles.

2.2.3. Functional Capacity Assessment

We assessed functional capacity using the Vignos scale [29], Brooke scale [30], 6MWT, NSAA [31], PUL (version 2.0) [32], timed function test (rise‐from‐floor, 10 m‐walk/run, and four stair climb), quantitative muscle strength test (grip strength, elbow flexion, knee extension, and ankle dorsiflexion), and respiratory muscle strength test (maximal inspiration pressure [MIP] and maximal expiration pressure [MEP]). All functional assessments were evaluated by a well‐trained physiotherapist (Y. C.). The 6MWT was conducted using a modified version of the guidelines recommended by the American Thoracic Society [33, 34, 35]. The participants rested in a chair for at least 10 min before the test. We recorded the total distance walked within 6 min or until the participant wanted to stop. The detailed procedures for the other tests are described in Supporting Information 2.

Separate examiners independently performed the ultrasonographic measurements and functional capacity assessments, with the sonographer (Y. J. I.) blinded to the results of the functional capacity assessments.

2.3. Statistical Analysis

We calculated the descriptive statistics to summarize the sample data. We employed the Shapiro–Wilk test to assess the normality of distribution. We computed correlations of the QMUS parameters with age, qualitative echogenicity analysis, and functional capacity assessments using Spearman's rank correlation test or Kendall's tau correlation test, as appropriate, considering the presence of tied ranks. For correlations of QMUS parameters with functional capacity assessments, we calculated age‐adjusted partial correlations. Correlation strengths were categorized by the absolute value of the correlation coefficient (CC): weak (0.20–0.39), moderate (0.40–0.59), strong (0.60–0.79), and very strong (0.80–1.00). Missing data were excluded from the analysis. Statistical significance was set at p < 0.05 (two‐tailed). We conducted all analyses using the statistical software R (version 4.4.1; R Foundation for Statistical Computing, Vienna, Austria) within the RStudio environment (version 2024.4.2.764; RStudio PBC, Boston, MA, USA).

The sample size was predetermined based on the available data from the main study. We hypothesized an expected correlation coefficient of 0.66, derived from previous literature [20]. Using the G*Power software (version 3.1.9.7; Heinrich‐Heine‐Universität Düsseldorf, Düsseldorf, Germany) [36], a sample size of 15 participants with a significance level of 0.05 was estimated to provide a statistical power of 0.81 to detect the correlation in a two‐tailed test.

3. Results

3.1. Demographics and Clinical Characteristics

A total of 23 patients (mean age at evaluation 10.7 ± 4.2 years) from 21 families were enrolled in the study. Table 1 presents their detailed characteristics. In this study population, the most common mutation of the DMD gene was a point mutation (n = 11). Twenty participants were receiving corticosteroids at the time of assessment, with an average duration of 4.6 ± 3.1 years. Under standard care, the mean daily doses were 8.8 ± 3.4 mg for prednisolone (n = 7) and 17.5 ± 4.2 mg for deflazacort (n = 13). At the time of baseline evaluation, 18 participants were ambulant.

TABLE 1.

Demographic and clinical characteristics of the participants.

Characteristic	All participants (N = 23)
Age (year), mean ± SD	10.7 ± 4.2
Anthropometric measurement
Height (cm), mean ± SD	126.2 ± 14.3
Weight (kg), mean ± SD	32.0 ± 13.2
BMI (kg/m2), mean ± SD	19.5 ± 5.7
Ambulation status
Ambulant, n (%)	18 (78.3)
Non‐ambulant, n (%)	5 (21.7)
DMD gene mutation type
Deletion, n (%)	9 (39.1)
Duplication, n (%)	3 (13.0)
Point mutation, n (%)	11 (47.8)
Use of corticosteroids
Previous use, n (%)	1 (4.3)
Steroid naïve, n (%)	2 (8.7)
On oral steroid, n (%)	20 (87.0)
Comorbidities
Scoliosis	4 (17.4)
Fragility fracture	4 (17.4)
Congestive heart failure	1 (5.0)

Abbreviations: BMI, body mass index; DMD, Duchenne muscular dystrophy; SD, standard deviation.

Table 2 summarizes the results of the functional capacity assessments. The median total score for the NSAA was 14.0, whereas the median total score for the PUL was 38.0, suggesting better preservation of upper limb function compared with ambulation in this study population.

TABLE 2.

Results of functional capacity assessments in the enrolled population.

Test	N	Results
Vignos scale	23
1—5, n (%)		17 (73.9)
6—10, n (%)		6 (26.1)
Brooke scale	23
1—3, n (%)		19 (82.6)
4—6, n (%)		4 (17.4)
Total walking distance (m), 6MWT, median (IQR)	17	312.0 (267.5–366.4)
NSAA total score, median (IQR)	23	14.0 (5.0–23.5)
Timed function test
Rise from supine (s), median (IQR)	18	7.0 (5.2–14.6)
4‐stair climb (s), median (IQR)	16	5.7 (2.9–12.8)
10‐m walk/run (s), median (IQR)	17	6.3 (5.4–8.8)
PUL total score, median (IQR)	22	38.0 (33.3–40.0)
Quantitative muscle strength test
Grip (kgf), median (IQR)	19	3.8 (3.1–6.0)
Elbow flexion (kgf), median (IQR)	23	4.1 (3.1–5.4)
Knee extension (kgf), median (IQR)	23	4.6 (2.8–6.0)
Ankle dorsiflexion (kgf), median (IQR)	23	4.0 (2.7–5.1)
Respiratory strength test
MEP (cmH20), median (IQR)	20	27.0 (18.0–34.3)
MIP (cmH20), median (IQR)	20	22.0 (17.8–34.0)

Abbreviations: 6MWT, 6‐min walk test; IQR, interquartile range; MEP, maximal expiratory pressure; MIP, maximal inspiratory pressure; NSAA, North Star Ambulatory Assessment; PUL, Performance of Upper Limb module.

Age showed moderate to strong negative correlations with NSAA and PUL tests, and moderate to strong positive correlations with the four‐stair climb test and rise‐from‐floor time tests (Table S2). Other assessments including the 6MWT, 10 m‐walk/run test, quantitative muscle strength tests, and respiratory muscle strength tests did not show significant correlations with age (Table S2).

3.2. Muscle Ultrasound

The ultrasound scans were conducted on 230 muscles among the 23 participants, without the use of sedative medications, and lasted approximately 20–30 min per participant.

3.2.1. Qualitative Analysis of Muscle Echogenicity

Figure 1 and Table S3 present the outcomes of the qualitative analysis for individual muscles. The rectus femoris and tibialis anterior muscles displayed abnormal Heckmatt scores (grade ≥ 2) in all patients, suggesting they were the most frequently affected muscles in this cohort.

FIGURE 1.

Echogenicity of scanned muscles based on qualitative analysis using the original (or modified) Heckmatt scale. The bar chart displays the percentage distribution of each Heckmatt grade (Grades 1–4) across different muscle groups, arranged from left to right in descending order of mean score. Each column represents a muscle group, with the color gradient indicating the proportion of muscles within each Heckmatt grade.

3.2.2. Quantitative Analysis of Muscle Echogenicity

Figure 2 displays representative images from the quantitative grayscale level analysis of the muscle ultrasound. The rectus femoris exhibited the highest median MGV among all muscle groups (Figure 3 and Table S3). The MGV exhibited a moderate to very strong positive correlation with age across all muscle groups, indicating increased muscle echogenicity with age (Table S4). Heckmatt scale scores also showed moderate to strong positive correlations with MGV, suggesting greater echogenicity with higher grades (Table S5).

FIGURE 2.

Representative images of the muscle ultrasound and associated quantitative grayscale level analysis. (A, B) Ultrasound images of the geniohyoid muscle (arrows) in two patients (Patient #15 and #20) with North Star Ambulatory Assessment (NSAA) scores of 0. (A) Patient #15 was graded as a 3 on the modified Heckmatt scale, with a mean grayscale value of 173.87, suggesting increased echogenicity. (B) Patient #20 was graded as 2 on the modified Heckmatt scale, with a mean grayscale value of 93.01. Despite identical NSAA scores, we observed differences in their muscle echogenicity (Dg: digastric muscle). (C, D) Ultrasound images of the biceps brachii muscle (arrows) in two patients (Patient #1 and #18) with high Performance of Upper Limb (PUL) module scores. (C) Patient #1, with a PUL score of 43, showed increased echogenicity, graded as a 2 on the original Heckmatt scale, with a mean grayscale value of 114.31. This implies that changes in echogenicity may precede a noticeable decline in PUL scores. (D) Patient #18, with a PUL score of 42, exhibited lower echogenicity, graded as a 1 on the original Heckmatt scale, with a mean grayscale level of 76.52.

FIGURE 3.

Box plots showing the echogenicity of muscles examined using quantitative grayscale level analysis. Higher mean grayscale values (MGV) indicate increased echogenicity, which is associated with structural changes in muscle tissue. Muscles are presented in descending order of median MGV from left to right.

3.2.3. Correlations Between Quantitative Muscle Ultrasound and Functional Capacity Assessments

Among the QMUS parameters, both the MGV‐leg (mean MGV of the rectus femoris, tibialis anterior, and gastrocnemius muscles) and MGV‐arm (mean MGV of the deltoid, biceps brachii, and flexor carpi radialis muscles) displayed statistically significant correlations with five functional assessments (Table 3). The MGV‐leg and MGV‐arm demonstrated strong negative correlations with the total walking distance in the 6MWT and moderate negative correlations with the NSAA total score (Table 3). Both MGV‐leg and MGV‐arm showed strong to very strong positive correlations with the time required for the four‐stair climb, 10‐m walk/run, and rise‐from‐floor task (Table 3). These findings indicate that increased echogenicity of the upper and lower limb muscles was associated with poor ambulatory function. Neither the MGV‐leg nor the MGV‐arm exhibited significant correlations with the PUL total score (Table 3). The MGV‐neck (mean MGV of the masseter, sternocleidomastoid, and geniohyoid muscles) was correlated with the NSAA and four‐stair climb time, but the correlation coefficients were lower than those observed for MGV‐leg and MGV‐arm (Table 3). The MGV of the diaphragm did not correlate significantly with any of the functional tests (Table 3).

TABLE 3.

Age‐adjusted correlation coefficients between composite quantitative muscle ultrasound measures and functional capacity assessments.

Test	N	MGV of QMUS
		MGV‐leg		MGV‐arm		MGV‐neck		MGV‐diaphragm
		CC	p	CC	p	CC	p	CC	p
6MWT (m)	17	−0.74	0.001**	−0.68	0.004**	−0.29	0.27	0.24	0.38
NSAA, total score	23	−0.44	0.004**	−0.53	< 0.001***	−0.35	0.02*	−0.09	0.58
Rise from floor (s)	18	0.76	< 0.001***	0.65	0.005**	0.23	0.39	−0.20	0.45
4‐stair climb (s)	16	0.83	< 0.001***	0.71	0.003**	0.68	0.005**	−0.10	0.71
10‐m walk/run (s)	17	0.87	< 0.001***	0.78	< 0.001***	0.45	0.08	−0.13	0.63
PUL, total score	22	−0.13	0.42	−0.28	0.07	−0.13	0.41	0.06	0.73

Note: *0.01 ≤ p < 0.05, **0.001 ≤ p < 0.01, ***p < 0.001.

Abbreviations: 6MWT, 6‐min walk test; CC, correlation coefficient; MGV, mean grayscale value; MGV‐arm, average of the mean grayscale value of the deltoid, biceps brachii, and flexor carpi radialis muscles; MGV‐diaphragm, mean grayscale value of diaphragm muscle; MGV‐leg, average of the mean grayscale value of rectus femoris, tibialis anterior, and gastrocnemius muscles; MGV‐neck, average of the mean grayscale value of masseter, sternocleidomastoid, and geniohyoid muscles; NSAA, North Star Ambulatory Assessment; PUL, Performance of Upper Limb module; QMUS, quantitative muscle ultrasound.

At the level of individual muscle groups, the MGV of gastrocnemius (medial head), deltoid, and flexor carpi radialis muscle exhibited weak to strong correlations with five functional assessments but was generally weaker than those observed for the MGV‐leg, except for the NSAA (Tables 3 and 4). Notably, the MGV of the flexor carpi radialis was the only QMUS parameter that showed a significant correlation with the PUL total score, demonstrating a weak negative correlation (Table 4).

TABLE 4.

Age‐adjusted correlation coefficients between quantitative muscle ultrasound of individual muscle groups and functional capacity assessments.

Test	N		MGV of QMUS
			RF	TA	GCM	D	BB	FCR	M	SCM	GH
6MWT (m)	17	CC	−0.39	−0.72	−0.54	−0.65	−0.41	−0.63	−0.34	0.10	−0.56
		p	0.13	0.001**	0.03*	0.007**	0.11	0.009**	0.20	0.73	0.02*
NSAA, total score	23	CC	−0.13	−0.30	−0.40	−0.45	−0.38	−0.43	−0.21	−0.19	−0.28
		p	0.39	0.05	0.01*	0.003**	0.01*	0.005**	0.18	0.22	0.07
Rise from floor (s)	18	CC	0.43	0.70	0.56	0.58	0.33	0.49	0.07	−0.06	0.52
		p	0.09	0.002**	0.02*	0.01*	0.20	0.04*	0.77	0.81	0.03*
4‐stair climb (s)	16	CC	0.68	0.65	0.57	0.77	0.37	0.58	0.54	0.44	0.68
		p	0.008**	0.01*	0.03*	0.001**	0.19	0.03*	0.048*	0.12	0.007**
10‐m walk/run (s)	17	CC	0.56	0.68	0.71	0.76	0.39	0.64	0.31	0.13	0.65
		p	0.02*	0.004**	0.002**	0.001**	0.14	0.008**	0.24	0.64	0.006**
PUL, total score	22	CC	−0.01	−0.19	−0.18	−0.24	−0.15	−0.35	−0.10	−0.08	−0.19
		p	0.97	0.23	0.24	0.13	0.35	0.02*	0.52	0.61	0.22

Note: *0.01 ≤ p < 0.05, **0.001 ≤ p < 0.01, ***p < 0.001.

Abbreviations: 6MWT, 6‐min walk test; BB, biceps brachii; CC, correlation coefficient; D, deltoid; FCR, flexor carpi radialis; GCM, gastrocnemius medial head; GH, geniohyoid; M, masseter; MGV, mean grayscale value; NSAA, North Star Ambulatory Assessment; PUL, Performance of Upper limb module; RF, rectus femoris; SCM, sternocleidomastoid; TA, tibialis anterior; QMUS, quantitative muscle ultrasound.

The quantitative muscle strength tests and pulmonary strength tests revealed no correlations with the MGV of the corresponding muscles (Table 5).

TABLE 5.

Age‐adjusted correlation coefficient between the quantitative muscle ultrasound of the corresponding muscle groups and quantitative strength tests.

Test	N	Mean grayscale values
		Muscle	CC	p
Quantitative muscle strength
Grip strength (kgf)	19	Flexor carpi radialis	−0.36	0.15
Elbow flexion (kgf)	23	Biceps brachii	−0.04	0.86
Knee extension (kgf)	23	Rectus femoris	−0.09	0.69
Ankle dorsiflexion (kgf)	23	Tibialis anterior	0.04	0.84
Respiratory muscle strength
MEP (cmH20)	20	Diaphragm	−0.07	0.78
MIP (cmH20)	20	Diaphragm	0.21	0.39

Abbreviations: CC, correlation coefficient; MEP, maximal expiratory pressure; MIP, maximal inspiratory pressure; QMUS, quantitative muscle ultrasound.

4. Discussion

This study provides evidence for the potential of QMUS as a clinically relevant imaging biomarker in DMD. We noted correlations between the MGV of the upper and lower limb muscles and established functional measures, including 6MWT, NSAA, and timed function tests. Additionally, the MGV of flexor carpi radialis demonstrated a weak negative correlation with the PUL total score. As these functional assessments are commonly employed in clinical trials for DMD [6], our findings suggest that QMUS could serve as a valuable complementary tool for evaluating disease progression and functional status in patients with DMD.

Previous studies exploring correlations between QMUS and ambulatory function (6MWT, NSAA, and timed function tests) in patients with DMD have shown variable statistical significance [19, 20, 21]. In our study, both the MGV‐arm and MGV‐leg demonstrated correlations with the 6MWT, NSAA, and timed function tests that were more robust than those observed using the MGV of a single muscle. Physical function tests evaluate overall motor performance rather than individual muscle function. As such, composite values from the muscles in each region of the body may provide a more comprehensive representation of function in patients with DMD. Additionally, the varying degrees of echogenicity across different muscles within individual patients highlight the need for comprehensive assessments in protocols for evaluating muscle ultrasound. Our study revealed a weaker correlation between the QMUS and NSAA compared with the correlation between the QMUS and 6MWT. This could be due to the floor effect of the NSAA in the advanced stages of the disease, as evidenced by five patients in our cohort who scored 0 on the NSAA. Notably, although the age‐related increase in MGV‐arm and MGV‐leg showed a reduced slope after 12 years of age (Figure S1), the values remained variable among individuals: MGV‐arm varied from 121.7 to 155.8, MGV‐leg from 134.3 to 166.5, and MGV‐neck from 93.6 to 140.7. This suggests that the QMUS can continue to reflect disease progression, even when the NSAA scores have reached their minimum, implying its utility in the long‐term monitoring of severely affected patients [6].

Although limited research has explored the relationships between the QMUS and PUL test outcomes, several cross‐sectional studies have demonstrated correlations between the upper extremity muscle fat fraction, as measured by magnetic resonance imaging (MRI), and the total PUL scores [37, 38]. In our study, the total PUL scores showed a weak negative correlation with the MGV of the flexor carpi radialis—a distal muscle—compared with the deltoid or biceps brachii in the upper limb. This correlation may be attributed to the broader range of echogenicity values recorded in the flexor carpi radialis muscle, in contrast to the deltoid and biceps muscles, which tend to be affected earlier and more severely in the disease process. In these severely affected proximal muscles, echogenicity may plateau or even decline with age due to extensive fibrofatty replacement at an advanced stage [39], thereby weakening its correlation with upper extremity functional assessments. Moreover, the lack of significant correlations between the PUL score and QMUS parameters (except for flexor carpi radialis muscle) could be due to the characteristics of our cohort, which mostly comprised individuals with relatively preserved upper extremity function. Despite the observation of abnormally increased echogenicity (Heckmatt scale: 2 or higher) in most upper extremity muscles, 19 patients maintained a Brooke scale score of 1 or 2, with a mean PUL total score of 38.1. This finding indicates that alterations in muscle echogenicity may precede detectable functional deficits as measured by the Brooke scale and PUL test (Figure S1). In patients with DMD, the PUL total score has previously demonstrated a marked decline only after the loss of ambulation [40]. In contrast, increased echogenicity of the biceps brachii muscle has been reported to increase in patients with DMD as young as 7 years old [13]. This temporal difference underscores the potential of QMUS to identify subclinical muscle alterations prior to the onset of functional deficits, especially in the upper extremities of ambulant patients. Consequently, QMUS can offer a sensitive measure of muscle pathology across the disease spectrum of DMD from early to advanced stages.

Additionally, our findings differ from those of previous studies which demonstrated a significant correlation between quantitative muscle strength and muscle echogenicity in patients with DMD (aged 2–14 years) [20]. This discrepancy may be attributed to the complex mechanisms underlying muscle force generation, which are influenced by muscle fat degeneration and anatomical factors such as muscle fiber size and length, all of which change with pediatric growth [41]. Lerario et al. observed that in patients with DMD isometric muscle strength could increase up to approximately 8–9 years of age before beginning to decline [42]. In our study, muscle echogenicity increased with age across all muscle groups. However, quantitative muscle strength and respiratory muscle strength did not show age‐related increases. The broad age range of our participants (5–18 years) may explain the absence of such correlations, as age‐related changes in muscle strength and echogenicity likely follow different patterns in younger patients.

Another non‐invasive tool for assessing hereditary muscular diseases is an MRI [15, 43, 44]. A systematic review has demonstrated correlations between MRI measurements and motor function in ambulant patients with DMD [44]. Following micro‐dystrophin gene therapy, treated participants showed markedly lower muscle fat infiltration on MRI compared with an age‐matched natural history cohort, suggesting that MRI biomarkers may be useful for detecting treatment effects in DMD [45]. However, MRI is more expensive and may require sedation in pediatric patients. In contrast, QMUS offers advantages including ease of repeated measurements, potentially making it more suitable for regular clinical follow‐up. Whether QMUS can similarly capture treatment‐induced improvements in muscle echogenicity remains underexplored. While findings from MRI studies suggest that improvements in DMD may correlate with favorable changes in muscle composition [45], increased muscle echogenicity on ultrasound may persist despite functional recovery in certain neurological disorders [46]. Therefore, further longitudinal studies are needed to establish whether QMUS serves as a reliable biomarker of therapeutic response in patients with DMD.

Several limitations of this study should be acknowledged. First, the cross‐sectional design precludes the assessment of longitudinal changes in QMUS measures and their association with physical function over time. Second, the relatively small sample size and single‐center nature of the study may limit the generalizability of our findings. Third, the absence of reference values for QMUS from a healthy population limits direct interpretation of absolute QMUS values. However, our study focused on the relationship between QMUS parameters and functional outcomes within the same disease cohort, independent of normative reference values.

In conclusion, this study identified QMUS as a promising imaging biomarker for DMD, offering insights into structural muscle changes and functional status. Its sensitivity in detecting muscle alterations (even in advanced stages of the disease or before functional deficits become apparent) positions QMUS as a promising tool for the comprehensive assessment of muscle status across the disease spectrum in DMD. While further research is needed to fully establish its role, QMUS has substantial promise in enhancing clinical care and research efforts in DMD.

Author Contributions

Yu Jin Im: conceptualization, methodology, investigation, formal analysis, data curation, writing – original draft, writing – review and editing. Yumi Choe: data curation, formal analysis, investigation, methodology, writing – review and editing. Jiwon Lee: conceptualization, investigation, methodology, writing – review and editing. Jeehun Lee: conceptualization, investigation, methodology, writing – review and editing. Jong Geol Do: conceptualization, funding acquisition, investigation, methodology, project administration, supervision, writing – original draft, writing – review and editing. Jeong‐Yi Kwon: conceptualization, investigation, writing – original draft, writing – review and editing, project administration, supervision, methodology.

Ethics Statement

We confirm that we have read the journal's guidelines on issues involved in ethical publication, and affirm that this report is consistent with these guidelines.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Information 1.

Supporting Information 2.

Supporting Information 3.

Im Y. J., Choe Y., Lee J., Lee J., Do J. G., and Kwon J.‐Y., “Quantitative Muscle Ultrasound: A Non‐Invasive Biomarker for Monitoring Duchenne Muscular Dystrophy,” Muscle & Nerve 72, no. 4 (2025): 606–615, 10.1002/mus.28469.

Data Availability Statement

The data are not publicly available due to privacy or ethical restrictions. Anonymized data not included in this article will be made available upon reasonable requests from qualified investigators, following the completion of the main study.