Ultrasound grayscale histogram analysis for early rheumatoid arthritis cartilage assessment

Abstract

Objective

To evaluate the feasibility of ultrasound grayscale histogram analysis in detecting metacarpal head (MH) cartilage damage in rheumatoid arthritis (RA) and its correlation with clinical parameters.

Methods

This cross-sectional study enrolled 50 RA patients and 50 healthy controls (HCs). Bilateral MH2-5 cartilages were scanned using a Mindray Resona R9 ultrasound system. Grayscale values were quantified via histogram analysis in standardized regions of interest. Intraclass correlation coefficients (ICC) assessed reliability. Mann-Whitney U tests compared groups; Spearman correlation analyzed associations with disease duration and body mass index (BMI).

Results

RA patients exhibited significantly higher grayscale values in total MH (median: 48.6 vs. 44.8, P < 0.001), MH2 (46.4 vs. 42.7, P = 0.015), and MH3 (48.6 vs. 43.4, P = 0.006) compared to HCs. ICC demonstrated excellent intra- (0.905–0.945) and inter-observer (0.861–0.908) reliability. Weak correlations were observed between grayscale values and disease duration (r = 0.120, P = 0.001) or BMI (r = − 0.136, P < 0.001).

Conclusion

Quantitative grayscale histogram analysis of the metacarpal head cartilage demonstrates promising feasibility and provides valuable imaging evidence to support the early diagnosis of rheumatoid arthritis.

Clinical trial number

Not applicable.

Introduction

Rheumatoid arthritis (RA), a chronic autoimmune disorder, is pathologically characterized by persistent synovial inflammation and progressive joint destruction [1, 2]. Cartilage degradation, recognized as an early manifestation of articular injury, serves as a critical determinant in the development of structural deformity and functional disability, ultimately leading to substantial deterioration in patient quality of life [3, 4]. Early detection and management of cartilage damage are therefore essential for improving clinical outcomes and delaying disease progression [5, 6]. Imaging modalities play a pivotal role in cartilage assessment. While conventional radiography demonstrates limited utility in direct cartilage visualization due to poor soft tissue contrast [7], magnetic resonance imaging (MRI) provides superior anatomical resolution but remains constrained by high costs and specialized hardware requirements [4, 8, 9]. In contrast, high-resolution ultrasonography has emerged as a prominent imaging modality for musculoskeletal evaluation, offering advantages including real-time dynamic imaging, absence of ionizing radiation and cost-effectiveness [10, 11]. The metacarpal head (MH) represents an optimal anatomical target for ultrasound evaluation, given its superficial location, standardized scanning planes, and established correlation with global joint functional status [5, 12, 13]. Current ultrasonographic approaches for MH cartilage assessment encompass both semi-quantitative and quantitative methodologies [14–20]. Semi-quantitative scoring systems, though time-efficient, demonstrate significant interobserver variability. Quantitative techniques focusing on cartilage thickness measurement exhibit improved reliability, yet remain susceptible to technical variations including transducer frequency selection and subjective delineation of chondral boundaries [19, 21–23]. This study aims to investigate the feasibility of grayscale histogram analysis in ultrasonographic quantification of MH cartilage damage in RA patients, proposing an objective computational approach to overcome existing methodological limitations.

Methods

Study participants

This cross-sectional study was conducted at West China Hospital of Sichuan University from April to October 2023. A total of 50 RA patients and 50 age-/sex-matched healthy controls (HCs) were prospectively enrolled. RA diagnosis was confirmed by rheumatologists according to the 2010 ACR/EULAR classification criteria. Participants underwent standardized screening by certified sonographers, with inclusion criteria requiring: (1) age 18–70 years; (2) capacity to maintain proper positioning during ultrasound examination. Exclusion criteria applied to both groups included: (1) concurrent metabolic arthritis (e.g., gout, pseudogout) or active joint infections; (2) history of traumatic injury or surgical intervention affecting hand joints; (3) severe joint deformity impairing imaging acquisition; (4) systemic inflammatory conditions other than RA (for HCs). The study protocol adhered to the Declaration of Helsinki and Good Clinical Practice guidelines, obtaining ethical approval from the Biomedical Ethics Review Committee of West China Hospital, Sichuan University (Approval No. 2023 − 1886). Written informed consent was obtained from all participants prior to enrollment.

US image acquisition

Bilateral metacarpal heads (MHs) of the second to fifth digits were systematically examined using a standardized scanning protocol. All participants maintained maximal metacarpophalangeal (MCP) joint flexion (fist-clenched position) during imaging. A certified sonographer performed dual-plane assessments (transverse and longitudinal orientations) for each MH, covering radial-to-ulnar and proximal-to-distal dimensions. Examinations were conducted using a Mindray Resona R9 ultrasound system (Mindray Medical International, Shenzhen, China) equipped with an L30-8U ultra-high-frequency linear array transducer (18–30 MHz). To optimize acoustic coupling, a standardized sonographic interface pad (Deji Huizhong Medical Equipment Co., Beijing; dimensions: 140 × 110 × 7 mm) was interposed between the transducer and skin surface, ensuring perpendicular beam orientation relative to the articular cartilage plane. Consistent acquisition parameters were maintained across all examinations: (1) depth setting: 1.0 cm; (2) overall gain: 30 dB; (3) magnification: 1.2x. Cartilage echogenicity was quantified using built-in histogram analysis software. For each MH, three consecutive measurements were obtained from the central chondral zone (excluding periosteal and cortical interfaces) within standardized regions of interest (ROIs; dimensions: 1.5 × 0.2 mm). Final grayscale values were calculated as the mean of triplicate measurements (Fig. 1). Intraclass correlation coefficients (ICC) were calculated for both intra- and interobserver variability analyses to evaluate measurement reproducibility, 20 randomly selected RA patients underwent repeat examinations 24 h post-initial assessment by a second blinded sonographer using identical equipment and protocols.

Fig. 1.

Measured ultrasound grayscale of left-hand MH5 at transverse section

Statistical analysis

All statistical analyses were performed using IBM SPSS Statistics software (version 26.0; IBM Corp., Armonk, NY). Continuous variables were expressed as median with interquartile range (IQR) due to non-normal distribution. Inter- and intra-observer reliability were assessed through ICC using a two-way mixed-effects model with absolute agreement. ICC interpretation followed established guidelines: <0.40 (poor), 0.40–0.59 (moderate), 0.60–0.79 (good), and ≥ 0.80 (excellent). Between-group comparisons were conducted using: independent sample t-test for normally distributed continuous variables; Mann-Whitney U test for non-normally distributed continuous variables; chi-square test for categorical variables. Bivariate correlation analyses were performed using: pearson’s correlation coefficient for normally distributed paired variables, otherwise spearman correlation coefficient was used; correlation strength was categorized as: |r| ≥0.80 (strong), 0.50–0.79 (moderate), 0.30–0.49 (weak), and < 0.30 (negligible). Multiple linear regression analysis to identify independent risk factors. All hypothesis tests were two-tailed, with statistical significance defined as P < 0.05.

Results

The study cohort comprised 50 RA patients and 50 HCs, with a total of 400 metacarpal head cartilage samples analyzed (8 joints per participant). Demographic and clinical characteristics are summarized in Table 1. RA patients demonstrated significantly lower BMI values compared to HCs (median (IQR): 22.2 (3.3) vs. 23.1 (4.5), P = 0.002), while no significant intergroup differences were observed in gender distribution and age (both P > 0.05). Quantitative ultrasound analysis and multiple linear regression analysis revealed statistically significant correlations between cartilage grayscale values and clinical parameters in RA patients, duration of disease and BMI are independent risk factors for it. (Table 2; Fig. 2). Negative correlation with BMI: r=-0.136 (95% CI: -0.205 to -0.065, P < 0.001), and positive correlation with disease duration: r = 0.120 (95% CI: 0.049–0.189, P = 0.001). Both correlations were classified as very weak according to predefined strength categories. No significant associations were detected between grayscale values and other laboratory/clinical markers, including CRP, Anti-CCP, RF, and DAS28-CRP scores (all P > 0.05).

Table 1.

Demographic and clinical characteristics of RA patients and healthy controls (n = 50)

Characteristic	RA patients	HCs	P-value
Gender (Male/Female), n	3/47	6/44	0.487
Age (years)	48.0 (15.3)	50.0(20.0)	0.898
BMI (kg/m2)	22.2(3.3)	23.1(4.5)	0.002
Duration of disease (months)	21.0(60.0)	/	/
CRP (mg/L)	4.4(13.9)	/	/
Anti-CCP (U/ml) (< 8.0/8.0-500.0/>500.0), n	7/19/24	/	/
RF (IU/ml) (< 20.0/>20.0), n	0/50	/	/
DAS28-CRP	3.5(3.0)	/	/

Data presented as median (IQR) unless specified. RA: rheumatoid arthritis; HCs: healthy controls; BMI: body mass index; CRP: c-reactive protein; Anti-CCP: anti-cyclic citrullinated peptide antibody; RF: rheumatoid factor; DAS28-CRP: disease activity score-28 with CRP

Table 2.

Correlation between MH cartilage grayscale values and clinical parameters in RA patients

Parameter	r	95%CI lower	95%CI upper	P-value
Age (years)	0.037	-0.035	0.108	0.300
Duration of disease (months)	0.120	0.049	0.189	0.001
BMI(kg/m2)	-0.136	-0.205	-0.065	< 0.001
CRP(mg/L)	-0.024	-0.095	0.048	0.500
Anti-CCP(U/ml)	0.005	-0.067	0.076	0.895
RF(IU/ml)	0.047	-0.024	0.118	0.181
DAS28-CRP	-0.040	-0.111	0.031	0.255

CI: confidence interval; BMI: body mass index; CRP: c-reactive protein; Anti-CCP: anti-cyclic citrullinated peptide antibody; RF: rheumatoid factor; DAS28-CRP: disease activity score-28 with CRP

Fig. 2.

Multivariate linear regression

Reliability analysis

The intra- and interobserver reliability of grayscale measurements demonstrated excellent agreement across all assessments (Table 3). ICC exceeded 0.80 for both transverse and longitudinal sectional measurements (all P < 0.001). Notably, transverse sections exhibited significantly higher grayscale values compared to longitudinal sections (P < 0.001).

Table 3.

Reliability of ultrasound grayscale measurements in RA patients

Assessment type	Section	ICC	95%CI	P-value
Intra-observer	Transverse	0.945	0.926–0.960	< 0.001
	longitudinal	0.905	0.872–0.930	< 0.001
Inter-observer	Transverse	0.908	0.877–0.932	< 0.001
	longitudinal	0.861	0.815–0.897	< 0.001

ICC: intraclass correlation coefficient; CI: confidence interval

Comparative analysis

A total of 400 MH cartilage samples (8 joints per participant) from 50 RA patients and 50 HCs were analyzed, with 100 samples per joint (MH2-MH5) in each group. As shown in Table 4, RA patients exhibited significantly elevated grayscale values at MH2 (P = 0.015), MH3 (P = 0.006), and overall MH2-5 measurements (P < 0.001) compared to HCs. No significant differences were observed at MH4 (P = 0.590) and MH5 (P = 0.180).

Table 4.

Comparative grayscale values between RA patients and HCs (n = 100)

Joint site	grayscale values		P-value
	RA patients	HCs
MH2-5	48.6(15.1)	44.8(9.5)	< 0.001
MH2	46.4(15.8)	42.7(8.5)	0.015
MH3	48.6(13.0)	43.4(9.0)	0.006
MH4	50.1(18.0)	46.2(8.3)	0.590
MH5	47.5(17.4)	46.8(9.5)	0.180

MH: metacarpal head; RA: rheumatoid arthritis; HCs: healthy controls

Discussion

Our findings demonstrate significantly elevated ultrasound grayscale values in the MH cartilage of RA patients compared to healthy controls, particularly at MH2 and MH3 sites. The quantitative grayscale measurement exhibited excellent intra- and interobserver reliability (ICC > 0.80), with transverse sections yielding systematically higher values than longitudinal orientations. Notably, we identified a very weak positive correlation between grayscale intensity and disease duration (r = 0.120, P = 0.001), alongside a weak inverse association with BMI (r=-0.136, P < 0.001) in RA patients.

These observations align with prior investigations documenting preferential cartilage degradation at MH2-3 in RA patients through semi-quantitative scoring and thickness measurements [7, 24]. This is likely attributable to their highest functional demand and exposure to sustained mechanical stress, which increases their susceptibility to microtrauma. While semi-quantitative assessments offer rapid clinical implementation, their subjective nature introduces substantial interrater variability. Quantitative thickness measurement, though more objective, remains technically constrained by transducer frequency selection and operator-dependent boundary delineation [19, 21–23]. Crucially, both conventional methods rely on precise identification of the cartilage-synovial interface obscured in early disease stages by inflammatory edema-induced boundary blurring [25]. Our grayscale quantification approach circumvents this limitation by assessing intrinsic tissue echogenicity rather than morphological changes.

Ultrasonographic visualization of hyaline cartilage typically reveals homogeneous hypoechogenicity due to its organized collagen-proteoglycan matrix [26–28]. RA-driven synovitis initiates extracellular matrix degradation through proteoglycan depletion and collagen network disruption, increasing cartilage permeability and water content [29, 30]. These compositional alterations enhance ultrasound backscattering, manifested as elevated grayscale values preceding macroscopic structural damage [31]. Our findings of increased MH cartilage echogenicity in RA patients likely reflect these early biochemical changes, providing a potential imaging for preclinical cartilage deterioration.

The demonstrated high reliability of grayscale quantification (ICC > 0.80 across all measurements) establishes its technical robustness for longitudinal monitoring. Contrary to previous reports linking obesity with cartilage degeneration [8, 32], we observed an inverse BMI correlation, possibly reflecting RA-specific pathophysiology characterized by progressive weight loss and accelerated joint destruction in advanced stages. The weak association with disease duration may suggest non-linear temporal progression of echogenicity changes, warranting further investigation.

Our study still has several limitations. Initially, the built-in histogram analysis function of the instrument was utilized for data analysis. However, specific details regarding the underlying software were not available. Two, single-center design with limited sample size restricts generalizability. Finally, this study focused solely on assessing the feasibility of the novel method; consequently, comparisons with established assessment methods like semi-quantitative scoring were not undertaken and remain a subject for future research. As other imaging modalities are not the primary modality for rheumatoid arthritis diagnosis, direct comparisons with MRI were not performed. Future studies should expand the sample size and employ multimodal imaging techniques. Additionally, these findings could be further validated through other assessment methods in future studies. Incorporating various assessment methods through a multi-center design would facilitate clinical translation.

Conclusion

Quantitative grayscale histogram analysis of the metacarpal head cartilage demonstrates promising feasibility and provides valuable imaging evidence to support the early diagnosis of rheumatoid arthritis.

Abbreviations

MH

Metacarpal head

RA

Rheumatoid arthritis

HCs

Healthy controls

ICC

Intraclass correlation coefficient

BMI

Body mass index

MRI

Magnetic resonance imaging

MCP

Metacarpophalangeal joint

IQR

Interquartile range

CRP

C-reactive protein

Anti-CCP

Anti-cyclic citrullinated peptide antibody

RF

Rheumatoid factor

DAS28-CRP

Disease activity score-28 with CRP

CI

Confidence interval

Author contributions

Yujia Yang, Xi Xiang, Linyan Zhang, Bihui Zhu, Zhenling Zhou contributed to provide cases.

Funding

This work was supported by Sichuan International Science and Technology Innovation Cooperation/Hong Kong, Macao and Taiwan Science and Technology Innovation Cooperation Program (2022YFH0112); and Panzhihua Municipal Guiding Science and Technology Program Projects (2024ZD-S-12).

Data availability

The data used to support the findings of this study are available from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study protocol adhered to the Declaration of Helsinki and Good Clinical Practice guidelines, obtaining ethical approval from the Biomedical Ethics Review Committee of West China Hospital, Sichuan University (Approval No. 2023 − 1886). Written informed consent was obtained from all participants prior to enrollment.

Competing interests

The authors declare no competing interests.