Advancements and challenges of ultrasound imaging in the management of thyroid-associated ophthalmopathy

Abstract

Thyroid-associated ophthalmopathy (TAO), an autoimmune disorder closely associated with thyroid dysfunction, requires timely diagnosis and ongoing accurate evaluation to improve patient outcomes. With the global incidence of TAO increasing and significantly affecting the quality of life of patients, there is an urgent need for effective diagnostic tools. As a noninvasive imaging technique, ultrasound plays a pivotal role in diagnosing and managing TAO, particularly in the early detection of and monitoring of disease progression. Despite its advantages, ultrasound faces challenges such as limited resolution for deep orbital structures and a lack of standardized protocols, which can lead to diagnostic inaccuracies. This paper reviews the current status of ultrasound applications in TAO, including diagnostic utility, recent technological advances, and key challenges. It proposes strategies for future research and improvement, emphasizing analysis of ultrasound imaging data to develop biomarker stratification models. We propose an integrated multimodal framework that combines ultrasound elastography with deep learning to improve diagnostic precision.

Key Words: Thyroid; Ophthalmopathy; Ultrasound imaging; Imaging; Diagnostics

Core Tip: Thyroid-associated ophthalmopathy, an autoimmune disorder linked to thyroid dysfunction, needs timely diagnosis. Ultrasound, a noninvasive tool, aids thyroid-associated ophthalmopathy diagnosis/management but has limitations. This minireview proposes future research strategies that emphasize analysis of ultrasound data to develop biomarker models. We propose an integrated multimodal framework combining ultrasound elastography with deep learning to enhance diagnostic precision.

INTRODUCTION

Thyroid-associated ophthalmopathy (TAO), a condition closely linked to Graves’ disease, presents significant diagnostic and management challenges. With rising incidence rates and substantial impact on quality of life, early and accurate diagnosis is critical. Current diagnostic approaches rely on clinical evaluation and imaging modalities such as computed tomography (CT) and magnetic resonance imaging (MRI), which have limitations, including radiation exposure and high cost. Ultrasound has emerged as a preferred alternative because of its accessibility, noninvasiveness, and real-time monitoring capabilities. However, its utility in TAO is constrained by limited resolution for deep orbital structures, a lack of standardized protocols that introduce inter-center variability, and operator dependence. In this manuscript, we examine how emerging technologies, including ultrasound elastography and artificial intelligence (AI)-driven analysis, can address these limitations. We highlight the unique advantages of ultrasound in assessing disease activity and severity, and propose strategies to standardize protocols and enhance diagnostic precision for personalized TAO management. These strategies form the basis of the diagnostic framework detailed in subsequent sections.

HIERARCHICAL ULTRASONOGRAPHY IN DIAGNOSIS AND STRATIFIED MANAGEMENT OF TAO

Standardization of morphometric evaluation protocols

Ultrasound imaging is essential for the morphological evaluation of TAO, chiefly by measuring standardized parameters such as extraocular muscle thickness, the anteroposterior diameter of the lacrimal gland, and optic nerve sheath diameter (ONSD). Research has shown that extraocular muscle thickness is significantly increased in individuals diagnosed with TAO and strongly correlates with disease activity[1-3]. During the active phase of TAO, enlargement and thickening of the ocular muscles are primarily driven by inflammatory processes, particularly affecting the extraocular muscles[4]. This thickening, together with the proliferation of retro-orbital adipose tissue, is a major contributor to proptosis observed in TAO patients[4]. Evidence indicates that extraocular muscle thickness is markedly greater during the active phase, with the inferior rectus muscle exhibiting particularly pronounced thickening compared with the inactive phase[3]. Consequently, regular assessment of inferior rectus muscle thickness is crucial for monitoring disease progression and evaluating treatment effectiveness. Ultrasound, a noninvasive imaging modality, can accurately quantify inferior rectus muscle thickness, and its results correlate positively with MRI, thereby aiding early diagnosis and ongoing treatment assessment[5].

Furthermore, the establishment of standardized measurement protocols and consistent ultrasound techniques is essential to ensure comparability of findings across medical facilities. Integrating ultrasound with other imaging techniques, such as MRI, can yield a more comprehensive assessment of extraocular muscle condition in patients with TAO[6-8]. In patients with TAO, the condition of the optic nerve is crucial for preserving visual function[9-11]. Measurement of ONSD is an effective method for evaluating optic nerve compression, particularly in TAO, where compression is often caused by inflammation or orbital swelling, leading to enlargement of the extraocular muscles[4,12]. An increase in ONSD may indicate potential compression. Increased ONSD is associated with pathological changes of the orbit, and ultrasound measurement of ONSD is noninvasive and cost-effective, enabling physicians to determine whether the optic nerve is compressed and to correlate this finding with clinical symptoms[13]. In patients presenting with vision loss or blindness, ultrasound can be used to assess the extent of optic nerve damage.

Lacrimal gland volume is a key indicator for assessing tear-secretion function. In patients with TAO, increased volume may indicate an active disease phase[14]. Lacrimal gland function is affected by fluctuations in thyroid hormone levels, and swelling or dysfunction can exacerbate dry eye symptoms. Both ultrasonography and MRI are effective for quantifying lacrimal gland volume. Changes in volume correlate closely with TAO activity; patients with active disease typically have larger volumes[14-16]. Incorporating lacrimal gland volume with other clinical parameters can provide a comprehensive assessment of tear-secretion function in TAO and support timely intervention.

Three-dimensional ultrasound reconstruction is a promising advance in medical imaging. This technique collects multi-angle data from two-dimensional ultrasound images and uses computational algorithms to reconstruct three-dimensional anatomical structures, thereby improving the precision of volume measurements. The technology provides comprehensive anatomical insights that clarify lesion spatial relationships. Its noninvasive nature and real-time capability are particularly valuable for evaluating ocular diseases[17]. Compared with conventional imaging modalities, three-dimensional ultrasound offers detailed structural information without tissue damage, which is essential for early diagnosis and for recurrent preoperative evaluation of gallbladder disorders, uterine pathologies, and ocular conditions[18-20]. Moreover, this technology can facilitate foreign-body extraction from favorable positions and support decompression surgery in patients with TAO[21]. The real-time imaging feature enables clinicians to monitor dynamic lesion changes instantaneously, thereby improving diagnostic accuracy and supporting timely treatment decisions. With ongoing technological progress, three-dimensional ultrasound is expected to open new avenues for early diagnosis and management of TAO.

Functional ultrasound biomarkers in TAO

In functional imaging, combining the Doppler blood flow grading system with elastography stiffness thresholds provides new insights for early detection of TAO. Notably, the Adler scoring Doppler blood flow grading system, originally developed for assessing blood flow in breast cancer[22], is now widely applied across conditions, including liver cancer, cervical cancer, renal cancer, rheumatoid arthritis, white matter lesions in sickle cell disease, and carpal tunnel syndrome[23-32]. It can also evaluate microcirculatory effects in patients with coronary artery disease and chronic heart failure after surgery and endovascular myocardial revascularization[33].

Consequently, this system may effectively assess blood flow in extraocular muscles. Studies report significantly increased blood flow in extraocular muscles in patients with TAO, closely correlating with inflammatory activity[34]. Elastography stiffness thresholds also offer new opportunities for early diagnosis of TAO. Elastography is a noninvasive imaging technique that evaluates lesions by measuring tissue stiffness, with results closely linked to pathological conditions such as fibrosis and tumors[35-37]. This technique is broadly divided into two types: Strain elastography, which provides qualitative stiffness assessments[35], and shear wave elastography, which provides quantitative stiffness values[38]. Research indicates that ultrasound shear wave elastography and perfusion measurements can serve as biomarkers for tumor response[39]. For conditions requiring long-term follow-up, such as tumors, fatty liver disease, and liver cirrhosis, this technology confers notable advantages[36,39-44]. Studies show that ultrasound findings in patients with TAO may correlate positively with fibrosis, providing important evidence for assessing the degree of fibrosis in this population[45]. The primary multidisciplinary clinical applications of the Adler-Doppler blood flow grading system are detailed in Table 1.

Table 1 Multidisciplinary clinical applications of the Adler-Doppler blood flow grading system¹.

Classification	Application	Implications in TAO
Origin research	Evaluating tumor vascularity in breast cancer[43]	Basis for grading orbital inflammation
Cross-disease validation	Hemodynamic assessment in oncology (liver, cervical, renal cancers)[44-53]	Adapted for TAO microcirculation monitoring
Cardiovascular surgery	Microcirculation evaluation in CVD post-revascularization[54]	Potential for assessing TAO-related ischemia
TAO specific validation	Detection of inflammatory-driven blood flow changes in extraocular muscles[69]	Correlates with disease activity and fibrosis

¹The Adler-Doppler blood flow grading system demonstrates extensive clinical utility across multiple disciplines. Originally developed for evaluating tumor vascularity in breast cancer, this imaging modality has been validated in hemodynamic assessments of hepatic, cervical, cardiovascular interventions and renal malignancies. Recent thyroid-associated ophthalmopathy-specific investigations reveal its capability to detect inflammatory-driven blood flow augmentation in extraocular muscles.

TAO: Thyroid-associated ophthalmopathy; CVD: Cardiovascular disease.

How to overcome he challenges faced by ultrasound

More recently, advances in high-frequency probes have improved resolution, enabling measurement of extraocular muscle thickness and detection of orbital changes in real time[46]. Ultrasound is increasingly recognized for its clinical utility in monitoring disease activity and guiding management decisions. Nevertheless, broader adoption of ultrasound in TAO remains limited by several intrinsic factors, including inadequate resolution for deep orbital structures[47], a lack of standardized acquisition protocols that lead to between-center variability[48], and high operator dependence that reduces reproducibility[49].

Recent technological progress offers opportunities to address these shortcomings. Ultrasound elastography enables quantitative assessment of tissue stiffness, potentially distinguishing active inflammatory changes from chronic fibrosis[35,38]. Moreover, advances in AI and machine learning have opened new avenues for automated image analysis, reducing operator dependence and improving diagnostic consistency[50-52]. These developments may expand the role of ultrasound from a supplementary tool to a central modality for diagnosis, staging, and longitudinal monitoring of TAO.

ULTRASOUND IN TAO PATHOMECHANISMS: KEY CONTRIBUTIONS

Dynamic monitoring of inflammatory activity

The inflammatory reaction in TAO is a core component of the pathogenic mechanism. Initially, abnormal thyroid hormone levels and autoimmune responses lead to infiltration of inflammatory cells in the orbit, consisting primarily of CD4+ T cells and monocytes/macrophages. Subsequently, thyroid-stimulating hormone receptor and insulin-like growth factor-1 synergistically promote fibroblast proliferation and differentiation into adipocytes, resulting in expansion of adipose tissue within extraocular muscles and other orbital fat and connective tissues. In addition, the ability of orbital fibroblasts to synthesize hydrophilic glycosaminoglycans, predominantly hyaluronic acid, increases, contributing to edema and further expansion of orbital tissues[53,54]. Research indicates that severe active TAO is significantly associated with infiltration of CD4+ T cells and monocytes/macrophages[55,56]. The Clinical Activity Score correlates positively with the degree of lymphocyte infiltration (including total lymphocytes, T cells, and B cells) and macrophages[57-59]. Monitoring inflammatory activity can be facilitated by imaging modalities such as ultrasound, which helps determine whether the disease is in an active phase[47]. Furthermore, ultrasound plays a crucial role in evaluating inflammatory biomarkers. Vascular endothelial growth factor (VEGF) is critical for maintaining physiological vascular homeostasis across diverse cell types and tissues. It has been implicated in the molecular mechanisms underlying tumor growth and metastasis, as well as in retinal disorders linked to several blinding conditions, such as age-related macular degeneration and diabetic and hypertensive retinopathy. Pathogenic effects associated with VEGF are primarily due to its role in modulating vascular permeability and promoting neovascularization[60]. Research indicates increased vascular density in acute TAO compared with chronic TAO and control orbital fat, suggesting an angiogenic response. This pro-angiogenic and pro-lymphangiogenic microenvironment may result from heightened expression of VEGF receptors, including VEGFR-2, VEGF-A, VEGF-C, and VEGF-D. These observations suggest that periorbital edema in acute TAO may be partially driven by immature blood vessels and lymphatic capillaries with insufficient functional capacity to adequately drain the stroma[61]. Numerous studies also indicate that ultrasound can be used to evaluate abnormal blood flow. Using techniques such as color Doppler imaging and pulsed Doppler mode, together with advances in nanoparticle technology, ultrasound can quantitatively assess hemodynamic alterations in ocular tissues, thereby providing insights into VEGF expression levels and facilitating evaluation of the inflammatory response[61-63].

The advantages of ultrasound for evaluating inflammation and associated biomarkers derive from its noninvasive nature, real-time capability, and cost-effectiveness. In monitoring patients with Crohn’s disease, ultrasound can assess structural alterations within the intestine while simultaneously evaluating disease activity and treatment effectiveness by comparing dynamic changes in inflammatory markers[62]. This multifaceted approach establishes ultrasound as a valuable tool that offers immediate feedback in the management of inflammatory conditions such as TAO. Dynamic monitoring is also a major application of ultrasound in assessing inflammatory response. By performing routine ultrasound examinations alongside biomarker assessments, clinicians can achieve a comprehensive evaluation of fluctuations in the inflammatory status of the patient, facilitating timely treatment modifications. This dynamic monitoring enhances the efficiency of disease management and markedly improves patient outcomes. For example, in the management of inflammatory bowel disease, ultrasound can provide objective data on disease activity even in the absence of clinical symptoms, thereby reducing unnecessary invasive procedures[63]. The clinical importance of dynamic monitoring is evident in evaluating treatment response. During active treatment, hormone therapies and monoclonal antibodies are frequently used. Ultrasound offers real-time feedback on the patient’s response to these medications, aiding clinicians in assessing treatment efficacy. In patients receiving biologic agents for Crohn’s disease, ultrasound can track changes in intestinal wall thickness and blood flow, enabling prompt assessment of treatment outcomes and guiding subsequent therapeutic decisions[64]. These real-time feedback capabilities position ultrasound as an essential tool for managing inflammatory diseases.

Assessment of fibrosis progression

During the inactive phase of TAO, common pathological alterations include adipose tissue hyperplasia, increased synthesis of hyaluronic acid, and conversion of fibroblasts into myofibroblasts, which ultimately leads to tissue fibrosis. These changes are primarily driven by interactions among T cells, B cells, and orbital fibroblasts[53,54,65]. The fibrotic stage is characterized by irreversible structural damage, with symptoms frequently associated with mechanical compression and increased tissue stiffness. Timely identification of the transition from the active to the fibrotic phase is essential for mitigating long-term complications, enhancing prognosis, and improving quality of life for patients. Research indicates that applying color Doppler to assess blood flow parameters in the ophthalmic artery, superior ophthalmic vein, and central retinal artery may aid in differentiating between the active and inactive phases of TAO. Patients with a Clinical Activity Score of 3 or higher (indicative of the active phase) demonstrate significantly elevated arterial Doppler parameters, including peak systolic velocity and end diastolic velocity of the ophthalmic artery, as well as peak systolic velocity of the central retinal artery, whereas the maximum velocity of the superior ophthalmic vein shows a significant decrease, suggesting ongoing inflammatory activity within the orbit[66]. However, the observed increase in arterial blood flow velocity may also reflect a secondary response to intraorbital inflammation from diverse etiologies[67]. As noted above, strain elastography qualitatively assesses tissue stiffness[35], whereas shear wave elastography quantitatively evaluates fibrosis status[38]. In patients with chronic liver disease, routine ultrasound examinations facilitate monitoring of liver fibrosis progression and treatment efficacy[68]. Evidence suggests that regular ultrasound assessments significantly improve early detection of liver fibrosis progression and inform treatment decisions. Clinicians can more thoroughly evaluate hepatic changes by combining elastography (for stiffness) with B-mode ultrasound (for anatomy)[69]. At the molecular level, transforming growth factor beta 1 (TGF-β1) is a key biomarker related to fibrosis. It plays a critical role in its development[70,71].

Numerous studies have established a significant correlation between elevated serum TGF-β1 levels and fibrosis severity, indicating its potential utility in fibrosis assessment[72,73]. Therefore, combining elastography with TGF-β1 detection may provide more precise information for early diagnosis and evaluation of TAO-related fibrosis. In addition, long-term follow-up should incorporate clinical manifestations of patients and biochemical indicators to formulate personalized management plans, thereby improving outcomes and enhancing quality of life for patients. Establishing a multidisciplinary collaborative team that integrates ultrasound technology with other imaging modalities, such as MRI, can facilitate dynamic monitoring and assessment of fibrosis progression, ensuring optimal treatment and care[74].

INNOVATIONS IN CLINICAL TRANSLATION GUIDED BY ULTRASOUND

Ultrasound imaging provides real-time guidance for muscle monitoring, injections, and treatment assessment. However, challenges remain in visualizing deep orbital structures and in standardizing protocols, which limit diagnostic consistency. Future progress will require enhanced resolution, unified criteria, and integration of nanotechnology and AI to transform ultrasound into a dynamic platform for stratified therapies. This evolution will bridge precision medicine with the complexity of TAO, paving the way for personalized management and improved outcomes. This roadmap (Figure 1) positions ultrasound as a central platform for precision management of TAO and may reduce healthcare costs by 30%-40% compared with current imaging paradigms.

Figure 1 Artificial intelligence-assisted ultrasound diagnostic schematic diagram. Ocular ultrasound combined with deep learning and other artificial intelligence technologies can extract a large amount of imaging data, and construct predictive models by integrating molecular markers such as vascular endothelial growth factor and interleukin-6 to assess the disease status of thyroid-associated ophthalmopathy. Deep learning processes ultrasound data alongside vascular endothelial growth factor/interleukin-6 levels to predict disease severity. This is more real-time and convenient compared to other imaging methods such as computed tomography and magnetic resonance imaging. AI: Artificial Intelligence; VEGF: Vascular endothelial growth factor; IL: Interleukin; CT: Computed tomography; MRI: Magnetic resonance imaging.

Comparison of clinical applications of ultrasound and CT/MRI in the diagnosis of TAO

Current diagnostic approaches rely primarily on clinical evaluation and radiologic imaging. CT was widely adopted for delineating extraocular muscle enlargement and bony anatomy; however, concerns about radiation exposure limit repeated use[75]. MRI subsequently emerged as a superior tool for soft-tissue assessment, enabling visualization of orbital inflammation and disease activity[76-78]. Despite these advantages, CT and MRI are costly, less accessible for routine follow-up, involve radiation exposure (for CT), and are impractical for dynamic monitoring. Drawing on research evidence for imaging techniques in TAO, this article analyzes key characteristics of three modalities, ultrasound, CT, and MRI, from the perspectives of technical parameters and comparative medicine. Evidence derived from nine core papers (data as of 2023) focuses on six dimensions: Spatial resolution, hemodynamic evaluation, disease activity staging, bone structure evaluation, cost and accessibility, and safety. Key comparisons are outlined in Table 2.

Table 2 Comparison of clinical applications of ultrasound and computed tomography/magnetic resonance imaging in the diagnosis of thyroid-associated orbitopathy.

Technical dimension	Ultrasound	CT/MRI	Ref.
Spatial resolution	Moderate (can distinguish muscle/fat layers)	High (MRI provides superior muscle texture detail)	[1,76]
Hemodynamic assessment	Superior (real-time Doppler quantification)	Limited (requires contrast-enhanced scanning)	[46,77]
Disease activity staging	Based on hemodynamic + structural dynamics	MRI T2 signal indicates edema	[1]
Bony structure assessment	Cannot visualize bony walls	Essential (preoperative surgical decompression planning)	[76]
Cost and accessibility	Superior (portable equipment)	Higher (especially MRI)	[46,76]
Safety	Superior (non-ionizing, repeatable)	CT radiation risk; MRI metallic contraindications	[46,76]

CT: Computed tomography; MRI: Magnetic resonance imaging.

Ultrasound-guided precision interventions

In the future, ultrasound-guided orbital decompression surgery may reduce surgical complications and shorten recovery times. Ultrasound for real-time monitoring of localized drug injections also enhances procedural precision and effectiveness[79-81]. This monitoring increases treatment safety and yields data to inform personalized strategies, thereby improving overall therapeutic efficacy. As a cost-effective and efficient imaging modality, ultrasound provides real-time imaging and is suitable for repeated use to support ongoing assessment of the condition[82]. Weekly ultrasound evaluations can monitor changes in extraocular muscle thickness, guiding adjustments to hormone dosages and ensuring the safety and effectiveness of treatment regimens. This approach promotes greater personalization and accuracy in treatment and contributes to improved patient outcomes.

Convergent innovation: Ultrasound and nanotechnology

In ultrasound medicine, the integration of targeted therapy and nanotechnology is a pivotal research focus, offering innovative strategies for precise management of TAO. Sonodynamic therapy, a leading technology in this field, activates chemical agents (sonosensitizers) through nonthermal ultrasound effects, enabling superior tissue penetration and spatial selectivity. This noninvasive modality has shown significant advantages in oncology[83-85]. Advances in multifunctional nanoparticles have further expanded the potential applications of this technology. These nanoparticles have been designed as targeted drug carriers for tumors, long-acting ultrasound contrast agents, and enhancers of ultrasound-mediated drug delivery, warranting further investigation as potential anticancer agents[86]. In the future, such nanoparticles may serve as targeted carriers for TAO-related inflammatory factors, such as insulin-like growth factor-1R and interleukin-6, facilitate real-time visualization and monitoring of treatment processes as long-acting ultrasound contrast agents, and improve drug permeation across biological barriers through ultrasound-mediated cavitation. Although research on localized nanoparticle drug delivery systems for TAO remains nascent, ongoing advances in ultrasound molecular imaging and responsive nanomaterials offer promising opportunities for developing an ultrasound-guided, nanoparticle-targeted therapy paradigm for TAO.

AI-driven revolution in ultrasound imaging

Intelligent technologies have markedly improved the quality and efficiency of medical care for the diagnosis and treatment of TAO[50-52]. The integration of AI, particularly in imaging, is progressively reshaping conventional medical paradigms. Research demonstrates that deep learning systems can distinguish papilledema, normal optic discs, and various disc abnormalities by analyzing fundus photographs obtained with pharmacological mydriasis. One system achieved an area under the curve of 0.96 [95% confidence interval (CI): 0.95-0.97], with a sensitivity of 96.4% (95%CI: 93.9-98.3) and a specificity of 84.7% (95%CI: 82.3-87.1)[87]. By harnessing deep learning and big data analytics, AI can extract relevant information from extensive imaging datasets, thereby aiding healthcare professionals in making more precise diagnoses of conditions such as coronavirus disease 2019 and solid cancers, as well as in evaluating coronary artery disease[88-94]. The incorporation of AI into ultrasound imaging analysis has significantly enhanced diagnostic accuracy, enabling the detection of subtle lesions with greater precision than traditional methods and thereby reducing risks of misdiagnosis and missed diagnoses. However, variability among models may lead to errors, and a high false-positive rate could necessitate substantial changes in clinical applications[94]. AI also supports the development of personalized treatment plans through data mining, which can improve treatment outcomes and enhance quality of life for patients. Nonetheless, implementation faces challenges related to data quality and quantity; limited case numbers may adversely affect model performance, and difficulties persist in the architectural design of deep learning algorithms for medical imaging. In addition, the “black box” nature of AI reduces transparency in the decision-making process, potentially fostering skepticism among patients and healthcare providers regarding the results. Consequently, improving the interpretability of AI models represents a critical direction for future research[95,96].

CONCLUSION

Ultrasound offers advantages for TAO assessment, but key limitations remain, including resolution, operator dependence, and protocol variability. Recent technological progress offers opportunities to address these shortcomings. Ultrasound elastography enables quantitative assessment of tissue stiffness, potentially distinguishing active inflammatory changes from chronic fibrosis. Advances in AI and machine learning have opened avenues for automated image analysis, reducing operator dependence and improving diagnostic consistency. These developments may expand the role of ultrasound from a supplementary tool to a central modality for the diagnosis, staging, and longitudinal monitoring of TAO. To advance the field, combining 3D ultrasound with shear-wave elastography and microvascular Doppler, developing validated algorithms using multicenter datasets, and establishing imaging-VEGF/interleukin-6 stratification models are essential and promising steps. This tripartite approach positions ultrasound as a cost-effective, radiation-free cornerstone for TAO management, with the potential to reduce healthcare costs in future medicine.