Value of high-frequency color Doppler ultrasound for preorbital mass diagnosis

Abstract

Purpose:

To assess the diagnostic efficacy of the 18 MHz Color Doppler ultrasound (CDU) probe (L8-18i) in distinguishing benign and malignant preorbital masses and compare its performance with the 11 MHz probe (L3-12).

Methods:

A retrospective study analyzed ultrasound data from 17 patients (7 male, 10 female; age 5–80 years) with preorbital masses between May 2021 and April 2024. Patients underwent imaging with both probes, assessing mass size, internal echo characteristics, and blood flow parameters. Results were compared with postoperative pathology using Wilcoxon signed-rank and Mann–Whitney U tests.

Results:

Mass locations included the upper eyelid (4), lower eyelid (10), inner canthus (1), brow (1), and conjunctiva (1). Anteroposterior diameters measured by both probes showed no significant difference (P = 0.320), nor did internal echo characteristics (P = 0.891). However, blood flow signals differed significantly (P = 0.014). Pathology revealed benign lesions (chronic inflammation, nevi, cysts, hemangiomas) and malignant tumors (sebaceous gland carcinoma, basal cell carcinoma, squamous cell carcinoma). The L8-18i probe detected significant differences in peak systolic and end-diastolic velocities between benign and malignant masses (P = 0.048), while the L3-12 probe did not.

Conclusions:

The 18 MHz CDU probe demonstrated superior visualization of mass structure and blood flow, improving differentiation between benign and malignant lesions. It enhanced clarity in detecting hemorrhage, necrosis, and septation, offering higher sensitivity and accuracy than the 11 MHz probe, aligning more closely with pathological findings.

The preorbital region of the face, comprising the anterior third of the orbit, encompasses key anatomical structures, including the brow arch, eyelids, lacrimal gland, lacrimal sac, and conjunctiva.[1,2,3] Preorbital masses are prevalent ophthalmic disorders, with a spectrum of pathologies spanning benign cysts to malignant tumors.[4,5,6,7] Early and accurate diagnosis of preorbital masses is essential to guide treatment strategies and improve patient prognosis; however, the intricate anatomy of the preorbital area, coupled with mass heterogeneity, presents diagnostic challenges. Conventional imaging modalities, such as computed tomography (CT) and magnetic resonance imaging (MRI), provide essential anatomical detail but cannot generally be used to evaluate preorbital lesion tissue composition and dynamic internal vascular characteristics.[8,9] Consequently, there is a growing demand for advanced diagnostic tools that provide refined, functional insights to allow effective differentiation of benign from malignant lesions, particularly at early diagnostic stages.

Color Doppler ultrasound (CDU) is a noninvasive imaging technique extensively used in ophthalmic diagnostics.[10,11,12,13,14] Relative to other imaging modalities, CDU can both provide detailed anatomical information on orbital masses and real-time hemodynamic activity visualization.[15,16,17] Recent technological advances have led to broader adoption of high-frequency ultrasound probes, contributing to significant progress in orbital mass diagnostics.[18,19] High-frequency probes provide enhanced spatial resolution, allowing precise visualization of fine tissue structures and improving evaluation of small masses in the orbit and preorbital regions.[20,21] Prior studies have highlighted the benefits of CDU in evaluating orbital tumors and vascular lesions[22,23]; however, there is a lack of data on the value of high-frequency probes for dynamic assessment and diagnosis of complex preorbital masses.

The aims of this study were to systematically assess the clinical value and application potential of an 18 MHz high-frequency CDU probe for addressing current gaps in preorbital mass diagnosis. Using high-resolution ultrasound imaging, we aimed to accurately delineate both the morphological and histological characteristics of preorbital masses and investigate the feasibility of detecting early malignancy markers via advanced hemodynamic analysis. Through analyses of clinical case data, we assessed the strengths and potential limitations of the 18 MHz probe in differentiating benign from malignant preorbital masses. Our findings provide robust empirical evidence that can assist ophthalmologists in clinical practice, while promoting the broader application and improvement of high-frequency ultrasound technology in diagnosing ophthalmic tumors. We also sought to expand the role of ultrasound imaging in ophthalmology and introduce novel approaches to advance imaging-based diagnosis of preorbital masses.

Methods

Ethical approval

This research was authorized by the Research Ethics Committee of our hospital (protocol No.KY-2025006). The procedures complied with the principles of the Declaration of Helsinki. After the purpose and procedure of the study were explained, all the subjects signed written informed consent forms.

Study subjects

This study included 17 patients (17 eyes) with preorbital masses who underwent CDU examination at our hospital between May 2021 and April 2024. All cases were confirmed via pathological diagnosis. The study group comprised 7 males and 10 females, aged 5–80 years (mean age, 53.8 years).

Equipment

Ultrasound examinations were conducted using the GE LOGIQ S8 CDU system, equipped with two linear array probes: the L8-18i and the L3-12. The L8-18i probe operates at a maximum frequency of 18 MHz in B-mode, with lateral resolution ≤2 mm, axial resolution ≤1 mm, detection depth ≥50 mm, and color flow imaging frequency 5.0 MHz. The L3-12 probe offers a maximum frequency of 11 MHz in B-mode, with lateral resolution ≤2 mm, axial resolution ≤1 mm, detection depth ≥100 mm, and color flow imaging frequency 4.2 MHz.

Methods

During the ultrasound examination, patients were positioned supine with instructions to gently close their eyes. An ultrasound coupling agent (a standard aqueous-based medical ultrasound gel compliant with China YY 0299-2016 standard) was applied liberally to the surface of the eyelid to ensure optimal acoustic coupling. For uncooperative pediatric patients, 6.5% chloral hydrate (1 ml/kg) was administered rectally and the examination conducted once the patient was asleep. Patients were examined using both the 3-12L and the high-frequency L8-18i probes. Initial B-mode ultrasound was used to determine mass location, morphology, boundaries, and acoustic characteristics, as well as to measure its anteroposterior diameter, providing a preliminary assessment of tumor features. Subsequently, Doppler ultrasound was initiated, overlaying two-dimensional color flow imaging to evaluate the distribution, morphology, and density of blood flow signals within and around the mass, thereby assessing vascular richness. The most prominent blood flow signal was identified, and pulsed Doppler was used to measure specific blood flow parameters, including peak systolic velocity, end-diastolic velocity, and resistance index.

Criteria for blood flow signal richness assessment

Blood flow signal richness was evaluated by multilayer scanning. Lesions exhibiting only faint blood flow signals, or with one to two observable signals, were categorized as having minimal blood flow. Lesions with consistently visible blood vessels across multiple layers, with two to three vessels appearing in each layer, were classified as having rich blood flow. Intermediate cases, exhibiting characteristics between these two categories, were classified as having moderate blood flow signals.

Statistical analysis

The significance of differences between benign and malignant tumors and between the L3-12 and L8-18i probes in anteroposterior diameter measurements, internal echo characteristics, and blood flow signal richness were assessed using the Wilcoxon signed-rank test, and blood flow parameters, including velocity and resistance index, were assessed using the Mann–Whitney U test. All analyses were performed using SPSS version 24.0.

Results

Characteristics of cases

In this study, 17 patients (7 male, 10 female) with preorbital masses were evaluated. Lesions were located in the right eye in 10 cases and in the left eye in 7 cases. Mass distribution included 4 in the upper eyelid, 10 in the lower eyelid, 1 at the inner canthus, 1 in the brow region, and 1 conjunctival mass [Table 1].

Table 1.

Characteristics of patients

		Sex		Age		Position of eye		Localizations
case 1		Female		48		Left		Conjunctiva
case 2		Female		72		Right		Upper eyelid
case 3		Male		12		Left		Upper eyelid
case 4		Male		60		Right		Inner canthus
case 5		Female		42		Left		Lower eyelid
case 6		Male		55		Right		Upper eyelid
case 7		Female		14		Left		Lower eyelid
case 8		Female		47		Right		Lower eyelid
case 9		Female		5		Right		Upper eyelid
case 10		Male		79		Right		Lower eyelid
case 11		Female		80		Left		Lower eyelid
case 12		Male		72		Right		Lower eyelid
case 13		Male		63		Left		Lower eyelid
case 14		Female		75		Right		Lower eyelid
case 15		Female		55		Right		Brow region
case 16		Male		72		Left		Lower eyelid
case 17		Female		63		Right		Lower eyelid

Measurement of anteroposterior mass diameter

Anteroposterior mass diameters were measured using CDU with both L3-12 and L8-18i probes. Using the L3-12 probe, the anteroposterior diameter ranged from 1.9 to 14.4 mm (mean ± standard deviation (SD) = 7.67 ± 3.36 mm), while the anteroposterior diameter measured using the L8-18i probe was in the range 1.7 to 16.0 mm (mean ± SD = 7.86 ± 3.34 mm). Comparison of paired anteroposterior diameters measurements obtained using the two probes indicated no significant difference between them (P = 0.320, Wilcoxon signed-rank test).

Echo characteristics and blood flow signal richness in masses

Internal echo characteristics detected by the L8-18i and L3-12 probes are presented in Table 2. There was no significant difference in internal echo characteristics detected using the two probes (Z = –0.137, P = 0.891; Wilcoxon signed-rank test). Comparison of blood flow signal richness using the L3-12 and L8-18i probes demonstrated a significant advantage for the L8-18i probe in detecting blood flow signals within masses [Table 2]. Specifically, the L8-18i probe identified rich blood flow signals in 70.6% of cases, which was a substantially higher proportion than detected using the L3-12 probe (35.3%). Further, the L8-18i probe detected minimal blood flow signals in 35.3% of cases, relative to 11.8% found with the L3-12 probe. This difference between the L8-18i and L3-12 probes in detecting blood flow signal richness within masses was significant (Z = –2.449, P = 0.014; Wilcoxon signed-rank test).

Table 2.

Comparison of L3-12 and L8-18i probe echo characteristics and blood flow signal richness in masses

Parameter		Variable		Observation		Number of Cases Detected
						L3-12 Probe		L8-18i Probe
Echoes within Mass		Homogeneity		Uniform		2		1
				Non-uniform		15		16
		Echo Intensity		Low		1		1
				Medium-Low		7		13
				Medium		8		1
				Medium-High		1		2
Blood Flow Signal				No Blood Flow Signal		5		3
				Minimal Blood Flow Signal		6		2
				Abundant Blood Flow Signal		6		12

Differences in blood flow velocity and resistance index values between benign and malignant tumors

No significant differences in blood flow parameters were detected between benign and malignant tumors using the L3-12 probe: peak systolic velocity (PSV) (P = 0.180; Mann–Whitney U test), end-diastolic velocity (EDV) (P = 0.149; Mann–Whitney U test), and resistance index (RI) (P = 0.198; Mann–Whitney U test). In contrast, the L8-18i probe revealed significant differences in PSV and EDV (both P = 0.048; Mann–Whitney U test), while no significant difference in RI was detected (P = 0.149; Mann–Whitney U test).

Pathological findings

Pathological analysis of the 17 preorbital masses identified 11 benign and 6 malignant tumors. Among the benign lesions, specific diagnoses included 1 case of chronic inflammation, 1 pigmented nevus, and 4 cystic masses (2 dermoid cysts, 1 epidermoid cyst, and 1 sweat gland cyst). Additionally, there were 3 cases of hemangioma (2 cavernous hemangioma and 1 myxoid fibrous hemangioma) as well as 1 case of seborrheic keratosis and 1 sebaceous adenoma. The malignant tumors comprised 3 sebaceous gland carcinomas, 1 basal cell carcinoma, 1 squamous cell carcinoma, and 1 mucosa-associated lymphoid tissue (MALT) lymphoma.

Imaging of representative cases using the L3-12 and L8-18i probes

Case 1

Case 1 was a patient with MALT lymphoma [characteristics of the patient in Table 1]; the clinical presentation featured multifocal, salmon-pink, minimally elevated conjunctival lesions involving all quadrants of the bulbar conjunctiva [Fig. 1a-d], consistent with the characteristic clinical appearance of conjunctival MALT lymphoma. Using the low-frequency CDU probe (L3-12), only areas with a larger anteroposterior diameter were visible, and the internal echo details of the mass were poorly defined. In contrast, the high-frequency L8-18i linear array CDU probe could clearly visualize the echogenic characteristics of the conjunctival mass in each quadrant, including its irregular shape, indistinct boundaries with adjacent tissues, and heterogeneous internal echoes, which primarily appeared as hypoechoic regions [Fig. 2]. Additionally, the high-frequency probe, with its enhanced sensitivity for blood flow detection, accurately displayed the abundant internal blood flow signals of the mass, which appeared as dot-like or strip-like flow patterns [Fig. 3], consistent with literature descriptions of the ultrasonographic characteristics of ocular adnexal MALT lymphoma.[24,25,26]

Figure 1.

Anterior segment and pathological imaging (a-d): Diffuse, salmon-pink (“fish-flesh”) e elevations with minimal height involve the fornices and bulbar conjunctiva in all quadrants. (e): HE × 100 Diffuse subconjunctival lymphoid proliferation consisted of monotonous, small-to-medium centrocyte-like cells with mildly irregular nuclei, finely dispersed chromatin, and inconspicuous nucleoli. (f): Immunohistochemistry ×200 Tumor cells were diffusely positive for CD20 and negative for CD3

Figure 2.

Comparison of two-dimensional ultrasound images (a): L3-12 probe: Only the portion of the lesion with a larger anteroposterior diameter was detected, and internal echo detail had limited clarity. (b-e): L8-18i probe: Internal echo details are clearly visible across all quadrants. The lesion exhibits an irregular shape with poorly defined boundaries relative to surrounding tissues. Internal echoes are heterogeneous, predominantly showing low echogenicity

Figure 3.

Comparison of color Doppler blood flow images generated using L3-12 and L8-18i probes (case 1) (a): Limited blood flow signals were detected within the tumor using the L3-12 probe displays. (b-d): Abundant blood flow signals, appearing as dotted or linear flow patterns, were detected across multiple quadrants lesion using the L8-18i probe

The final diagnosis, however, was established through histopathological and immunohistochemical analysis of a representative incisional biopsy specimen obtained from the conjunctival mass. The specimen revealed characteristic features of MALT lymphoma, including diffuse lymphoid proliferation in the subconjunctival region composed of monotonous, centrocyte-like tumor cells. The cells were small to medium-sized, with mildly irregular nuclei, finely dispersed chromatin, and inconspicuous nucleoli, some of which exhibited monocytoid B-cell morphology. Immunohistochemically, the tumor cells were positive for CD20 and CD79a and negative for CD3 [Fig. 1e and f]. Thus, although HFUS provided valuable preoperative anatomical and hemodynamic characterization and guided biopsy site selection, the diagnosis of MALT lymphoma was conclusively verified by histopathological examination.

Case 2

Case 2 presented with a basal cell carcinoma in the right upper eyelid [characteristics of the patient in Table 1]. The L8-18i linear array high-frequency probe allowed multiangled visualization of a round mass in the upper eyelid [Fig. 4]. Relative to imaging using the conventional L3-12 low-frequency color Doppler probe, the L8-18i probe provided clearer visualization of the internal echoes of the mass, facilitating comprehensive examination of areas difficult to assess with a low-frequency probe, particularly at the interface between the mass and normal tissue. This allowed more accurate understanding of the scope of the mass and the extent of its invasion into surrounding normal tissues, supplying clinicians with detailed imaging data essential for assessment and treatment planning [Fig. 5]. The higher blood flow frequency of the L8-18i high-frequency probe also provided a more accurate display of blood flow signals within lesions than those achieved using low-frequency probes. In Case 2, the internal echo characteristics and rich distribution of blood flow signals were consistent with the color Doppler imaging features typically associated with malignant basal cell carcinomas[27,28,29,30,31] [Fig. 6].

Figure 4.

Anterior segment and pathological imaging (a): Rounded lesion with a “rodent ulcer” appearance. (b) and (c): HE × 100. (b): Tumor cells resembled epidermal basal cells, with ovoid or short spindle-shaped appearance, and are arranged in irregular nests connected to the epidermis, with tumor cells at the periphery displaying a palisading arrangement (arrow). (c): Some cancer cells had a pseudoglandular arrangement, with luminal spaces containing light blue mucinous material (arrow)

Figure 5.

Comparison of two-dimensional ultrasound images (a): L3-12 probe: Heterogeneous internal echoes, patchy high echogenicity, poorly defined boundaries. (b-e): L8-18i probe: Multidirectional observation of lesion internal echogenicity facilitated enhanced visualization of structural details. The lesion displayed heterogeneous internal echoes with identifiable high-echogenic necrotic areas, and the boundary between the lesion and adjacent eyelid tissues was more distinct relative to visualization

Figure 6.

Comparison of color Doppler blood flow images generated using the L3-12 and L8-18i probes (Case 2) (a): The L3-12 probe revealed abundant blood flow signals within the lesion. (b-d): Application of the L8-18i probe, from multiple angles, revealed abundant internal blood flow signals

The characterization of internal vascularity held direct implications for surgical planning. Delineation of hypervascular zones aided in differentiating malignant from benign pathology, assessing biological aggressiveness, and determining the true margins of infiltration—particularly in regions where grayscale boundaries were ambiguous. Preoperative vascular mapping informed decisions regarding resection extent, anticipated intraoperative bleeding risk, and guided the design of reconstructive flaps by identifying dominant feeding vessels. In scenarios where vascular patterns suggested extensive or ill-defined infiltration, the surgical strategy could be adjusted toward wider excision or combined with intraoperative frozen section analysis to ensure clear margins. Thus, high-frequency color Doppler ultrasound not only supported diagnostic evaluation but also contributed critical hemodynamic information for individualized surgical management.

Discussion

The findings of this study demonstrate that the L8-18i probe offers greater sensitivity in detecting the richness of blood flow signals within masses, particularly in malignant tumors, than that of the L3-12 probe [Fig. 6]. The L8-18i probe detected abundant and minimal blood flow signals in masses at rates of 70.6% and 35.3%, respectively, which were both significantly higher than those achieved using the L3-12 probe. Relative to low-frequency CDU probes, high-frequency probes provide enhanced structural detail in two-dimensional imaging and more sensitive and accurate blood flow detection.[22] This increased sensitivity reduces the likelihood of missed blood flow signals, thus improving preoperative diagnostic accuracy. While abundant vascularity can be observed in both benign and malignant lesions, it is frequently associated with malignant tumors due to tumor-associated angiogenesis. In our cohort, all malignant lesions (6/6) exhibited rich vascularity on high-frequency ultrasound (HFUS), whereas only 36% of benign lesions showed a comparable degree of vascularity. The integration of additional sonographic features, such as waveform parameters (e.g. peak systolic velocity and end-diastolic velocity), significantly enhanced the differentiation between malignant and benign entities. Detailed blood flow analysis facilitates understanding of the vascular characteristics of a mass, enhancing early detection of malignancies and accurate differentiation between benign and malignant lesions. Hence, high-frequency CDU probes offer superior accuracy and clinical value in distinguishing between benign and malignant tumors. Consistently, our data indicate that blood flow velocity measurement using the L8-18i probe may provide adjunctive information valuable for differentiating between benign and malignant tumors.

High-frequency probes can also precisely reveal fine internal structures within masses, such as hemorrhage, necrosis, and septations, and clearly illustrate the boundary relationships between a mass and its adjacent tissues [Fig. 5]. Additionally, high-frequency CDU is advantageous for dynamic monitoring of blood flow distribution within a mass and assessment of vascular structures, providing essential imaging reference data for preoperative diagnosis and postoperative follow-up.

The compact size of the L8-18i probe enables flexible, multiangled observation of the internal echoes and blood flow of a lesion, thereby reducing the risk of information loss in small preseptal masses that may be obscured by larger ultrasound probes [Figs. 5 and 6]. This flexibility enhances the accuracy of preoperative assessment and is particularly beneficial in sensitive areas, such as the eyelids, where the high resolution and blood flow sensitivity of the probe substantially aid preoperative evaluation and differential diagnosis of basal cell carcinoma and other malignancies.[32]

Preorbital masses are typically located relatively superficially and can occur in areas including the brow, eyelids, lacrimal gland region, lacrimal sac, and conjunctiva. In this study, eyelid masses were most prevalent, with 14 cases observed. Preorbital masses are generally smaller than those found in the intraorbital area. The L8-18i probe employed in this study enabled detection of masses as small as 1.7 mm in anteroposterior diameter, which was the smallest conjunctival lesion.

Traditional B-mode ultrasound operates at a frequency of approximately 10 MHz, while low-frequency probes used in CDU typically function at around 11 MHz. Due to their relatively low frequency, these devices have limited capacity for displaying the internal structure of smaller masses, which can hinder preoperative assessment of lesion pathological characteristics.

Overall, the L8-18i linear array high-frequency probe demonstrated superiority over the low-frequency L3-12 probe in imaging preorbital masses, and particularly in assessing the structural details of malignant tumors, their relationships with surrounding tissues, and their blood flow characteristics; however, the limited penetration of the L8-18i high-frequency probe presents a challenge when visualizing posterior structures and blood flow characteristics of larger masses, which may impact diagnostic accuracy. For lesions with a larger anteroposterior diameter, although high-frequency probes provide excellent resolution and detailed blood flow imaging, their limited depth of penetration can hinder complete visualization of deeper structures, potentially leading to incomplete assessment of the full depth and extent of the mass. Combining imaging data from low- and high-frequency probes could offer an effective solution to address this limitation. Although low-frequency probes lack the resolution of high-frequency probes, their greater penetration depth enables effective visualization of larger masses and their posterior structures. By leveraging the complementary strengths of high- and low-frequency probes, a more comprehensive assessment of the full extent and pathological characteristics of a mass becomes possible. This combined frequency approach could enhance diagnostic accuracy and thoroughness, particularly for masses with larger anteroposterior diameters, and likely holds significant clinical value. Further, combining high-frequency and low-frequency probes with other imaging modalities (e.g. MRI or CT) could also mitigate the limitations of high-frequency probes in visualizing deeper structures. This multimodal imaging approach has potential to deliver comprehensive, precise imaging support for clinical decision-making, staging, and individualized treatment planning.

This study has limitations. The sample size was limited, which may constrain the generalizability of our findings and limit their applicability in broader clinical contexts. Furthermore, this study was restricted to comparison of benign and malignant tumors and different pathological types of mass were not grouped, precluding analysis of imaging variations among different types of mass. Future studies should include a larger patient cohort encompassing diverse age groups, both sexes, and various pathological types of mass to increase sample representativeness. A larger sample size would also enhance the statistical power of the study, thereby improving the robustness of the results. Grouping masses by pathological type in future research would allow for more precise identification of specific imaging characteristics associated with each type, thereby improving diagnostic differentiation among various lesions using high-frequency CDU.

Conclusion

The findings of this study demonstrate that an 18 MHz high-frequency CDU probe provides substantial diagnostic value for imaging of preorbital masses. With its superior resolution and dynamic blood flow assessment capabilities, this high-frequency probe provides high sensitivity and specificity for characterization of preorbital masses. We recommend the 18 MHz high-frequency CDU probe is adopted as a primary screening tool for preorbital masses in clinical practice, supplemented by other imaging modalities for a comprehensive diagnosis.

Authors’ contributions

Nan Zhang and Rui Hao contributed to concept and design. Nan Zhang, Rui Hao and Jinyong Lin were involved in acquisition, analysis, or interpretation of data. Nan Zhang and Jinyong Lin drafted the manuscript. Nan Zhang, Rui Hao and Jinyong Lin performed critical revision of the manuscript for important intellectual content. Nan Zhang conducted statistical analysis. Rui Hao contributed to administrative, technical, or material support. Rui Hao were involved in supervision. All the authors read and approved the final manuscript.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Supported by Tianjin Key Medical Discipline (Specialty) Construction Project (No.TJYXZDXK-016A).