Ultrafast ultrasound localization microscopy for differential diagnosis of reactive hyperplasia and metastatic cervical lymph nodes: a pilot study

Introduction

Cervical lymph nodes (LNs) play a pivotal role in the immune response and act as critical pathways for the metastasis of malignancies (1), especially cancers of the head and neck. Accurate preoperative differentiation of reactive versus metastatic LNs is crucial for effective clinical management (2) and can reduce unnecessary invasive biopsies and surgery. Ultrasound (US) imaging has emerged as a non-invasive and cost-effective diagnostic tool to evaluate the morphology and vascular characteristics of LNs (3). However, the sensitivity and specificity of traditional US methods are often insufficient for visualization of the microvasculature (4). Alternative imaging modalities, such as computed tomography, magnetic resonance imaging, and positron emission tomography, are effective but limited due to the high cost and incapacity to capture detailed information about the microvasculature (5).

Current US techniques, such as color Doppler flow imaging (CDFI), can provide insight into the vascularity of LNs (6), but only limited information of the microvascular (7). Superb microvascular imaging (SMI), designed to visualize microvessels without a contrast agent, offers improved sensitivity for detection of low-flow microvessels as compared to conventional Doppler methods (8). Despite these advantages, SMI and other traditional US techniques struggle to capture vessels with diameters smaller than 200 µm and flow velocities less than 0.1 cm/s (9). Contrast-enhanced US (CEUS) can highlight contrast-enhanced patterns within LNs to guide biopsy (10), but the precision to characterize the microvascular structure remains insufficient (11).

To address these limitations, ultrafast ultrasound localization microscopy (ULM) is a groundbreaking approach that utilizes super-resolution techniques to visualize microvascular structures with unprecedented clarity (12). ULM achieves spatial resolution beyond the acoustic diffraction limit, detecting vessels as small as 10–50 µm (13), a capability unmatched by other imaging modalities. This technology holds significant potential to improve the differential diagnosis of reactive versus metastatic LNs, while enabling detailed visualization of the microvascular density (MVD), microvascular diameter, microscopic patterns, and microvascular distribution. A preliminary study has shown that ULM can clearly display the microvascular structure and blood flow within LNs. Quantitative analysis of these characteristics can effectively identify metastatic LNs (14). Despite potential benefits, clinical application of ULM is limited by the lack of standardization, validation, and integration into routine diagnostic workflows. Furthermore, robust data are needed to compare the diagnostic efficacy of ULM as compared to two-dimensional (2D) US combined with SMI.

This study addresses several core questions in the evolving field of LN diagnostics: (I) can ULM significantly enhance diagnostic accuracy to differentiate reactive versus metastatic cervical LNs as compared to current imaging methods? (II) Are ULM-derived microvascular parameters, such as microvascular distribution and MVD, correlated with metastatic involvement, and can these parameters serve as independent predictive indicators? And (III) what are the clinical challenges and potential benefits of ULM integrated into routine diagnostic protocols?

Hence, the primary aim of this research was to evaluate the feasibility and clinical utility of ULM to differentiate reactive versus metastatic cervical LNs by focusing on detailed microvascular features. Additionally, the diagnostic power of ULM combined with traditional US and SMI indices was assessed to validate the potential as a superior imaging modality to differentiate reactive versus metastatic cervical LNs and to provide a foundation for future studies and clinical applications in oncology diagnostics. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2876/rc).

Methods

Study approval and patient consent

The study protocol was approved by the Institutional Review Committee of Renmin Hospital of Wuhan University (No. WDRY2024-K109) and conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Prior to inclusion in this study, informed consent was obtained from all participants.

Study cohort

The study cohort was limited to adult patients who underwent puncture biopsy of LNs at Renmin Hospital of Wuhan University between December 2023 and October 2024. The inclusion criteria were (I) US-guided puncture biopsies of LNs; (II) complete clinical information and pathological data; and (III) no contraindications for CEUS. The exclusion criteria were (I) pathological results not indicating metastatic or reactive LNs; (II) history of chemotherapy or radiotherapy; and (III) contraindications for contrast agents or biopsy. In total, 135 cases were included for analysis (Figure 1, Table 1).

Figure 1 Flowchart of patient selection. 2D US, two-dimensional ultrasound; LN, lymph node; SMI, superb microvascular imaging; ULM, ultrasound localization microscopy.

Acquisition of 2D US images

A well-trained, experienced sonographer conducted all US examinations using a Resona A20 Pro US diagnostic system (Mindray Bio-Medical Electronics Co., Ltd., Shenzhen, China) equipped with an LM18-5WU transducer operating at 5–18 MHz. The dynamic range, gain, and depth were adjusted in accordance with the manufacturer’s specifications. Standard 2D US was used to examine the LNs, prioritizing the longitudinal axis view to avoid interference from adjacent large arteries and prevent pulsation artifacts to ensure image quality, and to assess the morphology and size of the target lesions, whereas CDFI was employed to evaluate blood flow. Color gain was adjusted to display small vessels without artifacts, while avoiding compression of the nodules to ensure optimal visualization of blood flow. Recorded US characteristics included the long-to-short axis ratio (L/S ratio), presence or absence of a hilum, calcifications, cystic changes, peripheral vessels, and the maximum section showing the minimal lumen diameter near the cortex capsule.

Acquisition of SMI images

The same LM18-5WU transducer (Resona A20 Pro; Mindray Bio-Medical Electronics Co., Ltd.), operating at 5–18 MHz, was used to examine the LNs in SMI mode, while adjusting for the dynamic range, gain, and depth in accordance with the manufacturer’s specifications. Color gain was adjusted to display small vessels without artifacts, while avoiding compression of the nodules to ensure optimal visualization of blood flow. The smallest peripheral microvascular lumen diameter and vascular index (VI) were measured. The VI is an automatically calculated Doppler-derived parameter to quantify signal flow as the ratio of color pixels within the vascular lumen to the total number of pixels (15). The average value of three measurements at the areas with the richest blood flow signals in the upper, middle, and lower poles of the LNs was used for analysis. Each region of interest (ROI) was 5 mm × 5 mm.

Acquisition of ULM images

The SL10-3U transducer (Resona A20 Pro; Mindray Bio-Medical Electronics Co., Ltd.), operating at 3–10 MHz in microvascular imaging mode with a mechanical index of 0.116, was carefully placed over the LNs. A 2.0-mL bolus of SonoVue^® contrast agent (Bracco S.p.A., Milan, Italy) was intravenously administered via the antecubital vein and immediately followed by a saline flush to ensure uniform distribution. During real-time 30-s super-resolution contrast imaging, ultrafast ULM was immediately triggered upon achieving peak enhancement of the LNs. The system simultaneously activated the “microvascular imaging” mode and the patient was instructed to not to breathe during the 3-s dynamic acquisition phase. This process captured 1,500 frames at 500/s with motion correction algorithms providing real-time compensation for respiratory and tissue displacement. The system then automatically generated super-resolution images of the LNs with precise visualization of the microvascular architecture and hemodynamic parameters. If image quality was unsatisfactory, the procedure was repeated after a 20-min interval, allowing complete dissipation of contrast agent microbubbles, until satisfactory images were obtained. These images included detailed microvascular visualizations and color-coded maps showing blood flow velocity and direction. All parameters were analyzed using the manufacturer’s software with the following specific parameters:

• MVD: the ratio of the total microvascular area to the ROI within the microvascular map of the LNs. The ROI (5 mm × 5 mm) was manually selected at areas with the highest MVD in the upper, middle, and lower poles of the LNs. The average of three measurements was calculated for analysis. This approach minimizes partial volume effects from adjacent tissues and accounts for vascular distribution heterogeneity (e.g., hilar vs. peripheral zones).
• Microscopic pattern: A20 in vivo analysis software tracks individual microbubbles, which remain within the vascular space and have flow dynamics similar to red blood cells, generating super-resolution maps that show microvascular blood flow distribution, velocity, and direction. The distribution of microvessels within the LNs reflects the arrangement of the vessels. “Tree-like” refers to the distribution of microvessels along the hilum of the LN, extending outward from the inner region in a tree-like pattern. “Peripheral or disordered” indicates that the microvessels either follow the periphery of the LN or are situated centrally and internally, with a disordered blood flow direction.
• Microvascular distribution: the evenness of microvascular distribution within the largest cross-section of the LNs was assessed. Uniform distribution was defined by the absence or presence of significant microvessel-deficient areas, while non-uniform distribution was indicated by distinct regions of microvessel-deficient regions. For clinical relevance, quantitative uniformity scores were correlated with blinded visual ratings from two radiologists using a 5-point Likert scale (1= highly uniform; 5= highly irregular) (16) Strong agreement was observed (Spearman’s ρ=0.82; P<0.001).

Statistical analysis

Statistical analyses were performed using IBM SPSS Statistics for Windows (version 26.0; IBM Corporation, Armonk, NY, USA) and MedCalc software (version 22.0.1; MedCalc Software, Ltd., Ostend, Belgium). Continuous normally distributed variables are reported as the mean ± standard deviation. The independent samples t-test was used for between-group comparisons. Non-normally distributed quantitative data were analyzed with a non-parametric test and are presented as the median and interquartile range. Categorical data were analyzed using the chi-square test and are expressed as frequencies or percentages. Logistic regression analysis was used to identify US parameters associated with LN differentiation. Receiver operating characteristic curve analysis was performed to assess the diagnostic performance of the identified factors. A probability (P) value <0.05 was considered statistically significant.

Results

US image characteristics

This study included 61 reactive LNs and 74 metastatic cervical LNs. US analysis revealed no significant difference in focal hyperechogenicity or cystic change between groups (P>0.05). However, significant differences were observed in the L/S ratio, hilum presence, calcification, and peripheral blood flow (all, P<0.01) (Table 2). On 2D US, reactive LNs exhibited well-defined corticomedullary differentiation with central blood flow patterns (Figure 2A,2B). In contrast, metastatic LNs more often demonstrated round-shaped morphology, hilum absence, calcification, and peripheral blood flow patterns (Figure 2C,2D).

Figure 2 All images are of the maximum cross-sectional plane of the LNs. Images of reactive LNs: (A) 2D, (B) CDFI. Images of metastatic LNs: (C) 2D, (D) CDFI. 2D, two-dimensional ultrasound; CDFI, color Doppler flow imaging; LN, lymph node.

Microvascular quantitative indices under different parameters

US analysis revealed no significant difference in vessel diameter between CDFI and SMI (P>0.05). However, ULM demonstrated significant differences in vessel diameter, VI, MVD, microvascular flow rate, microscopic pattern, and microvascular distribution (all P<0.01) (Table 3). Reactive LNs exhibited central flow on both SMI and ULM. Compared to metastatic LNs, reactive LNs had a significantly higher VI and lower MVD (P<0.05) (Table 3; Figure 3A,3B). Conversely, metastatic LNs showed significantly lower VI and higher MVD (P<0.05) (Table 3; Figure 3C,3D). ULM revealed a finer microvascular distribution in reactive LNs, characterized by faster flow, tree-like branching extending from the hilum to the periphery, uniform distribution, and centrifugal blood flow direction (Table 3; Figure 4A,4B). In contrast, metastatic LNs displayed peripheral flow with a radial or disorganized vascular pattern extending from the periphery towards the center, along with a heterogeneous microvascular distribution. ULM further demonstrated a coarser microvascular distribution, slower flow rate, and chaotic blood flow direction in metastatic LNs (Table 3; Figure 4C,4D).

Figure 3 Vascular metrics. ROIs 123 were selected in areas of richest blood flow within the upper, middle, and lower poles of the LNs. Each ROI measured 5 mm × 5 mm. (A) VI in reactive LNs; (B) MVD in reactive LNs; (C) VI in metastatic LNs; (D) MVD in metastatic LNs. LN, lymph node; MVD, microvascular density; ROI, region of interest; VI, vascular index.

Figure 4 Flow dynamics. (A) Velocity map (reactive); (B) flow direction (reactive); (C) velocity map (metastatic); (D) flow direction (metastatic). In images (A) and (C), color coding indicates flow velocity with yellow representing higher velocity and blue indicating lower velocity. In images (B) and (D), red signifies flow towards the transducer, while blue indicates flow away from it. (A) Vmax of reactive LNs. (B) Microvascular flow direction of reactive LNs. (C) Vmax of metastatic LNs. (D) Microvascular flow direction of metastatic LNs. LN, lymph node.

Binary logistic regression analysis of clinical features of metastatic LNs

To improve differentiation of metastatic from reactive cervical LNs, binary logistic regression analysis was performed, including all variables that reached statistical significance (P<0.05). The analysis identified L/S ratio, VI, MVD, and microvascular distribution as independent predictive factors significantly contributing to distinguishing metastatic from reactive cervical LNs. An L/S ratio ≥2 [odds ratio (OR) =0.236, 95% confidence interval (CI): 0.067–0.840, P=0.026], lower VI (OR =0.938, 95% CI: 0.889–0.991, P=0.022), higher MVD (OR =1.046, 95% CI: 1.011–1.084, P=0.011), and uneven microvascular distribution (OR =29.977, 95% CI: 6.927–129.726, P<0.001) were strong indicators of metastatic LNs (Table 4). As compared to the reference group (L/S <2), L/S ≥2 was associated with a 76.4% lower risk of metastatic LNs. Each 1% decrease in VI and increase in MVD were associated with a 6.2% and 4.6% greater risk of metastasis, respectively. LNs with uneven, as compared to uniform, microvascular distribution indicated a nearly 30-fold greater risk of metastasis. These findings highlight the utility of specific US imaging characteristics to identify metastatic cervical LNs.

Comparison of diagnostic performance of imaging combinations

Diagnostic performance of various classification methods was compared based on the area under the curve (AUC) (Figure 5), sensitivity, specificity, and P values (Table 5). All methods showed statistically significant performance (all, P<0.05). The 2D + SMI + ULM combination achieved the highest AUC (0.913; 95% CI: 0.867–0.959), followed by 2D + ULM (AUC =0.900; 95% CI: 0.845–0.955) and ULM alone (AUC =0.888; 95% CI: 0.829–0.947). The 2D imaging method had the lowest AUC (0.660; 95% CI: 0.566–0.753). ULM alone had the highest sensitivity (83.8%), while combinations involving ULM (2D + ULM and 2D + SMI + ULM) maintained moderate sensitivity (78.4%) but significantly improved specificity (95.1% vs. 91.8% for ULM alone). Specificity increased progressively with modality integration, reaching 95.1% for 2D + ULM and 2D + SMI + ULM as compared to 59.0% for 2D alone. These findings suggest that multimodal approaches, especially 2D + SMI + ULM, optimize diagnostic accuracy by balancing sensitivity and specificity, with statistically and clinically meaningful improvements to AUC values.

Figure 5 Diagnostic performance evaluation of combinations of three imaging modalities. 2D, two-dimensional; ROC, receiver operating characteristic; SMI, superb microvascular imaging; ULM, ultrasound localization microscopy.

Discussion

This study demonstrates significant potential of ultrafast ULM to distinguish reactive versus metastatic LNs. ULM enhanced visualization of microvascular characteristics (i.e., vessel diameter, density, and distribution patterns) that are crucial to differentiate benign from malignant LNs. Previous research has shown the feasibility of ULM for clinical imaging and the potential to characterize LNs by detection of irregularities in the direction of local blood flow (14). The present study demonstrated that the diagnostic accuracy of ULM was superior to conventional techniques for detection of metastatic LNs, which are characterized by coarser terminal microvessels, elevated MVD, disorganized microvascular architecture, and diminished microcirculatory flow velocity. These results emphasize the potential of ULM as an advanced diagnostic tool to enhance the precision of pathological identification of LNs.

As compared to traditional imaging modalities, like CDFI and SMI, ULM excels at detecting subtle microvascular architecture changes (17). While providing valuable insights into blood flow, CDFI and SMI have limited capacity to visualize low-velocity microvascular flow or detect vessels <200 µm (18). A preliminary study has highlighted the limitations of CDFI and SMI to identify early vascular alterations to metastatic LNs (19). In contrast, the present study demonstrated that ULM overcomes these constraints by providing super-resolution imaging of vessels as small as 100 µm, enabling detailed assessment of vascular abnormalities (20). The unparalleled resolution and flow-independent tracking capabilities of ULM explain the distinct terminal microvascular diameter differences between reactive and metastatic LNs, which are undetectable by CDFI and SMI. Traditional methods measure larger parent vessels affected by hemodynamic variability, while ULM quantifies true terminal microvessels where pathological angiogenesis is most pronounced. This aligns with histopathological evidence showing that the microvasculature of metastatic LN exhibits focal hyperplasia and irregular diameters—features resolvable only at super-resolution (21). The advantages of ULM for imaging of the microvascular have garnered growing attention for evaluation of the tumor microvascular, neuroimaging (22), renal cortical microcirculation assessment (23), and diabetic complication studies (24). Future studies integrating ULM findings with immunohistochemical analyses are needed to validate these observations. The primary advantage of ULM is the ability to acquire high-frame-rate imaging data, enabling post-processing reconstruction of microvascular networks with subwavelength resolution (20). Using microbubbles as contrast agents, ULM captures detailed microvascular structures at the subwavelength level (17). Ultrafast ULM leverages ultrafast imaging techniques allowing acquisition of hundreds of frames per second, enabling high temporal resolution and accuracy (25). The Resona A20 Pro US system uses contrast-enhanced ultra-high frame acquisition to capture microbubble trajectories, followed by computational reconstruction to generate super-resolution microvascular visualization. This methodology facilitates the detection of early-stage neovascularization and hemodynamic alterations associated with malignant tumors, providing critical insights for clinical diagnosis (26). Studies have confirmed the diagnostic efficacy of ULM to distinguish benign from malignant thyroid nodules (16), breast masses (17), and gastrointestinal tract tumors (27).

ULM was utilized to identify significant differences in microvascular characteristics between reactive and metastatic LNs. Metastatic LNs displayed irregular vascular patterns, disordered blood flow direction, increased MVD, and reduced microvascular flow velocity. In contrast, reactive LNs exhibited a more uniform, tree-like distribution of microvessels, with microvascular flow directed from the hilar to the peripheral regions and higher microvascular flow velocity. Additionally, the marginal microvessels of metastatic LNs were significantly coarser than those of reactive LNs, providing further evidence of malignancy-associated neovascularization and hemodynamic alterations. This distinct microvascular architecture provides a clear differential diagnosis between reactive and metastatic LNs, which remains challenging with traditional imaging techniques.

Logistic regression analysis further confirmed the significance of ULM-derived parameters to predict the presence of metastatic LNs. The L/S ratio, VI, MVD, and microvascular distribution were all statistically significant indicators. Notably, MVD and uneven microvascular distribution were the strongest predictors of metastatic LNs. These findings highlight the capacity of ULM to enhance diagnostic clarity through precise visualization of microvascular abnormalities.

The addition of ULM significantly enhanced differentiation of reactive from metastatic LNs. When combined with 2D US, ULM improved the AUC to 0.900, demonstrating impressive diagnostic performance with 78.4% sensitivity and 95.1% specificity. Clinically, the superior AUC of multimodal approaches (e.g., 2D + SMI + ULM) reflects balanced optimization of sensitivity and specificity, while reducing both false-negative and false-positive diagnoses. This significant improvement highlights the capability of ULM to provide additional diagnostic clarity where other techniques fall short. While providing valuable morphological and vascular information, 2D US and SMI cannot accurately capture vessels <200 µm or visualize finer aspects of blood flow dynamics. The detailed microvascular structure and flow pattern visualization of ULM can improve classification of LN abnormalities.

As a key study strength, clinical application of ULM provides non-invasive, high-resolution imaging that significantly improves diagnostic accuracy. Incorporating ULM into clinical practice provides critical benefits for managing LN-related diseases. By providing non-invasive and high-resolution detection of microvascular changes, ULM can reduce the dependence on invasive procedures, like biopsies (28). This is especially relevant when it is difficult to distinguish reactive from metastatic LNs. The capability of ULM to detect early vascular changes could enable earlier and more accurate interventions, potentially improving patient outcomes. Moreover, the capability of ULM to quantify microvascular characteristics (e.g., MVD) provides a powerful tool to monitor therapeutic responses, especially in oncology, where the success of chemotherapy or radiation can be assessed based on changes to the vascularity of LNs (16).

While this study highlights the diagnostic potential of ULM, there were some limitations that must be addressed. First, the relatively small, single-center sample size may limit generalizability of these findings. Multi-center studies with larger sample sizes are needed to validate these results and evaluate the broader applicability of ULM in LN diagnostics. Additionally, the contrast agent requirement of ULM may pose challenges for patients with contraindications. Therefore, future research should explore non-contrast-based super-resolution techniques. Finally, although ULM offers superior diagnostic accuracy, clinical application remains in the early stages. Future studies should focus on standardizing ULM protocols and evaluating the cost-effectiveness in routine clinical practice. Further research into the use of ULM use in other applications—such as detecting LN recurrence or monitoring post-treatment responses—could expand utility.

Conclusions

In conclusion, ULM represents a significant advancement in differentiating reactive from metastatic LNs. By providing detailed visualization of the microvascular, ULM enhances diagnostic accuracy as a valuable tool for early detection and treatment monitoring. As technology evolves, integration of ULM into clinical practice could revolutionize LN diagnostics, ultimately improving patient care and outcomes.

Acknowledgments

We would like to thank the radiologists at Renmin Hospital of Wuhan University for blinded image assessments.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2876/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2876/dss

Funding: This study was supported by the National Natural Science Foundation of China (Nos. 81971624 and 82271999) and the Cross-Innovation Talent Project of the People’s Hospital of Wuhan University (No. JCRCYR-2022-011).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2876/coif). The authors have no conflicts of interest to declare.

Ethical Statement: 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 was approved by the Institutional Review Committee of Renmin Hospital of Wuhan University (No. WDRY2024-K109) and conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Prior to inclusion in this study, informed consent was obtained from all participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Liu J, Shao T, Zhang J, Liu Q, Hua H, Zhang H, Wang J, Luo T, Shi YE, Jiang Y. Gamma synuclein promotes cancer metastasis through the MKK3/6-p38MAPK cascade. Int J Biol Sci 2022;18:3167-77. [Crossref] [PubMed]
2. Liu Y, Chen J, Zhang C, Li Q, Zhou H, Zeng Y, Zhang Y, Li J, Xv W, Li W, Zhu J, Zhao Y, Chen Q, Huang Y, Li H, Huang Y, Yang G, Huang P. Ultrasound-Based Radiomics Can Classify the Etiology of Cervical Lymphadenopathy: A Multi-Center Retrospective Study. Front Oncol 2022;12:856605. [Crossref] [PubMed]
3. Furukawa MK, Furukawa M. Diagnosis of lymph node metastases of head and neck cancer and evaluation of effects of chemoradiotherapy using ultrasonography. Int J Clin Oncol 2010;15:23-32. [Crossref] [PubMed]
4. Cutolo M, Smith V. Detection of microvascular changes in systemic sclerosis and other rheumatic diseases. Nat Rev Rheumatol 2021;17:665-77. [Crossref] [PubMed]
5. Hu J, Yu Y, Liu W, Zhong J, Zhou X, Xi H. The survival benefit of different lymph node yields in radical prostatectomy for pN1M0 prostate cancer patients: Implications from a population-based study. Front Oncol 2022;12:953069. [Crossref] [PubMed]
6. Huang T, Huang PT, Luo ZY, Lv JF, Jin PL, Zhang T, Zhao YL, Wang Y, Hong YR. Use superb microvascular imaging to diagnose and predict metastatic cervical lymph nodes in patients with papillary thyroid carcinoma. J Cancer Res Clin Oncol 2024;150:268. [Crossref] [PubMed]
7. Ma Y, Li G, Li J, Ren WD. The Diagnostic Value of Superb Microvascular Imaging (SMI) in Detecting Blood Flow Signals of Breast Lesions: A Preliminary Study Comparing SMI to Color Doppler Flow Imaging. Medicine (Baltimore) 2015;94:e1502. [Crossref] [PubMed]
8. Seskute G, Jasionyte G, Rugiene R, Butrimiene I. The Use of Superb Microvascular Imaging in Evaluating Rheumatic Diseases: A Systematic Review. Medicina (Kaunas) 2023.
9. Fu Z, Zhang J, Lu Y, Wang S, Mo X, He Y, Wang C, Chen H. Clinical Applications of Superb Microvascular Imaging in the Superficial Tissues and Organs: A Systematic Review. Acad Radiol 2021;28:694-703. [Crossref] [PubMed]
10. Li R, Lan X, Xie X, Wu J, Hu R, Guo J. Diagnostic value of high-frame-rate contrast-enhanced ultrasound and contrast vector imaging for superficial lymph node lesions. BMC Cancer 2025;25:785. [Crossref] [PubMed]
11. Harput S, Christensen-Jeffries K, Li Y, Brown J, Eckersley RJ, Dunsby C, Tang MX. Two stage sub-wavelength motion correction in human microvasculature for CEUS imaging. 2017 IEEE International Ultrasonics Symposium (IUS); 2017.
12. Errico C, Pierre J, Pezet S, Desailly Y, Lenkei Z, Couture O, Tanter M. Ultrafast ultrasound localization microscopy for deep super-resolution vascular imaging. Nature 2015;527:499-502. [Crossref] [PubMed]
13. Hingot V, Brodin C, Lebrun F, Heiles B, Chagnot A, Yetim M, Gauberti M, Orset C, Tanter M, Couture O, Deffieux T, Vivien D. Early Ultrafast Ultrasound Imaging of Cerebral Perfusion correlates with Ischemic Stroke outcomes and responses to treatment in Mice. Theranostics 2020;10:7480-91. [Crossref] [PubMed]
14. Zhu J, Zhang C, Christensen-Jeffries K, Zhang G, Harput S, Dunsby C, Huang P, Tang MX. Super-Resolution Ultrasound Localization Microscopy of Microvascular Structure and Flow for Distinguishing Metastatic Lymph Nodes - An Initial Human Study. Ultraschall Med 2022;43:592-8. [Crossref] [PubMed]
15. Spiesecke P, Schmidt J, Peters R, Fischer T, Hamm B, Lerchbaumer MH. Assessment of quantitative microflow Vascular Index in testicular cancer. Eur J Radiol 2024;176:111513. [Crossref] [PubMed]
16. Shin T, Smyth TB, Ukimura O, Ahmadi N, de Castro Abreu AL, Ohe C, Oishi M, Mimata H, Gill IS. Diagnostic accuracy of a five-point Likert scoring system for magnetic resonance imaging (MRI) evaluated according to results of MRI/ultrasonography image-fusion targeted biopsy of the prostate. BJU Int 2018;121:77-83. [Crossref] [PubMed]
17. Shin Y, Lowerison MR, Wang Y, Chen X, You Q, Dong Z, Anastasio MA, Song P. Context-aware deep learning enables high-efficacy localization of high concentration microbubbles for super-resolution ultrasound localization microscopy. Nat Commun 2024;15:2932. [Crossref] [PubMed]
18. Foiret J, Zhang H, Ilovitsh T, Mahakian L, Tam S, Ferrara KW. Ultrasound localization microscopy to image and assess microvasculature in a rat kidney. Sci Rep 2017;7:13662. [Crossref] [PubMed]
19. Lowerison MR, Sekaran NVC, Zhang W, Dong Z, Chen X, Llano DA, Song P. Aging-related cerebral microvascular changes visualized using ultrasound localization microscopy in the living mouse. Sci Rep 2022;12:619. [Crossref] [PubMed]
20. Yan J, Huang B, Tonko J, Toulemonde M, Hansen-Shearer J, Tan Q, Riemer K, Ntagiantas K, Chowdhury RA, Lambiase PD, Senior R, Tang MX. Transthoracic ultrasound localization microscopy of myocardial vasculature in patients. Nat Biomed Eng 2024;8:689-700. [Crossref] [PubMed]
21. Zhang G, Lei YM, Li N, Yu J, Jiang XY, Yu MH, Hu HM, Zeng SE, Cui XW, Ye HR. Ultrasound super-resolution imaging for differential diagnosis of breast masses. Front Oncol 2022;12:1049991. [Crossref] [PubMed]
22. Renaudin N, Demené C, Dizeux A, Ialy-Radio N, Pezet S, Tanter M. Functional ultrasound localization microscopy reveals brain-wide neurovascular activity on a microscopic scale. Nat Methods 2022;19:1004-12. [Crossref] [PubMed]
23. Bodard S, Denis L, Hingot V, Chavignon A, Hélénon O, Anglicheau D, Couture O, Correas JM. Ultrasound localization microscopy of the human kidney allograft on a clinical ultrasound scanner. Kidney Int 2023;103:930-5. [Crossref] [PubMed]
24. Zhang H, Huang L, Yang Y, Qiu L, He Q, Liu J, Qian L, Luo J. Evaluation of Early Diabetic Kidney Disease Using Ultrasound Localization Microscopy: A Feasibility Study. J Ultrasound Med 2023;42:2277-92. [Crossref] [PubMed]
25. Yi HM, Lowerison MR, Song PF, Zhang W. A Review of Clinical Applications for Super-resolution Ultrasound Localization Microscopy. Curr Med Sci 2022;42:1-16. [Crossref] [PubMed]
26. Zhang G, Yu J, Lei YM, Hu JR, Hu HM, Harput S, Guo ZZ, Cui XW, Ye HR. Ultrasound super-resolution imaging for the differential diagnosis of thyroid nodules: A pilot study. Front Oncol 2022;12:978164. [Crossref] [PubMed]
27. Zhang C, Lei S, Ma A, Wang B, Wang S, Liu J, Shang D, Zhang Q, Li Y, Zheng H, Ma T. Evaluation of tumor microvasculature with 3D ultrasound localization microscopy based on 2D matrix array. Eur Radiol 2024;34:5250-9. [Crossref] [PubMed]
28. Huang C, Lowerison MR, Trzasko JD, Manduca A, Bresler Y, Tang S, Gong P, Lok UW, Song P, Chen S. Short Acquisition Time Super-Resolution Ultrasound Microvessel Imaging via Microbubble Separation. Sci Rep 2020;10:6007. [Crossref] [PubMed]