Contrast-free Super-Resolution Ultrasound Imaging for Microvascular Assessment of Murine Subcutaneous Tumour Models in Preclinical Oncology

Contrast-free Super-Resolution Ultrasound Imaging for Microvascular Assessment of Murine Subcutaneous Tumour Models in Preclinical Oncology | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Contrast-free Super-Resolution Ultrasound Imaging for Microvascular Assessment of Murine Subcutaneous Tumour Models in Preclinical Oncology Amy McDermott, Iman Taghavi, Natalia Perez Jimenez, Anne Skovsbo Clausen, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8989062/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract Introduction . Tumour vascularity is an important biomarker of tumour growth and therapeutic response. Current in vivo imaging methods have a limited ability to detect individual tumour vessels due to the small vessel size and low flow rates commonly observed in murine oncology models. This paper presents three pilot studies which together aim to assess the performance of a developing imaging method, Super-resolution Ultrasound using Erythrocytes (SURE), for subcutaneous tumour imaging. Methods . Across three experiments, we qualitatively compared the vascular imaging sensitivity and clinical applicability of SURE imaging with colour Doppler imaging and ex vivo immunohistochemistry. Two murine oncology models were investigated: the LNCaP (lymph node carcinoma of the prostate) xenograft model (n = 3) and the B16-F10 melanoma model (n = 7). Optimal SURE post-processing parameters for subcutaneous tumour imaging were characterised. Results . SURE outperformed Doppler imaging, detecting flow with higher spatial resolution in corresponding tissue regions. SURE successfully visualised tumour microvasculature, detecting vessels approximately 30 µm in diameter while enabling haemodynamic characterisation. Optimal post-processing involved a trade-off between vascular imaging sensitivity, final signal-to-noise ratio, and image artefact prevalence. Vascular density measurements obtained using SURE did not differ significantly from those obtained by immunohistochemistry. Conclusions . SURE represents a significant advance in in vivo microvascular imaging. While post-processing time, image artefacts, and limited quantitative analysis currently limit preclinical application, the technique demonstrates strong potential for tumour vascular characterisation in both preclinical research and clinical oncology. Super-resolution ultrasound Contrast-free ultrasound Microvascular imaging Preclinical oncology Erythrocyte tracking Murine tumour model Figures Figure 1 Figure 2 Figure 3 Figure 4 INTRODUCTION Angiogenesis plays a critical role in tumour development, as the formation of new blood vessels provides the tissue perfusion required to sustain growth. Tumour vascular characteristics, such as tortuosity, density, and haemodynamics, are specific to tumour type and stage and can change in response to treatment. Neoplastic vasculature differs markedly from that of healthy tissue. Tumour-associated vessels are typically smaller in diameter, highly tortuous, and exhibit reduced and heterogeneous flow velocities [ 1 – 3 ]. These features often place tumour microvasculature below the spatial and temporal resolution limits of current in vivo imaging modalities, hindering comprehensive clinical detection and characterisation. Consequently, quantitative assessment of tumour vascularity remains challenging, leading to reliance on surrogate measures such as gross tumour size or molecular tracer uptake in both clinical assessment and preclinical murine oncology studies [ 4 ]. This occurs despite well-established evidence that even partial vascular assessment using conventional imaging techniques, such as Doppler ultrasound, contrast-enhanced CT and MRI, provides significant clinical value for the diagnosis, treatment planning, and therapeutic monitoring of conditions involving solid tumours [ 5 – 8 ]. Emerging ultrasound-based imaging techniques, including superb-microvascular imaging and other methods that utilise microbubble contrast agents to enhance vascular detection, have demonstrated increased diagnostic value compared with conventional clinical imaging modalities [ 5 , 9 ]. In preclinical studies, the combination of microbubbles with high-frame-rate imaging systems has overcome the traditional diffraction limit of ultrasound, enabling in vivo visualisation at the capillary level in solid tumours [ 10 ]. This approach, ultra-localised microscopy (ULM), enables microvascular imaging from approximately 10 minutes of acquired data [ 11 , 12 ]. ULM can resolve tumour vessels as small as 10 um and provides comprehensive characterisation of flow direction, velocity, vessel morphology, and regional tumour perfusion [ 13 ]. The enhanced haemodynamic and vascular information afforded by ULM offers significant value in clinical oncology, having already shown promise in the monitoring of murine oncology models [ 14 ]. However, its clinical translatability remains limited by the need for prolonged imaging times and the use of exogenous contrast agents. Recent advances in contrast-free ultrasound imaging have made significant progress in addressing the translatability limitations associated with ULM. Super-Resolution Ultrasound using Erythrocytes (SURE) has emerged as a non-invasive technique that exploits erythrocyte backscattering rather than microbubble localisation [ 15 , 16 ]. This approach combines high-frame-rate imaging with specialised post-processing algorithms to amplify previously undetectable vascular signals, reducing the required data acquisition time to seconds while still enabling capillary-level imaging. To date, SURE has enabled novel in vivo microvasculature characterisation of the rat kidney and human lymph node [ 15 , 17 ]. In this study, we present pilot testing of the SURE method for preclinical oncology and optimise the technique for future translation to clinical imaging of subcutaneous tumours. We investigate SURE performance in both low- and high-vascularity tumour models and assess the impact of tumour vascularity on imaging performance. We will also present a semi-automated method for quantitative vascular density analysis using SURE images and compare these measurements with those obtained from paired immunofluorescent immunohistochemistry. METHODS Study Population Three experiments were conducted using two murine tumour models. Pilot study 1 employed an LNCaP xenograft Balb/c athymic nude mouse model (n = 3). Pilot study 2 and the main study utilised C57BL/6 mice inoculated with the B16-F10 melanoma cell line (ATCC® CRL-6475™) (n = 2 and n = 5, respectively). All animals were purchased from Janvier Labs (Le Genest-Saint-Isle, France) and housed in the University of Copenhagen core animal facility with ad libitum access to food and water. Tumours were established on the right flank and allowed to grow until approximately 1cm in diameter before inclusion in the study. Each experiment had a distinct objective, and different imaging planes were acquired accordingly. The pilot studies focused on comparing imaging performance between tumour models and identifying optimal SURE imaging parameters. The experimental designs were as follows, with no overlap in population: Pilot Study 1 One randomly selected imaging plane and one preidentified highly vascular imaging plane, identified using colour Doppler imaging, were acquired per LNCaP tumour (n = 3). Pilot Study 2 Two randomly selected ultrasound imaging planes were acquired per B16-F10 tumour (n = 2). Main Study Quantitative vascular analysis using SURE was compared with ex vivo immunohistochemistry. Three randomly selected imaging planes were selected blindly for both ultrasound imaging and histological analysis for each B16-F10 tumour (n = 5). All imaging experiments lasted approximately one hour, after which animals were euthanised by cervical dislocation. In the main study, tumours were excised postmortem and fixed in formalin for 24 hours before paraffin embedding. Ultrasound Data Collection Mice were sedated using inhaled isoflurane and maintained at 2.0%. Hair within the target scanning window was removed, and animals were immobilised in lateral recumbency. A probe holder was used to position a 168-channel GE L8-18iD linear array transducer (GE HealthCare, USA) approximately 1 cm from the skin’s surface, with ultrasound gel providing an interface. The transducer, with a wavelength pitch of 150 µm, was operated at a transmission frequency of 10 MHz. The immobilised probe position was maintained while acquiring both SURE and conventional ultrasound images from the same imaging plane. A commercial GE LOGIQ E9 scanner (GE HealthCare, USA) was used to acquire B-mode and colour Doppler images. The same transducer was then connected to a Verasonics Vantage 256 system (Verasonics, Inc., Kirkland, WA, USA) for SURE data acquisition. SURE data were acquired using a pulse inversion sequence comprising 24 emissions from 2×12 virtual sources [ 18 ]. The acquisition duration was 1 min, with a transmission voltage of 95 V. Data were acquired using both manual and automated time-gain compensation (TGC) settings to assess their impact on image quality. TGC adjustments, whether automated or manual, were performed to minimise signal clipping. All manual adjustments were completed by a single operator to reduce operator-induced bias. SURE Data Processing A windowed motion-correction method was employed, in which multiple low-motion data segments were manually selected from the 1-min acquisition. These segments were combined to produce approximately 10 s of low-motion data for subsequent processing and transverse oscillation motion correction. Each SURE scan data set was processed twice to illustrate the need for clinically tailored processing settings, similar to Doppler ultrasound methods. Singular value decomposition (SVD) processing is used for separating out tissue, flow, and noise. An upper threshold separates out tissue from flow, and a lower threshold is used for removing noise, where the lower singular values are disregarded. Specifically, the SVD range and dataset signal threshold for each scan were adjusted to optimise processing for two distinct clinical objectives: (i) high-sensitivity processing, maximising detection of small, low-flow vessels, and (ii) high clinical accuracy, minimising inclusion of erroneous background signal. Data processing was based on the method described by Jensen et al. [ 15 ], with several modifications introduced to optimise the technique for oncology applications. To address clipping artefacts, clipped values in the raw radiofrequency data were identified and set to zero across all frames as a preprocessing step prior to SURE pipeline processing. Element exclusion was determined using beamformed clipped data rather than raw radiofrequency data alone. Elements contributing to clipping within the beamformed region of interest were excluded, while elements outside this region were retained to avoid unnecessary data loss. Excessive element exclusion was observed when clipping extended into the lower part of the image; in this region, the selection criteria were refined such that elements were removed only when their contributions degraded the beamformed region. SURE Image Analysis For qualitative comparison with colour Doppler imaging, SURE images were overlaid onto corresponding B-mode images. SURE images were binarised, post-processed, and analysed using Fiji (ImageJ). Images were converted to 8-bit grayscale and smoothed using a Gaussian blur (σ = 1.0 px). Local adaptive thresholding was applied to account for spatial variations in background intensity, using the Phansalkar method (local window radius of 45 px, k = 0.25, and r = 0.5). Post-processing was minimal and aimed to improve vessel continuity and suppress residual noise. Binary closing (one iteration) was applied to bridge small discontinuities within vessels, followed by removal of isolated bright pixels using the “Remove Outliers” filter (radius 2 px). Vessel contrast was enhanced by multiplying pixel intensities by a factor of 1.4 prior to final mask extraction. Vessel-specific masks were generated by converting the binary image into a selection and applying it to the original image, with all pixels outside the selection cleared to produce a transparent vessel mask suitable for B-mode overlay. In the main study, vascular density was measured using high-sensitivity SURE scans. Images were binarised using the method described above, and the tumour boundaries were manually defined using the free region-of-interest (ROI) tool in Fiji. Significant imaging artefacts were excluded manually. Vascular density was calculated as the percentage of ROI pixels containing vascular signal relative to the total number of ROI pixels. Quantitative analysis of blood haemodynamics was also performed in the main study using the velocity-based SURE processing and analysis method outlined by Naji et al. [ 16 ]. One scan per mouse was selected based on the imaging plane demonstrating the highest detectable vascularity on both SURE and colour Doppler imaging. Quantitative data were collected from multiple vessels within each SURE velocity map to analyse inter- and intra-subject haemodynamic variation across a single tumour imaging plane. Blood velocity and vessel lumen diameter were measured at manually selected vascular locations, with at least 9 sampling points selected per SURE image. Further haemodynamics characterisation was performed by estimating the volumetric blood flow rate for each vessel using the following equations: $$Bloodflow\left(\mu L/s\right)=\frac{Velocity\left(mm/s\right)\times Area\left(\mu m\right)²}{1{0}^{9}}$$ $$VesselArea\left(\mu{m}^{2}\right)=\pi\times Radius{\left(\mu m\right)}^{2}$$ Histology Slide preparation and staining were performed by the Histology Laboratory at the Department of Biomedical Sciences, University of Copenhagen. Formalin-fixed paraffin-embedded (FFPE) samples, collected during the main study, were sectioned along the longitudinal axis, and three distinct regions were randomly selected for analysis. From each region, two slides were prepared using adjacent tissue sections. One slide was stained with hematoxylin and eosin (H&E) to guide the identification of tumour boundaries. The second slide underwent fluorescent immunohistochemistry targeting CD-31 to identify the vascular endothelium. This used anti-CD31 primary antibody (clone EPR17259; ab182981, Abcam) and an Alexa Fluor® 568-conjugated secondary antibody. Whole-section slide imaging was performed using the AxioScan Z1 (ZEISS) at x10 magnification. Quantitative analysis of vascular density in digitised fluorescent immunohistochemistry images was performed using Fiji. No post-processing of images was performed. Tumour ROIs were manually delineated in Fiji, and any evident imaging artefacts were excluded. Images were segmented using a manually selected threshold to separate the fluorescent signal from the background prior to vascular density calculation. Vascular density was calculated as the percentage of ROI pixels containing fluorescent signal relative to the total number of ROI pixels. Statistical Analysis All data analysed in this study were assessed for normality using the Shapiro-Wilk test. For comparisons between tumours, data that were not normally distributed were analysed using the Kruskal-Wallis test with Dunn’s post-hoc test, while normally distributed data were analysed using a one-way analysis of variance (ANOVA) with Tukey’s post-hoc test. The Wilcoxon signed-rank test was used to analyse paired data collected from the same tumour. Throughout the study, a p-value of < 0.05 was considered statistically significant. All statistical analyses were conducted using RStudio. Graphical representations indicate statistical significance using asterisks: *p < 0.05, **p < 0.01, and ***p < 0.001. RESULTS Feasibility Across Tumour Models In the LNCaP xenograft model, flow was detected in one tumour using both colour Doppler and SURE imaging (Fig. 1a – 1c). In this individual, the vascular signal was identifiable using both Doppler-guided and blind SURE acquisitions in matching tissue regions (Fig. 1a & 1c). In the remaining two tumours, no flow was detected using either modality. In the B16-F10 melanoma model, both colour Doppler and SURE consistently detected vascular signal across all imaging planes for both guided and blind acquisitions (Fig. 1d − 1f). These results show imaging feasibility to differ between tumour models and some consistency across ultrasound modalities. SURE and Colour Doppler Imaging Performance In the B16-F10 melanoma model, SURE detected vascular signals in all acquired scans. Across all imaging planes, SURE demonstrated comparable or improved vascular imaging compared with colour Doppler, revealing structures not detected with Doppler alone. In regions where flow was detected using colour Doppler, SURE provided enhanced spatial detail, enabling clearer delineation of vessel boundaries and branches (Fig. 1d and 1f). The higher spatial detail also allowed discrimination between small vessels in close proximity, which were not resolvable in the corresponding colour Doppler images (Fig. 2; green box overlay). Variability in Image Quality Data collected with both automated and manual TGC settings were affected by signal clipping, resulting in clipping artefacts. A small improvement in the data signal was observed in data collected with manual TGC selection, demonstrating that this parameter affects final image quality. Data collected using the manual TGC method were used for subsequent analysis and figure generation. As one operator completed all manual TGC, no analysis of operator variability can be performed. The optimal SVD and signal threshold processing settings needed to generate high-sensitivity or high-accuracy images varied between individuals and imaging planes. This could not be predicted and required screening of all combinations for each scan, similar to parameter optimisation in colour Doppler. Both high-sensitivity and high-accuracy SURE processing modes exceeded the vascular detection achieved with colour Doppler in most tissue regions (Fig. 2). High-sensitivity processing enabled detection of low-intensity vascular signals but resulted in increased background signal in deeper image regions, limiting its utility for B-mode overlay and interpretability. High-accuracy processing improved vessel contrast by reducing background signal, but reduced the detection of lower-intensity vascular signals from small vessels. Microvascular Haemodynamic Characterisation SURE velocity processing enabled identification of flow direction across the imaging plane, and point-specific measurement of blood velocity and vessel diameter in vessels with sufficient signal strength (Fig. 3a). The majority of vessels included in high-accuracy scans met the signal threshold required for haemodynamic characterisation, with the number of analysable vessels per scan ranging from 9 to 18. Blood velocity, vessel diameter and flow direction were successfully measured for all selected vessels, enabling assessment of haemodynamic variability both within individual tumours and between animals (Fig. 3b-d). Comparable haemodynamic measurements could not be obtained using spectral Doppler, as no measurable flow was detected in any imaging planes. Across B16-F10 tumours in the main study, blood velocity measurements had a mean of 2.75 mm/s (median 2.4 mm/s, range 1.2–7.5 mm/s) and differed significantly between tumours (Kruskal–Wallis χ²(4) = 11.65, p = 0.020*; Fig. 3b), with post-hoc testing identifying a significant difference in average velocity between tumours B and E. Mean vessel lumen diameter across all measurements was 121.31 µm (range 52.8–251.1 µm) and also differed significantly between tumours (one-way ANOVA F(4,72) = 4.118, p = 0.00463**; Fig. 3c), with tumour E exhibiting larger diameters than tumour B. Estimated blood flow rates showed significant differences across tumours (Kruskal–Wallis χ²(4) = 20.88, p = 0.00034***; Fig. 3d), with tumour E demonstrating higher flow rates than tumours A and B. Vascular Density Quantification and Histological Comparison Quantitative vascular density measurements derived from in vivo SURE imaging did not differ significantly from those obtained using ex vivo immunohistochemistry (Fig. 4a). Qualitatively, fluorescent immunohistochemistry revealed a greater number of small vessels and finer vascular detail compared with SURE images (Fig. 4b-c). SURE images were affected by signal banding artefacts, which persisted under both manual and automated time-gain compensation settings. The banding artefact can be seen in the lower right quadrant of Fig. 4b radiating into the image. This can be caused by any strong signal reflector (e.g., air) outside of the imaging plane. Across the main study dataset (15 measurements from five tumours), SURE-derived vascular density measurements had a mean value of 4.42% (SD 2.33%, range 1.30–8.61%). Vascular density varied between tumours but did not differ significantly across individuals (Kruskal–Wallis χ²(4) = 9.13, p = 0.058). Immunohistochemistry-derived vascular density measurements showed a mean value of 3.00% (SD 0.66%, range 2.01–4.07%), with comparatively low variability across tumours. Paired comparison using the Wilcoxon signed-rank test demonstrated no statistically significant differences between SURE- and immunohistochemistry-derived vascular density measurements within any tumour. DISCUSSION This study demonstrates the feasibility and capabilities of contrast-free SURE as an in vivo tool for analysing and assessing the vascularity of subcutaneous murine tumours. We demonstrate superior imaging resolution and microvascular detection capabilities compared with comparable methods, e.g. colour Doppler. It is confirmed that SURE can be utilised at a dataset level to characterise tumour haemodynamics (flow direction, velocity, and volumetric flow) and vascular density. We confirmed the value of this tool in structural vascular density analysis by comparing it with immunohistochemistry results obtained using a compatible method; no significant difference was observed in the final datasets. Tumour-model-dependent feasibility was observed, with both SURE and colour Doppler failing to detect vasculature in two of three LNCaP tumours, whereas all B16-F10 tumours showed detectable vasculature in every individual using both imaging methods. We hypothesise that this reflects known differences in vascular architecture and haemodynamics between the two models. LNCaP tumours are characterised by lower vascular density and slower, often heterogeneous blood flow [ 19 ]. The SURE method relies on the ability to distinguish moving erythrocyte signals from those of surrounding stationary tissues. As such, it is likely that lower flow rates impair the ability to identify LNCaP vessels, despite their presence within the spatial detection limits. Current SURE detection limits remain unknown, and rapid advances in erythrocyte tracking and motion correction methods will likely lower the limit for detectable vessels. The higher spatial resolution of SURE enabled visualisation and qualitative assessment of tumour microvasculature beyond the capabilities of conventional colour Doppler imaging. SURE provided enhanced delineation of vessel morphology, allowing clear identification of in-plane vascular branching, tortuosity, and avascular regions within tumours, features typically accessible only with contrast-enhanced ultrasound or ex vivo methods. In regions where flow was detected by both modalities, SURE frequently provided greater spatial detail, improving the visualisation of vessel boundaries and spatial relationships. Despite these advantages, colour Doppler imaging retains practical strengths. Doppler provides real-time visualisation of blood flow, whereas SURE currently requires extended data acquisition (approximately one minute) and subsequent post-processing prior to image generation. Consequently, SURE is currently less suited to rapid, real-time assessment. However, SURE offers additional functional information, including flow direction mapping comparable to that provided by Doppler imaging, while remaining contrast-free. SURE images were more frequently affected by artefacts, most prominently, signal banding, which persisted under both manual and automated TGC adjustment. The prevalence of SURE artefacts is impacted by the selection og processing parameters and is most prevalent in high-sensitivity vascular imaging. Together, these findings indicate that SURE complements rather than replaces colour Doppler, providing improved microvascular detail at the expense of increased acquisition and processing time. The clinical tailoring of SURE processing parameters and the SURE-B-mode overlay method are novel applications of the method and represent efforts to optimise the SURE method for clinical use. While tailoring of SURE processing parameters currently represents a considerable computational endeavour, rapid technical advancements are increasing the feasibility of this type of processing. The ability to tailor processing parameters to clinical aims increases the method’s usability and the potential for large-dataset preclinical use. Further studies across varied tumour models are needed to confirm whether manually tailoring (as used in this study) is always required to identify the optimal processing parameters in a data set. Microvascular haemodynamic characterisation has currently not been extensively utilised in preclinical oncology. To date, pharmacology and core research are mostly limited to structural assessment of the gross tumour, monitoring molecular tracer uptake, or imaging of large vessels. Consequently, despite knowing that perfusion and angiogenesis are key components of tumour growth, they are rarely included in oncology research due to limitations in imaging. This limitation is exemplified in this work by the failure of spectral Doppler to quantify flow in any tumour. SURE has the potential to address this capability gap. Quantitative vascular density measurements derived from SURE imaging showed agreement with those obtained using immunohistochemistry, supporting the ability of SURE to capture overall vascular burden in vivo. While no statistically significant differences were observed between the two methods, qualitative comparison demonstrated that histological imaging revealed a greater number of small vessels and finer microvascular detail. This difference is expected given the higher spatial resolution of ex vivo histological techniques and the absence of physiological motion and acoustic attenuation. The observed differences highlight the complementary nature of in vivo SURE imaging and histological analysis rather than a discrepancy between methods. Histology provides high-resolution structural detail but is limited to end-point assessment, whereas SURE enables non-invasive, in vivo visualisation of tumour vasculature under physiological conditions. As such, histology should not be considered a direct ground truth for in vivo imaging, but rather a complementary reference that contextualises the strengths and limitations of SURE. This study has several limitations. The sample size was small, reflecting the pilot nature of the work, and limits the generalisability of quantitative findings. The low sample size also resulted in an underpowered statistical analysis; consequently, all statistical tests described must be considered exploratory. SURE imaging is sensitive to physiological motion, requiring careful selection of low-motion data segments for processing. The impact of low-motion window selection on image quality was not assessed here. These limitations should be considered when interpreting the results and highlighting areas for further technical refinement. CONCLUSIONS SURE represents a significant advance in contrast-free in vivo microvascular imaging. Despite current limitations, including processing time, susceptibility to image artefacts, and restricted quantitative analysis, the findings of this study demonstrate strong potential for longitudinal assessment of tumour microvasculature in preclinical oncology models. The ability to visualise microvascular architecture and haemodynamic behaviour in vivo, without the need for exogenous contrast agents, supports repeated imaging for monitoring tumour progression and therapeutic response. Future work will focus on optimising acquisition and processing workflows to enhance robustness, reduce computational burden, and improve reproducibility. In particular, the development of automated parameter selection and advanced motion correction strategies will be critical for improving reliability and usability. Progress in these areas could accelerate the adoption of SURE in both preclinical research and clinical oncology, where contrast-free, high-resolution vascular imaging may provide valuable insights into tumour biology, vascular dynamics, and treatment response. Declarations Ethical Approval This study was approved by the Danish Animal Experiments Inspectorate ethics committee (licence no. 2021-15-0101-01041) and conducted in accordance with national legislation and the European Union Directive 2010/63/EU on the protection of animals used for scientific purposes. All efforts were made to minimise animal suffering and to reduce the number of animals used. Consent to Participate This study did not involve human participants. Therefore, ethical approval related to human subjects, informed consent, clinical trial registration, and Consent to Publish statements are not applicable. Consent to Publish This manuscript has been reviewed and approved by all co-authors. Conflicts of Interest Statement The authors declare the following employment relationships outside the submitted work: A.M. is employed by ICON PLC, I.T. by W.S. Audiology, and M.A.N. by GE Healthcare. These affiliations did not influence the design, conduct, analysis, or reporting of this study. The authors declare no other conflicts of interest. Competing Interests The authors declare the following employment relationships outside the submitted work: A.M. is employed by ICON PLC, I.T. by W.S. Audiology, and M.A.N. by GE Healthcare. These affiliations did not influence the design, conduct, analysis, or reporting of this study. The authors declare no other conflicts of interest. Funding Statement This study was funded by the ERC Synergy grant SURE, project no. 854796. Author Contribution A.M.: conceptualisation, methodology, investigation, formal analysis, writing — original draft, visualisation. I.T.: software, resources, data collection, data curation. N.P.J.: software, resources, data curation. A.S.C.: resources, Investigation — preparation of in vivo model. M.A.N.: software, resources, data curation. A.K.: supervision, writing — review. M.B.N.: supervision, funding acquisition, writing — review. J.A.J.: supervision, funding acquisition, software (lead), writing — review. C.M.S.: supervision (lead), project administration, funding acquisition, writing — review and editing. Data Availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. References Forster JC, et al. A review of the development of tumor vasculature and its effects on the tumor microenvironment. Hypoxia. 2017;5:21–32. Folkman J. Role of angiogenesis in tumor growth and metastasis. Seminars in Oncology, 2002. 29(6, Supplement 16): pp. 15–18. Harris AL, et al. Accessing the vasculature in cancer: revising an old hallmark. Trends Cancer. 2024;10(11):1038–51. Florea A, Mottaghy FM. Bauwens Molecular Imaging of Angiogenesis in Oncology: Current Preclinical and Clinical Status . Int J Mol Sci. 2021;22. 10.3390/ijms22115544 . Park AY, et al. The utility of ultrasound superb microvascular imaging for evaluation of breast tumour vascularity: comparison with colour and power Doppler imaging regarding diagnostic performance. Clin Radiol. 2018;73(3):304–11. Thompson G, et al. Imaging biomarkers of angiogenesis and the microvascular environment in cerebral tumours. Br J Radiol. 2014;84(specialissue2):S127–44. Anderson H, et al. Measuring changes in human tumour vasculature in response to therapy using functional imaging techniques. Br J Cancer. 2001;85(8):1085–93. Yang C-b, et al. Dual energy spectral CT imaging for the evaluation of small hepatocellular carcinoma microvascular invasion. Eur J Radiol. 2017;95:222–7. Eisenbrey JR, et al. Contrast-enhanced ultrasound (CEUS) in HCC diagnosis and assessment of tumor response to locoregional therapies. Abdom Radiol. 2021;46(8):3579–95. Porte C et al. Ultrasound Localization Microscopy for Cancer Imaging. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2024. 71(12: Breaking the Resolution Barrier in Ultrasound): pp. 1785–1800. Couture O, et al. Ultrasound Localization Microscopy and Super-Resolution: A State of the Art. IEEE Trans Ultrason Ferroelectr Freq Control. 2018;65(8):1304–20. Dencks S, Schmitz G. Ultrasound localization microscopy. Z Med Phys. 2023;33(3):292–308. Opacic T, et al. Motion model ultrasound localization microscopy for preclinical and clinical multiparametric tumor characterization. Nat Commun. 2018;9(1):1527. Hoyt K. Super-Resolution Ultrasound Imaging for Monitoring the Therapeutic Efficacy of a Vascular Disrupting Agent in an Animal Model of Breast Cancer. J Ultrasound Med. 2024;43(6):1099–107. Jensen JA, et al. Super-Resolution Ultrasound Imaging Using the Erythrocytes-Part I: Density Images. IEEE Trans Ultrason Ferroelectr Freq Control. 2024;71(8):925–44. Naji MA, et al. Super-Resolution Ultrasound Imaging Using the Erythrocytes-Part II: Velocity Images. IEEE Trans Ultrason Ferroelectr Freq Control. 2024;71(8):945–59. Amin Naji M, et al. Human lymph node microvascular imaging using a fast contrast-free super-resolution ultrasound technique. Sci Rep. 2025;15(1):23061. Jensen JA, et al. Universal Synthetic Aperture Sequence for Anatomic, Functional, and Super Resolution Imaging. IEEE Trans Ultrason Ferroelectr Freq Control. 2023;70(7):708–20. Walsh JC, et al. The clinical importance of assessing tumor hypoxia: relationship of tumor hypoxia to prognosis and therapeutic opportunities. Antioxid Redox Signal. 2014;21(10):1516–54. Additional Declarations Competing interest reported. The authors declare the following employment relationships outside the submitted work: A.M. is employed by ICON PLC, I.T. by W.S. Audiology, and M.A.N. by GE Healthcare. These affiliations did not influence the design, conduct, analysis, or reporting of this study. The authors declare no other conflicts of interest. Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 12 May, 2026 Reviewers agreed at journal 11 May, 2026 Reviews received at journal 21 Apr, 2026 Reviewers agreed at journal 16 Apr, 2026 Reviewers agreed at journal 15 Apr, 2026 Reviewers invited by journal 27 Mar, 2026 Editor assigned by journal 19 Mar, 2026 Submission checks completed at journal 17 Mar, 2026 First submitted to journal 16 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8989062","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":613447678,"identity":"4dbda2f6-77b9-4d01-90e1-83d1e99e78f6","order_by":0,"name":"Amy McDermott","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA70lEQVRIiWNgGAWjYPACZihdASISQDxmA2K0MDYcOEOyloNtRGjhn3bGgOlGjbWc+bTDxx9/nFeXuJ09gdmYh8HaGJcWids5Bsw5x9KNZW6nJTYc3HY4cWfPA+ZkHoZ0M5xOAmnJbTicOEM6xxCo5UDihhsJzId5GA7b4NIhD9VSD9Eyp46wFgOolgQJsJYGZrAWoMMO43SY4e20gsNAvxjOkE5LnHHm2GHjDWceNhvOMUjH6X2528kbH+fUWMtLSCcf+FBRUye74XjyYYk3FdaGDTj9z2FwAE2EEagYb0SyP8AnOwpGwSgYBaOAgQEATqxaoXqgMTkAAAAASUVORK5CYII=","orcid":"","institution":"University of Copenhagen","correspondingAuthor":true,"prefix":"","firstName":"Amy","middleName":"","lastName":"McDermott","suffix":""},{"id":613447679,"identity":"ab8b0bb9-34f6-42e0-9c5b-823d38a3305e","order_by":1,"name":"Iman Taghavi","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Iman","middleName":"","lastName":"Taghavi","suffix":""},{"id":613447680,"identity":"90f66d7b-0b2f-4f37-ac3e-ce9420a04150","order_by":2,"name":"Natalia Perez Jimenez","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Natalia","middleName":"Perez","lastName":"Jimenez","suffix":""},{"id":613447681,"identity":"abda22ce-12f6-4471-bb8c-a5f44f2e6266","order_by":3,"name":"Anne Skovsbo Clausen","email":"","orcid":"","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Anne","middleName":"Skovsbo","lastName":"Clausen","suffix":""},{"id":613447682,"identity":"9d397699-65b6-43b8-93db-ca3600cd7b2e","order_by":4,"name":"Mostafa Amin Naji","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Mostafa","middleName":"Amin","lastName":"Naji","suffix":""},{"id":613447683,"identity":"943e459e-f610-4980-9f57-467d057a8cd7","order_by":5,"name":"Andreas Kjaer","email":"","orcid":"","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Andreas","middleName":"","lastName":"Kjaer","suffix":""},{"id":613447684,"identity":"73b88bac-0c97-4655-9b93-bed081494932","order_by":6,"name":"Michael Bachmann Nielsen","email":"","orcid":"","institution":"Rigshospitalet","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"Bachmann","lastName":"Nielsen","suffix":""},{"id":613447685,"identity":"df3a6c5a-4062-4918-a896-916ab83766f1","order_by":7,"name":"Jørgen Arendt Jensen","email":"","orcid":"","institution":"Technical University of Denmark","correspondingAuthor":false,"prefix":"","firstName":"Jørgen","middleName":"Arendt","lastName":"Jensen","suffix":""},{"id":613447686,"identity":"252b9e29-bba9-46cf-b851-87c44843c421","order_by":8,"name":"Charlotte Mehlin Sørensen","email":"","orcid":"","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Charlotte","middleName":"Mehlin","lastName":"Sørensen","suffix":""}],"badges":[],"createdAt":"2026-02-27 14:23:44","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8989062/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8989062/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105878707,"identity":"573170d9-96ca-463e-8f3c-e2bdaff796e9","added_by":"auto","created_at":"2026-04-01 06:21:05","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":321641,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8989062/v1/a340455053b28afe5e13b14f.png"},{"id":105878710,"identity":"821f1e2c-88f4-4d86-ab40-037674ee408b","added_by":"auto","created_at":"2026-04-01 06:21:05","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":307168,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8989062/v1/ef3a289c76e93a2393b62ca0.png"},{"id":105878708,"identity":"fe812aa1-6ab8-4a47-888a-ac9830175d46","added_by":"auto","created_at":"2026-04-01 06:21:05","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":167067,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-8989062/v1/1d0b49bc4926fc6fddd082fe.png"},{"id":105878709,"identity":"0ad22348-3f18-4c22-95ae-87b605dfc7d4","added_by":"auto","created_at":"2026-04-01 06:21:05","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":207017,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-8989062/v1/5b60b877150ea2db9a5faf4f.png"},{"id":105905509,"identity":"ec7b269a-8309-448c-b66e-ef86b5eb4832","added_by":"auto","created_at":"2026-04-01 10:12:29","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1601558,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8989062/v1/5cb35495-57bc-41bc-b494-2ebea1a60a2d.pdf"}],"financialInterests":"Competing interest reported. The authors declare the following employment relationships outside the submitted work: A.M. is employed by ICON PLC, I.T. by W.S. Audiology, and M.A.N. by GE Healthcare. These affiliations did not influence the design, conduct, analysis, or reporting of this study. The authors declare no other conflicts of interest.","formattedTitle":"Contrast-free Super-Resolution Ultrasound Imaging for Microvascular Assessment of Murine Subcutaneous Tumour Models in Preclinical Oncology","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eAngiogenesis plays a critical role in tumour development, as the formation of new blood vessels provides the tissue perfusion required to sustain growth. Tumour vascular characteristics, such as tortuosity, density, and haemodynamics, are specific to tumour type and stage and can change in response to treatment. Neoplastic vasculature differs markedly from that of healthy tissue. Tumour-associated vessels are typically smaller in diameter, highly tortuous, and exhibit reduced and heterogeneous flow velocities [\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. These features often place tumour microvasculature below the spatial and temporal resolution limits of current in vivo imaging modalities, hindering comprehensive clinical detection and characterisation. Consequently, quantitative assessment of tumour vascularity remains challenging, leading to reliance on surrogate measures such as gross tumour size or molecular tracer uptake in both clinical assessment and preclinical murine oncology studies [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. This occurs despite well-established evidence that even partial vascular assessment using conventional imaging techniques, such as Doppler ultrasound, contrast-enhanced CT and MRI, provides significant clinical value for the diagnosis, treatment planning, and therapeutic monitoring of conditions involving solid tumours [\u003cspan additionalcitationids=\"CR6 CR7\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eEmerging ultrasound-based imaging techniques, including superb-microvascular imaging and other methods that utilise microbubble contrast agents to enhance vascular detection, have demonstrated increased diagnostic value compared with conventional clinical imaging modalities [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. In preclinical studies, the combination of microbubbles with high-frame-rate imaging systems has overcome the traditional diffraction limit of ultrasound, enabling in vivo visualisation at the capillary level in solid tumours [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThis approach, ultra-localised microscopy (ULM), enables microvascular imaging from approximately 10 minutes of acquired data [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. ULM can resolve tumour vessels as small as 10 um and provides comprehensive characterisation of flow direction, velocity, vessel morphology, and regional tumour perfusion [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. The enhanced haemodynamic and vascular information afforded by ULM offers significant value in clinical oncology, having already shown promise in the monitoring of murine oncology models [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. However, its clinical translatability remains limited by the need for prolonged imaging times and the use of exogenous contrast agents.\u003c/p\u003e \u003cp\u003eRecent advances in contrast-free ultrasound imaging have made significant progress in addressing the translatability limitations associated with ULM. Super-Resolution Ultrasound using Erythrocytes (SURE) has emerged as a non-invasive technique that exploits erythrocyte backscattering rather than microbubble localisation [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. This approach combines high-frame-rate imaging with specialised post-processing algorithms to amplify previously undetectable vascular signals, reducing the required data acquisition time to seconds while still enabling capillary-level imaging.\u003c/p\u003e \u003cp\u003eTo date, SURE has enabled novel in vivo microvasculature characterisation of the rat kidney and human lymph node [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. In this study, we present pilot testing of the SURE method for preclinical oncology and optimise the technique for future translation to clinical imaging of subcutaneous tumours. We investigate SURE performance in both low- and high-vascularity tumour models and assess the impact of tumour vascularity on imaging performance. We will also present a semi-automated method for quantitative vascular density analysis using SURE images and compare these measurements with those obtained from paired immunofluorescent immunohistochemistry.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy Population\u003c/h2\u003e \u003cp\u003eThree experiments were conducted using two murine tumour models. Pilot study 1 employed an LNCaP xenograft Balb/c athymic nude mouse model (n\u0026thinsp;=\u0026thinsp;3). Pilot study 2 and the main study utilised C57BL/6 mice inoculated with the B16-F10 melanoma cell line (ATCC\u0026reg; CRL-6475\u0026trade;) (n\u0026thinsp;=\u0026thinsp;2 and n\u0026thinsp;=\u0026thinsp;5, respectively). All animals were purchased from Janvier Labs (Le Genest-Saint-Isle, France) and housed in the University of Copenhagen core animal facility with ad libitum access to food and water. Tumours were established on the right flank and allowed to grow until approximately 1cm in diameter before inclusion in the study.\u003c/p\u003e \u003cp\u003eEach experiment had a distinct objective, and different imaging planes were acquired accordingly. The pilot studies focused on comparing imaging performance between tumour models and identifying optimal SURE imaging parameters. The experimental designs were as follows, with no overlap in population:\u003c/p\u003e \u003cp\u003e \u003cstrong\u003ePilot Study 1\u003c/strong\u003e \u003cp\u003eOne randomly selected imaging plane and one preidentified highly vascular imaging plane, identified using colour Doppler imaging, were acquired per LNCaP tumour (n\u0026thinsp;=\u0026thinsp;3).\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003ePilot Study 2\u003c/strong\u003e \u003cp\u003eTwo randomly selected ultrasound imaging planes were acquired per B16-F10 tumour (n\u0026thinsp;=\u0026thinsp;2).\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eMain Study\u003c/strong\u003e \u003cp\u003eQuantitative vascular analysis using SURE was compared with ex vivo immunohistochemistry. Three randomly selected imaging planes were selected blindly for both ultrasound imaging and histological analysis for each B16-F10 tumour (n\u0026thinsp;=\u0026thinsp;5).\u003c/p\u003e \u003c/p\u003e \u003cp\u003eAll imaging experiments lasted approximately one hour, after which animals were euthanised by cervical dislocation. In the main study, tumours were excised postmortem and fixed in formalin for 24 hours before paraffin embedding.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eUltrasound Data Collection\u003c/h3\u003e\n\u003cp\u003eMice were sedated using inhaled isoflurane and maintained at 2.0%. Hair within the target scanning window was removed, and animals were immobilised in lateral recumbency. A probe holder was used to position a 168-channel GE L8-18iD linear array transducer (GE HealthCare, USA) approximately 1 cm from the skin\u0026rsquo;s surface, with ultrasound gel providing an interface. The transducer, with a wavelength pitch of 150 \u0026micro;m, was operated at a transmission frequency of 10 MHz. The immobilised probe position was maintained while acquiring both SURE and conventional ultrasound images from the same imaging plane.\u003c/p\u003e \u003cp\u003eA commercial GE LOGIQ E9 scanner (GE HealthCare, USA) was used to acquire B-mode and colour Doppler images. The same transducer was then connected to a Verasonics Vantage 256 system (Verasonics, Inc., Kirkland, WA, USA) for SURE data acquisition. SURE data were acquired using a pulse inversion sequence comprising 24 emissions from 2\u0026times;12 virtual sources [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. The acquisition duration was 1 min, with a transmission voltage of 95 V. Data were acquired using both manual and automated time-gain compensation (TGC) settings to assess their impact on image quality. TGC adjustments, whether automated or manual, were performed to minimise signal clipping. All manual adjustments were completed by a single operator to reduce operator-induced bias.\u003c/p\u003e\n\u003ch3\u003eSURE Data Processing\u003c/h3\u003e\n\u003cp\u003eA windowed motion-correction method was employed, in which multiple low-motion data segments were manually selected from the 1-min acquisition. These segments were combined to produce approximately 10 s of low-motion data for subsequent processing and transverse oscillation motion correction. Each SURE scan data set was processed twice to illustrate the need for clinically tailored processing settings, similar to Doppler ultrasound methods. Singular value decomposition (SVD) processing is used for separating out tissue, flow, and noise. An upper threshold separates out tissue from flow, and a lower threshold is used for removing noise, where the lower singular values are disregarded. Specifically, the SVD range and dataset signal threshold for each scan were adjusted to optimise processing for two distinct clinical objectives: (i) high-sensitivity processing, maximising detection of small, low-flow vessels, and (ii) high clinical accuracy, minimising inclusion of erroneous background signal.\u003c/p\u003e \u003cp\u003eData processing was based on the method described by Jensen et al. [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e], with several modifications introduced to optimise the technique for oncology applications. To address clipping artefacts, clipped values in the raw radiofrequency data were identified and set to zero across all frames as a preprocessing step prior to SURE pipeline processing. Element exclusion was determined using beamformed clipped data rather than raw radiofrequency data alone. Elements contributing to clipping within the beamformed region of interest were excluded, while elements outside this region were retained to avoid unnecessary data loss. Excessive element exclusion was observed when clipping extended into the lower part of the image; in this region, the selection criteria were refined such that elements were removed only when their contributions degraded the beamformed region.\u003c/p\u003e\n\u003ch3\u003eSURE Image Analysis\u003c/h3\u003e\n\u003cp\u003eFor qualitative comparison with colour Doppler imaging, SURE images were overlaid onto corresponding B-mode images. SURE images were binarised, post-processed, and analysed using Fiji (ImageJ). Images were converted to 8-bit grayscale and smoothed using a Gaussian blur (σ\u0026thinsp;=\u0026thinsp;1.0 px). Local adaptive thresholding was applied to account for spatial variations in background intensity, using the Phansalkar method (local window radius of 45 px, \u003cem\u003ek\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.25, and \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.5). Post-processing was minimal and aimed to improve vessel continuity and suppress residual noise. Binary closing (one iteration) was applied to bridge small discontinuities within vessels, followed by removal of isolated bright pixels using the \u0026ldquo;Remove Outliers\u0026rdquo; filter (radius 2 px). Vessel contrast was enhanced by multiplying pixel intensities by a factor of 1.4 prior to final mask extraction. Vessel-specific masks were generated by converting the binary image into a selection and applying it to the original image, with all pixels outside the selection cleared to produce a transparent vessel mask suitable for B-mode overlay.\u003c/p\u003e \u003cp\u003eIn the main study, vascular density was measured using high-sensitivity SURE scans. Images were binarised using the method described above, and the tumour boundaries were manually defined using the free region-of-interest (ROI) tool in Fiji. Significant imaging artefacts were excluded manually. Vascular density was calculated as the percentage of ROI pixels containing vascular signal relative to the total number of ROI pixels.\u003c/p\u003e \u003cp\u003eQuantitative analysis of blood haemodynamics was also performed in the main study using the velocity-based SURE processing and analysis method outlined by Naji et al. [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. One scan per mouse was selected based on the imaging plane demonstrating the highest detectable vascularity on both SURE and colour Doppler imaging. Quantitative data were collected from multiple vessels within each SURE velocity map to analyse inter- and intra-subject haemodynamic variation across a single tumour imaging plane. Blood velocity and vessel lumen diameter were measured at manually selected vascular locations, with at least 9 sampling points selected per SURE image.\u003c/p\u003e \u003cp\u003eFurther haemodynamics characterisation was performed by estimating the volumetric blood flow rate for each vessel using the following equations:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$Bloodflow\\left(\\mu L/s\\right)=\\frac{Velocity\\left(mm/s\\right)\\times Area\\left(\\mu m\\right)\u0026sup2;}{1{0}^{9}}$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$VesselArea\\left(\\mu{m}^{2}\\right)=\\pi\\times Radius{\\left(\\mu m\\right)}^{2}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\n\u003ch3\u003eHistology\u003c/h3\u003e\n\u003cp\u003eSlide preparation and staining were performed by the Histology Laboratory at the Department of Biomedical Sciences, University of Copenhagen. Formalin-fixed paraffin-embedded (FFPE) samples, collected during the main study, were sectioned along the longitudinal axis, and three distinct regions were randomly selected for analysis. From each region, two slides were prepared using adjacent tissue sections. One slide was stained with hematoxylin and eosin (H\u0026amp;E) to guide the identification of tumour boundaries. The second slide underwent fluorescent immunohistochemistry targeting CD-31 to identify the vascular endothelium. This used anti-CD31 primary antibody (clone EPR17259; ab182981, Abcam) and an Alexa Fluor\u0026reg; 568-conjugated secondary antibody. Whole-section slide imaging was performed using the AxioScan Z1 (ZEISS) at x10 magnification.\u003c/p\u003e \u003cp\u003eQuantitative analysis of vascular density in digitised fluorescent immunohistochemistry images was performed using Fiji. No post-processing of images was performed. Tumour ROIs were manually delineated in Fiji, and any evident imaging artefacts were excluded. Images were segmented using a manually selected threshold to separate the fluorescent signal from the background prior to vascular density calculation. Vascular density was calculated as the percentage of ROI pixels containing fluorescent signal relative to the total number of ROI pixels.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eAll data analysed in this study were assessed for normality using the Shapiro-Wilk test. For comparisons between tumours, data that were not normally distributed were analysed using the Kruskal-Wallis test with Dunn\u0026rsquo;s post-hoc test, while normally distributed data were analysed using a one-way analysis of variance (ANOVA) with Tukey\u0026rsquo;s post-hoc test. The Wilcoxon signed-rank test was used to analyse paired data collected from the same tumour. Throughout the study, a p-value of \u0026lt;\u0026thinsp;0.05 was considered statistically significant. All statistical analyses were conducted using RStudio. Graphical representations indicate statistical significance using asterisks: *p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, **p\u0026thinsp;\u0026lt;\u0026thinsp;0.01, and ***p\u0026thinsp;\u0026lt;\u0026thinsp;0.001.\u003c/p\u003e \u003c/div\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eFeasibility Across Tumour Models\u003c/h2\u003e \u003cp\u003eIn the LNCaP xenograft model, flow was detected in one tumour using both colour Doppler and SURE imaging (Fig.\u0026nbsp;1a \u0026ndash; 1c). In this individual, the vascular signal was identifiable using both Doppler-guided and blind SURE acquisitions in matching tissue regions (Fig.\u0026nbsp;1a \u0026amp; 1c). In the remaining two tumours, no flow was detected using either modality.\u003c/p\u003e \u003cp\u003eIn the B16-F10 melanoma model, both colour Doppler and SURE consistently detected vascular signal across all imaging planes for both guided and blind acquisitions (Fig.\u0026nbsp;1d \u0026minus;\u0026thinsp;1f). These results show imaging feasibility to differ between tumour models and some consistency across ultrasound modalities.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eSURE and Colour Doppler Imaging Performance\u003c/h2\u003e \u003cp\u003eIn the B16-F10 melanoma model, SURE detected vascular signals in all acquired scans. Across all imaging planes, SURE demonstrated comparable or improved vascular imaging compared with colour Doppler, revealing structures not detected with Doppler alone. In regions where flow was detected using colour Doppler, SURE provided enhanced spatial detail, enabling clearer delineation of vessel boundaries and branches (Fig.\u0026nbsp;1d and 1f). The higher spatial detail also allowed discrimination between small vessels in close proximity, which were not resolvable in the corresponding colour Doppler images (Fig.\u0026nbsp;2; green box overlay).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eVariability in Image Quality\u003c/h2\u003e \u003cp\u003eData collected with both automated and manual TGC settings were affected by signal clipping, resulting in clipping artefacts. A small improvement in the data signal was observed in data collected with manual TGC selection, demonstrating that this parameter affects final image quality. Data collected using the manual TGC method were used for subsequent analysis and figure generation. As one operator completed all manual TGC, no analysis of operator variability can be performed.\u003c/p\u003e \u003cp\u003eThe optimal SVD and signal threshold processing settings needed to generate high-sensitivity or high-accuracy images varied between individuals and imaging planes. This could not be predicted and required screening of all combinations for each scan, similar to parameter optimisation in colour Doppler.\u003c/p\u003e \u003cp\u003eBoth high-sensitivity and high-accuracy SURE processing modes exceeded the vascular detection achieved with colour Doppler in most tissue regions (Fig.\u0026nbsp;2). High-sensitivity processing enabled detection of low-intensity vascular signals but resulted in increased background signal in deeper image regions, limiting its utility for B-mode overlay and interpretability. High-accuracy processing improved vessel contrast by reducing background signal, but reduced the detection of lower-intensity vascular signals from small vessels.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eMicrovascular Haemodynamic Characterisation\u003c/h2\u003e \u003cp\u003eSURE velocity processing enabled identification of flow direction across the imaging plane, and point-specific measurement of blood velocity and vessel diameter in vessels with sufficient signal strength (Fig.\u0026nbsp;3a). The majority of vessels included in high-accuracy scans met the signal threshold required for haemodynamic characterisation, with the number of analysable vessels per scan ranging from 9 to 18. Blood velocity, vessel diameter and flow direction were successfully measured for all selected vessels, enabling assessment of haemodynamic variability both within individual tumours and between animals (Fig.\u0026nbsp;3b-d). Comparable haemodynamic measurements could not be obtained using spectral Doppler, as no measurable flow was detected in any imaging planes.\u003c/p\u003e \u003cp\u003eAcross B16-F10 tumours in the main study, blood velocity measurements had a mean of 2.75 mm/s (median 2.4 mm/s, range 1.2\u0026ndash;7.5 mm/s) and differed significantly between tumours (Kruskal\u0026ndash;Wallis χ\u0026sup2;(4)\u0026thinsp;=\u0026thinsp;11.65, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.020*; Fig.\u0026nbsp;3b), with post-hoc testing identifying a significant difference in average velocity between tumours B and E. Mean vessel lumen diameter across all measurements was 121.31 \u0026micro;m (range 52.8\u0026ndash;251.1 \u0026micro;m) and also differed significantly between tumours (one-way ANOVA F(4,72)\u0026thinsp;=\u0026thinsp;4.118, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.00463**; Fig.\u0026nbsp;3c), with tumour E exhibiting larger diameters than tumour B. Estimated blood flow rates showed significant differences across tumours (Kruskal\u0026ndash;Wallis χ\u0026sup2;(4)\u0026thinsp;=\u0026thinsp;20.88, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.00034***; Fig.\u0026nbsp;3d), with tumour E demonstrating higher flow rates than tumours A and B.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eVascular Density Quantification and Histological Comparison\u003c/h2\u003e \u003cp\u003eQuantitative vascular density measurements derived from in vivo SURE imaging did not differ significantly from those obtained using ex vivo immunohistochemistry (Fig.\u0026nbsp;4a). Qualitatively, fluorescent immunohistochemistry revealed a greater number of small vessels and finer vascular detail compared with SURE images (Fig.\u0026nbsp;4b-c). SURE images were affected by signal banding artefacts, which persisted under both manual and automated time-gain compensation settings. The banding artefact can be seen in the lower right quadrant of Fig.\u0026nbsp;4b radiating into the image. This can be caused by any strong signal reflector (e.g., air) outside of the imaging plane.\u003c/p\u003e \u003cp\u003eAcross the main study dataset (15 measurements from five tumours), SURE-derived vascular density measurements had a mean value of 4.42% (SD 2.33%, range 1.30\u0026ndash;8.61%). Vascular density varied between tumours but did not differ significantly across individuals (Kruskal\u0026ndash;Wallis χ\u0026sup2;(4)\u0026thinsp;=\u0026thinsp;9.13, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.058). Immunohistochemistry-derived vascular density measurements showed a mean value of 3.00% (SD 0.66%, range 2.01\u0026ndash;4.07%), with comparatively low variability across tumours. Paired comparison using the Wilcoxon signed-rank test demonstrated no statistically significant differences between SURE- and immunohistochemistry-derived vascular density measurements within any tumour.\u003c/p\u003e \u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThis study demonstrates the feasibility and capabilities of contrast-free SURE as an in vivo tool for analysing and assessing the vascularity of subcutaneous murine tumours. We demonstrate superior imaging resolution and microvascular detection capabilities compared with comparable methods, e.g. colour Doppler. It is confirmed that SURE can be utilised at a dataset level to characterise tumour haemodynamics (flow direction, velocity, and volumetric flow) and vascular density. We confirmed the value of this tool in structural vascular density analysis by comparing it with immunohistochemistry results obtained using a compatible method; no significant difference was observed in the final datasets.\u003c/p\u003e \u003cp\u003eTumour-model-dependent feasibility was observed, with both SURE and colour Doppler failing to detect vasculature in two of three LNCaP tumours, whereas all B16-F10 tumours showed detectable vasculature in every individual using both imaging methods. We hypothesise that this reflects known differences in vascular architecture and haemodynamics between the two models. LNCaP tumours are characterised by lower vascular density and slower, often heterogeneous blood flow [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. The SURE method relies on the ability to distinguish moving erythrocyte signals from those of surrounding stationary tissues. As such, it is likely that lower flow rates impair the ability to identify LNCaP vessels, despite their presence within the spatial detection limits. Current SURE detection limits remain unknown, and rapid advances in erythrocyte tracking and motion correction methods will likely lower the limit for detectable vessels.\u003c/p\u003e \u003cp\u003eThe higher spatial resolution of SURE enabled visualisation and qualitative assessment of tumour microvasculature beyond the capabilities of conventional colour Doppler imaging. SURE provided enhanced delineation of vessel morphology, allowing clear identification of in-plane vascular branching, tortuosity, and avascular regions within tumours, features typically accessible only with contrast-enhanced ultrasound or ex vivo methods. In regions where flow was detected by both modalities, SURE frequently provided greater spatial detail, improving the visualisation of vessel boundaries and spatial relationships.\u003c/p\u003e \u003cp\u003eDespite these advantages, colour Doppler imaging retains practical strengths. Doppler provides real-time visualisation of blood flow, whereas SURE currently requires extended data acquisition (approximately one minute) and subsequent post-processing prior to image generation. Consequently, SURE is currently less suited to rapid, real-time assessment. However, SURE offers additional functional information, including flow direction mapping comparable to that provided by Doppler imaging, while remaining contrast-free. SURE images were more frequently affected by artefacts, most prominently, signal banding, which persisted under both manual and automated TGC adjustment. The prevalence of SURE artefacts is impacted by the selection og processing parameters and is most prevalent in high-sensitivity vascular imaging. Together, these findings indicate that SURE complements rather than replaces colour Doppler, providing improved microvascular detail at the expense of increased acquisition and processing time.\u003c/p\u003e \u003cp\u003eThe clinical tailoring of SURE processing parameters and the SURE-B-mode overlay method are novel applications of the method and represent efforts to optimise the SURE method for clinical use. While tailoring of SURE processing parameters currently represents a considerable computational endeavour, rapid technical advancements are increasing the feasibility of this type of processing. The ability to tailor processing parameters to clinical aims increases the method\u0026rsquo;s usability and the potential for large-dataset preclinical use. Further studies across varied tumour models are needed to confirm whether manually tailoring (as used in this study) is always required to identify the optimal processing parameters in a data set.\u003c/p\u003e \u003cp\u003eMicrovascular haemodynamic characterisation has currently not been extensively utilised in preclinical oncology. To date, pharmacology and core research are mostly limited to structural assessment of the gross tumour, monitoring molecular tracer uptake, or imaging of large vessels. Consequently, despite knowing that perfusion and angiogenesis are key components of tumour growth, they are rarely included in oncology research due to limitations in imaging. This limitation is exemplified in this work by the failure of spectral Doppler to quantify flow in any tumour. SURE has the potential to address this capability gap.\u003c/p\u003e \u003cp\u003eQuantitative vascular density measurements derived from SURE imaging showed agreement with those obtained using immunohistochemistry, supporting the ability of SURE to capture overall vascular burden in vivo. While no statistically significant differences were observed between the two methods, qualitative comparison demonstrated that histological imaging revealed a greater number of small vessels and finer microvascular detail. This difference is expected given the higher spatial resolution of ex vivo histological techniques and the absence of physiological motion and acoustic attenuation. The observed differences highlight the complementary nature of in vivo SURE imaging and histological analysis rather than a discrepancy between methods. Histology provides high-resolution structural detail but is limited to end-point assessment, whereas SURE enables non-invasive, in vivo visualisation of tumour vasculature under physiological conditions. As such, histology should not be considered a direct ground truth for in vivo imaging, but rather a complementary reference that contextualises the strengths and limitations of SURE.\u003c/p\u003e \u003cp\u003eThis study has several limitations. The sample size was small, reflecting the pilot nature of the work, and limits the generalisability of quantitative findings. The low sample size also resulted in an underpowered statistical analysis; consequently, all statistical tests described must be considered exploratory. SURE imaging is sensitive to physiological motion, requiring careful selection of low-motion data segments for processing. The impact of low-motion window selection on image quality was not assessed here. These limitations should be considered when interpreting the results and highlighting areas for further technical refinement.\u003c/p\u003e"},{"header":"CONCLUSIONS","content":"\u003cp\u003eSURE represents a significant advance in contrast-free in vivo microvascular imaging. Despite current limitations, including processing time, susceptibility to image artefacts, and restricted quantitative analysis, the findings of this study demonstrate strong potential for longitudinal assessment of tumour microvasculature in preclinical oncology models. The ability to visualise microvascular architecture and haemodynamic behaviour in vivo, without the need for exogenous contrast agents, supports repeated imaging for monitoring tumour progression and therapeutic response.\u003c/p\u003e \u003cp\u003eFuture work will focus on optimising acquisition and processing workflows to enhance robustness, reduce computational burden, and improve reproducibility. In particular, the development of automated parameter selection and advanced motion correction strategies will be critical for improving reliability and usability. Progress in these areas could accelerate the adoption of SURE in both preclinical research and clinical oncology, where contrast-free, high-resolution vascular imaging may provide valuable insights into tumour biology, vascular dynamics, and treatment response.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003e\u003cstrong\u003eEthical Approval\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThis study was approved by the Danish Animal Experiments Inspectorate ethics committee (licence no. 2021-15-0101-01041) and conducted in accordance with national legislation and the European Union Directive 2010/63/EU on the protection of animals used for scientific purposes. All efforts were made to minimise animal suffering and to reduce the number of animals used.\u003c/p\u003e\n\u003ch2\u003eConsent to Participate\u003c/h2\u003e\n\u003cp\u003eThis study did not involve human participants. Therefore, ethical approval related to human subjects, informed consent, clinical trial registration, and Consent to Publish statements are not applicable.\u003c/p\u003e\n\u003ch2\u003eConsent to Publish\u003c/h2\u003e\n\u003cp\u003eThis manuscript has been reviewed and approved by all co-authors.\u003c/p\u003e\n\u003ch2\u003eConflicts of Interest \u003cstrong\u003eStatement\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe authors declare the following employment relationships outside the submitted work: A.M. is employed by ICON PLC, I.T. by W.S. Audiology, and M.A.N. by GE Healthcare. These affiliations did not influence the design, conduct, analysis, or reporting of this study. The authors declare no other conflicts of interest.\u003c/p\u003e\n\u003ch2\u003eCompeting Interests\u003c/h2\u003e\n\u003cp\u003eThe authors declare the following employment relationships outside the submitted work: A.M. is employed by ICON PLC, I.T. by W.S. Audiology, and M.A.N. by GE Healthcare. These affiliations did not influence the design, conduct, analysis, or reporting of this study. The authors declare no other conflicts of interest.\u003c/p\u003e\n\u003ch2\u003eFunding Statement\u003c/h2\u003e\n\u003cp\u003eThis study was funded by the ERC Synergy grant SURE, project no. 854796.\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eA.M.: conceptualisation, methodology, investigation, formal analysis, writing \u0026mdash; original draft, visualisation. I.T.: software, resources, data collection, data curation. N.P.J.: software, resources, data curation. A.S.C.: resources, Investigation \u0026mdash; preparation of in vivo model. M.A.N.: software, resources, data curation. A.K.: supervision, writing \u0026mdash; review. M.B.N.: supervision, funding acquisition, writing \u0026mdash; review. J.A.J.: supervision, funding acquisition, software (lead), writing \u0026mdash; review. C.M.S.: supervision (lead), project administration, funding acquisition, writing \u0026mdash; review and editing.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eThe datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eForster JC, et al. A review of the development of tumor vasculature and its effects on the tumor microenvironment. Hypoxia. 2017;5:21\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFolkman J. \u003cem\u003eRole of angiogenesis in tumor growth and metastasis.\u003c/em\u003e Seminars in Oncology, 2002. 29(6, Supplement 16): pp. 15\u0026ndash;18.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarris AL, et al. Accessing the vasculature in cancer: revising an old hallmark. Trends Cancer. 2024;10(11):1038\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFlorea A, Mottaghy FM. Bauwens \u003cem\u003eMolecular Imaging of Angiogenesis in Oncology: Current Preclinical and Clinical Status\u003c/em\u003e. Int J Mol Sci. 2021;22. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ijms22115544\u003c/span\u003e\u003cspan address=\"10.3390/ijms22115544\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePark AY, et al. The utility of ultrasound superb microvascular imaging for evaluation of breast tumour vascularity: comparison with colour and power Doppler imaging regarding diagnostic performance. Clin Radiol. 2018;73(3):304\u0026ndash;11.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThompson G, et al. Imaging biomarkers of angiogenesis and the microvascular environment in cerebral tumours. Br J Radiol. 2014;84(specialissue2):S127\u0026ndash;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnderson H, et al. Measuring changes in human tumour vasculature in response to therapy using functional imaging techniques. Br J Cancer. 2001;85(8):1085\u0026ndash;93.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang C-b, et al. Dual energy spectral CT imaging for the evaluation of small hepatocellular carcinoma microvascular invasion. Eur J Radiol. 2017;95:222\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEisenbrey JR, et al. Contrast-enhanced ultrasound (CEUS) in HCC diagnosis and assessment of tumor response to locoregional therapies. Abdom Radiol. 2021;46(8):3579\u0026ndash;95.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePorte C et al. \u003cem\u003eUltrasound Localization Microscopy for Cancer Imaging.\u003c/em\u003e IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2024. 71(12: Breaking the Resolution Barrier in Ultrasound): pp. 1785\u0026ndash;1800.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCouture O, et al. Ultrasound Localization Microscopy and Super-Resolution: A State of the Art. IEEE Trans Ultrason Ferroelectr Freq Control. 2018;65(8):1304\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDencks S, Schmitz G. Ultrasound localization microscopy. Z Med Phys. 2023;33(3):292\u0026ndash;308.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOpacic T, et al. Motion model ultrasound localization microscopy for preclinical and clinical multiparametric tumor characterization. Nat Commun. 2018;9(1):1527.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoyt K. Super-Resolution Ultrasound Imaging for Monitoring the Therapeutic Efficacy of a Vascular Disrupting Agent in an Animal Model of Breast Cancer. J Ultrasound Med. 2024;43(6):1099\u0026ndash;107.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJensen JA, et al. Super-Resolution Ultrasound Imaging Using the Erythrocytes-Part I: Density Images. IEEE Trans Ultrason Ferroelectr Freq Control. 2024;71(8):925\u0026ndash;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNaji MA, et al. Super-Resolution Ultrasound Imaging Using the Erythrocytes-Part II: Velocity Images. IEEE Trans Ultrason Ferroelectr Freq Control. 2024;71(8):945\u0026ndash;59.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmin Naji M, et al. Human lymph node microvascular imaging using a fast contrast-free super-resolution ultrasound technique. Sci Rep. 2025;15(1):23061.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJensen JA, et al. Universal Synthetic Aperture Sequence for Anatomic, Functional, and Super Resolution Imaging. IEEE Trans Ultrason Ferroelectr Freq Control. 2023;70(7):708\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalsh JC, et al. The clinical importance of assessing tumor hypoxia: relationship of tumor hypoxia to prognosis and therapeutic opportunities. Antioxid Redox Signal. 2014;21(10):1516\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"discover-imaging","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Discover Imaging](https://link.springer.com/journal/44352)","snPcode":"44352","submissionUrl":"https://submission.springernature.com/new-submission/44352/3","title":"Discover Imaging","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Super-resolution ultrasound, Contrast-free ultrasound, Microvascular imaging, Preclinical oncology, Erythrocyte tracking, Murine tumour model","lastPublishedDoi":"10.21203/rs.3.rs-8989062/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8989062/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e \u003cb\u003eIntroduction\u003c/b\u003e. Tumour vascularity is an important biomarker of tumour growth and therapeutic response. Current in vivo imaging methods have a limited ability to detect individual tumour vessels due to the small vessel size and low flow rates commonly observed in murine oncology models. This paper presents three pilot studies which together aim to assess the performance of a developing imaging method, Super-resolution Ultrasound using Erythrocytes (SURE), for subcutaneous tumour imaging. \u003cb\u003eMethods\u003c/b\u003e. Across three experiments, we qualitatively compared the vascular imaging sensitivity and clinical applicability of SURE imaging with colour Doppler imaging and ex vivo immunohistochemistry. Two murine oncology models were investigated: the LNCaP (lymph node carcinoma of the prostate) xenograft model (n\u0026thinsp;=\u0026thinsp;3) and the B16-F10 melanoma model (n\u0026thinsp;=\u0026thinsp;7). Optimal SURE post-processing parameters for subcutaneous tumour imaging were characterised. \u003cb\u003eResults\u003c/b\u003e. SURE outperformed Doppler imaging, detecting flow with higher spatial resolution in corresponding tissue regions. SURE successfully visualised tumour microvasculature, detecting vessels approximately 30 \u0026micro;m in diameter while enabling haemodynamic characterisation. Optimal post-processing involved a trade-off between vascular imaging sensitivity, final signal-to-noise ratio, and image artefact prevalence. Vascular density measurements obtained using SURE did not differ significantly from those obtained by immunohistochemistry. \u003cb\u003eConclusions\u003c/b\u003e. SURE represents a significant advance in in vivo microvascular imaging. While post-processing time, image artefacts, and limited quantitative analysis currently limit preclinical application, the technique demonstrates strong potential for tumour vascular characterisation in both preclinical research and clinical oncology.\u003c/p\u003e","manuscriptTitle":"Contrast-free Super-Resolution Ultrasound Imaging for Microvascular Assessment of Murine Subcutaneous Tumour Models in Preclinical Oncology","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-01 06:21:01","doi":"10.21203/rs.3.rs-8989062/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-05-12T21:39:11+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"45594792588627704136883189949988616132","date":"2026-05-11T05:21:49+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-21T09:40:02+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"313001874725822709173766934271139562467","date":"2026-04-16T08:39:18+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"161058763327278692356591723077036245917","date":"2026-04-15T15:22:14+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-27T14:33:09+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-19T10:34:11+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-17T04:51:22+00:00","index":"","fulltext":""},{"type":"submitted","content":"Discover Imaging","date":"2026-03-16T22:08:06+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"discover-imaging","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Discover Imaging](https://link.springer.com/journal/44352)","snPcode":"44352","submissionUrl":"https://submission.springernature.com/new-submission/44352/3","title":"Discover Imaging","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"9439d2d1-09a5-48fc-9ab3-cfb7d9e08ee1","owner":[],"postedDate":"April 1st, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-12T21:39:11+00:00","index":67,"fulltext":""},{"type":"reviewerAgreed","content":"45594792588627704136883189949988616132","date":"2026-05-11T05:21:49+00:00","index":66,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-01T06:21:01+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-01 06:21:01","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8989062","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8989062","identity":"rs-8989062","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}