Localization of High-concentration Microbubbles for Ultrasound Localization Microscopy by Self-Supervised Deep Learning

Li, Yongshuai; Huang, Lijie; Zhang, Jingke; Huang, Chengwu; Chen, Shigao; Luo, Jianwen

doi:10.1109/ius52206.2021.9593750

Cited by 10 publications

(4 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Unlike our presented work, it does not include biological factors such as tumor-induced angiogenesis or the effects of anti-angiogenic substances. Other recent and more technical publications include the use of the CAM model for the application of ultrasound localization microscopy, which utilizes a CNN to visualize microvessels [ 59 , 60 ]. Li et al talk about the difficulty of obtaining the large in vivo training datasets required to train previously used cross-correlation-based localization methods combatted by the improved learning scheme of CNNs, though only using ex ovo CAM assay training data.…”

Section: Discussionmentioning

confidence: 99%

Deep Learning-Based Image Analysis for the Quantification of Tumor-Induced Angiogenesis in the 3D In Vivo Tumor Model—Establishment and Addition to Laser Speckle Contrast Imaging (LSCI)

Kuri

Pion

Mahl

et al. 2022

Cells

View full text Add to dashboard Cite

(1) Background: angiogenesis plays an important role in the growth and metastasis of tumors. We established the CAM assay application, an image analysis software of the IKOSA platform by KML Vision, for the quantification of blood vessels with the in ovo chorioallantoic membrane (CAM) model. We added this proprietary deep learning algorithm to the already established laser speckle contrast imaging (LSCI). (2) Methods: angiosarcoma cell line tumors were grafted onto the CAM. Angiogenesis was measured at the beginning and at the end of tumor growth with both measurement methods. The CAM assay application was trained to enable the recognition of in ovo CAM vessels. Histological stains of the tissue were performed and gluconate, an anti-angiogenic substance, was applied to the tumors. (3) Results: the angiosarcoma cells formed tumors on the CAM that appeared to stay vital and proliferated. An increase in perfusion was observed using both methods. The CAM assay application was successfully established in the in ovo CAM model and anti-angiogenic effects of gluconate were observed. (4) Conclusions: the CAM assay application appears to be a useful method for the quantification of angiogenesis in the CAM model and gluconate could be a potential treatment of angiosarcomas. Both aspects should be evaluated in further research.

show abstract

Section: Discussionmentioning

confidence: 99%

Deep Learning-Based Image Analysis for the Quantification of Tumor-Induced Angiogenesis in the 3D In Vivo Tumor Model—Establishment and Addition to Laser Speckle Contrast Imaging (LSCI)

Kuri

Pion

Mahl

et al. 2022

Cells

View full text Add to dashboard Cite

show abstract

“…The radial symmetry localization method used in this article can localize overlapped MBs to some extent 37 . More accurate localization methods, such as deep‐learning‐based methods, can also be used for separation of high‐concentration MBs 53 . Note that, high concentration of MBs brings out sufficient MBs events in the superb subtle vessels, such as the vessels located at 0–0.5 mm away from the renal capsule.…”

Section: Discussionmentioning

confidence: 99%

“…37 More accurate localization methods, such as deep-learning-based methods, can also be used for separation of high-concentration MBs. 53 Note that, high concentration of MBs brings out sufficient MBs events in the superb subtle vessels, such as the vessels located at 0-0.5 mm away from the renal capsule. In other words, a higher filling rate of these vessels can be guaranteed, which has the potential to shorten data acquisition time.…”

Section: Discussionmentioning

confidence: 99%

Evaluation of Early Diabetic Kidney Disease Using Ultrasound Localization Microscopy: A Feasibility Study

Zhang

Huang

Yang

et al. 2023

J of Ultrasound Medicine

Self Cite

View full text Add to dashboard Cite

ObjectiveThe purpose of this study is to detect the hemodynamic changes of microvessels in the early stage of diabetic kidney disease (DKD) and to test the feasibility of ultrasound localization microscopy (ULM) in early diagnosis of DKD.MethodsIn this study, streptozotocin (STZ) induced DKD rat model was used. Normal rats served as the control group. Conventional ultrasound, contrast‐enhanced ultrasound (CEUS), and ULM data were collected and analyzed. The kidney cortex was divided into four segments, which are 0.25–0.5 mm (Segment 1), 0.5–0.75 mm (Segment 2), 0.75–1 mm (Segment 3), and 1–1.25 mm (Segment 4) away from the renal capsule, respectively. The mean blood flow velocities of arteries and veins in each segment were separately calculated, and also the velocity gradients and overall mean velocities of arteries and veins. Mann–Whitney U test was used for comparison of the data.ResultsQuantitative results of microvessel velocity obtained by ULM show that the arterial velocity of Segments 2, 3, and 4, and the overall mean arterial velocity of the four segments in the DKD group are significantly lower than those in the normal group. The venous velocity of Segment 3 and the overall mean venous velocity of the four segments in the DKD group are higher than those in the normal group. The arterial velocity gradient in the DKD group is lower than that in the normal group.ConclusionULM can visualize and quantify the blood flow and may be used for early diagnosis of DKD.

show abstract

“…Such a scenario in the phantoms that we were imaging will yield a good super-resolution image, but the tracking algorithm will not work properly, which implies that a too high concentration is used. Certain algorithms such as sparsity or deep-learning-based algorithms may be better adept for accurately localizing high concentrations of MBs and could be used with the microfluidic-inspired phantoms developed here. , Therefore, the concentration was optimized to ensure accuracy in all of the other experiments. Next, velocity maps for three different flow rates in phantoms with different bifurcation angles showed the expected trend of a faster velocity for a larger flow rate.…”

Section: Discussionmentioning

confidence: 99%

Versatile Ultrasound-Compatible Microfluidic Platform for In Vitro Microvasculature Flow Research and Imaging Optimization

Mano,

Grutman,

Ilovitsh

2023

ACS Omega

View full text Add to dashboard Cite

Ultrasound localization microscopy (ULM) enables the creation of super-resolved images and velocity maps by localizing and tracking microbubble contrast agents through a vascular network over thousands of frames of ultrafast plane wave images. However, a significant challenge lies in developing ultrasound-compatible microvasculature phantoms to investigate microbubble flow and distribution in controlled environments. In this study, we introduce a new class of gelatin-based microfluidic-inspired phantoms uniquely tailored for ULM studies. These devices allow for the creation of complex and reproducible microvascular networks featuring channel diameters as small as 100 μm. Our experiments focused on microbubble behavior under ULM conditions within bifurcating and converging vessel phantoms. We evaluated the impact of bifurcation angles (25, 45, and 55°) and flow rates (0.01, 0.02, and 0.03 mL/min) on the acquisition time of branching channels. Additionally, we explored the saturation time effect of narrow channels branching off larger ones. Significantly longer acquisition times were observed for the narrow vessels, with an average increase of 72% when a 100 μm channel branched off from a 300 μm channel and an average increase of 90% for a 200 μm channel branching off from a 500 μm channel. The robustness of our fabrication method is demonstrated through the creation of two trifurcating microfluidic phantoms, including one that converges back into a single channel, a configuration that cannot be achieved through traditional methods. This new class of ULM phantoms serves as a versatile platform for noninvasively studying complex flow patterns using ultrasound imaging, unlocking new possibilities for in vitro microvasculature research and imaging optimization.

show abstract

Localization of High-concentration Microbubbles for Ultrasound Localization Microscopy by Self-Supervised Deep Learning

Cited by 10 publications

References 20 publications

Deep Learning-Based Image Analysis for the Quantification of Tumor-Induced Angiogenesis in the 3D In Vivo Tumor Model—Establishment and Addition to Laser Speckle Contrast Imaging (LSCI)

Deep Learning-Based Image Analysis for the Quantification of Tumor-Induced Angiogenesis in the 3D In Vivo Tumor Model—Establishment and Addition to Laser Speckle Contrast Imaging (LSCI)

Evaluation of Early Diabetic Kidney Disease Using Ultrasound Localization Microscopy: A Feasibility Study

Versatile Ultrasound-Compatible Microfluidic Platform for In Vitro Microvasculature Flow Research and Imaging Optimization

Contact Info

Product

Resources

About