Acoustic response and ambient pressure sensitivity characterization of SonoVue for noninvasive pressure estimation

Subharmonic aided pressure estimation (SHAPE) is a noninvasive pressure measurement technique based on the pressure dependent subharmonic signal from contrast microbubbles. Here, SonoVue microbubble with a sulfur hexafluoride (SF6) core, was investigated for use in SHAPE. The study uses excitations of 25-700 kPa peak negative pressure (PNP) and 3 MHz frequency over eight pressurization cycles between atmospheric pressure and overpressures, ranging from 0 to 25 kPa (0 to 186 mm Hg). The SonoVue subharmonic response was characterized into two types. Unlike other microbubbles, SonoVue showed significant subharmonic signals at low excitations (PNPs, 25-400 kPa), denoted here as type I subharmonic. It linearly decreased with increasing overpressure (-0.52 dB/kPa at 100 kPa PNP). However, over multiple pressurization-depressurization cycles, type I subharmonic changed; its value at atmospheric pressure decreased over multiple cycles, and at later cycles, it recorded an increase in amplitude with overpressure (highest, +13 dB at 50 kPa PNP and 10 kPa overpressure). The subharmonic at higher excitations (PNP > 400 kPa), denoted here as type II subharmonic, showed a consistent decrease with the ambient pressure increase with strongest sensitivity of -0.4 dB/kPa at 500 kPa PNP.

Determining Ultrasound Parameters for Bursting Polymer Microbubbles for Future Use in Spinal Cord Injury

OBJECTIVE

We believe our poly(lactic acid) (PLA) microbubbles are well suited for therapeutic delivery to spinal cord injury (SCI) using ultrasound-triggered bursting. We investigated the feasibility of clinical ultrasound bursting in situ, the optimal bursting parameters in vitro and the loading and release of a model bio-active DNA.

METHODS

Microbubbles were tested using clinical ultrasound in a rat cadaver SCI model. Burst pressure thresholds were determined using the change in enhancement after ultrasound exposure. Resonance frequency, acoustic enhancement, sizing and morphology were evaluated by comparing two microbubble porogens, ammonium carbonate and ammonium carbamate. Oligonucleotides were loaded into the shell and released using the found optimized ultrasound bursting parameters.

RESULTS

In situ imaging and bursting were successful. In vitro bursting thresholds using frequencies 1, 2.25 and 5 MHz were identified between peak negative pressures 0.2 and 0.5 MPa, believed to be safe for spinal cord. The pressure threshold decreased with decreasing frequencies. PLA bursting was optimized near the resonance frequency of 2.5 to 3.0 MHz using 2.25 MHz and not at lower frequencies. PLA microbubbles, initially with a mean size of approximately 2 µm, remained in one piece, collapsed to between 0.5 and 1 µm and did not fragment. Significantly more oligonucleotide was released after ultrasound bursting of loaded microbubbles. Microbubble-sized debris was detected when using ammonium carbamate, leading to inaccurate microbubble concentration measurements.

CONCLUSION

PLA microbubbles made with ammonium carbonate and burst at appropriate parameters have the potential to safely improve intrathecal therapeutic delivery to SCI using targeted ultrasound.

Percutaneous Microwave Ablation versus Cryoablation for Small Renal Masses (≤4 cm): 12-Year Experience at a Single Center

PURPOSE

To determine whether microwave ablation (MWA) has equivalent outcomes to those of cryoablation (CA) in terms of technical success, adverse events, local tumor recurrence, and survival in adult patients with solid enhancing renal masses ≤4 cm.

MATERIALS AND METHODS

A retrospective review was performed of 279 small renal masses (≤4 cm) in 257 patients (median age, 71 years; range, 40-92 years) treated with either CA (n = 191) or MWA (n = 88) between January 2008 and December 2020 at a single high-volume institution. Evaluations of adverse events, treatment effectiveness, and therapeutic outcomes were conducted for both MWA and CA. Disease-free, metastatic-free, and cancer-specific survival rates were tabulated. The estimated glomerular filtration rate was employed to examine treatment-related alterations in renal function.

RESULTS

No difference in patient age (P = .99) or sex (P = .06) was observed between the MWA and CA groups. Cryoablated lesions were larger (P < .01) and of greater complexity (P = .03). The technical success rate for MWA was 100%, whereas 1 of 191 cryoablated lesions required retreatment for residual tumor. There was no impact on renal function after CA (P = .76) or MWA (P = .49). Secondary analysis using propensity score matching demonstrated no significant differences in local recurrence rates (P = .39), adverse event rates (P = .20), cancer-free survival (P = .76), or overall survival (P = .19) when comparing matched cohorts of patients who underwent MWA and CA.

CONCLUSIONS

High technical success and local disease control were achieved for both MWA and CA. Cancer-specific survival was equivalent. Higher adverse event rates after CA may reflect the tendency to treat larger, more complex lesions with CA.

Contrast-enhanced ultrasound liver imaging reporting and data system: clinical validation in a prospective multinational study in North America and Europe

BACKGROUND AND AIMS

The objective of this study is to determine the diagnostic accuracy of the American College of Radiology Contrast-Enhanced Ultrasound (CEUS) Liver Imaging Reporting and Data System LR-5 characterization for HCC diagnosis in North American or European patients.

APPROACH AND RESULTS

A prospective multinational cohort study was performed from January 2018 through November 2022 at 11 academic and nonacademic centers in North America and Europe. Patients at risk for HCC with at least 1 liver observation not previously treated, identified on ultrasound (US), or multiphase CT or MRI performed as a part of standard clinical care were eligible for the study. All participants were examined with CEUS of the liver within 4 weeks of CT/MRI or tissue diagnosis to characterize up to 2 liver nodules per participant using ACR CEUS Liver Imaging Reporting and Data System. Definite HCC diagnosis on the initial CT/MRI, imaging follow-up, or histology for CT/MRI-indeterminate nodules were used as reference standards. A total of 545 nodules had confirmed reference standards in 480 patients, 73.8% were HCC, 5.5% were other malignancies, and 20.7% were nonmalignant. The specificity of CEUS LR-5 for HCC was 95.1% (95% CI 90.1%-97.7%), sensitivity 62.9% (95% CI 57.9%-67.7%), positive predictive value 97.3% (95% CI 94.5%-98.7%), and negative predictive value 47.7% (95% CI 41.7%-53.8%). In addition, benign CEUS characterization (LR-1 or LR-2) had 100% specificity and 100% positive predictive value for nonmalignant liver nodules.

CONCLUSIONS

CEUS Liver Imaging Reporting and Data System provides an accurate categorization of liver nodules in participants at risk for HCC.

Contrast-enhanced US Evaluation of Hepatocellular Carcinoma Response to Chemoembolization: A Prospective Multicenter Trial

Background Contrast-enhanced (CE) US has been studied for use in the detection of residual viable hepatocellular carcinoma (HCC) after locoregional therapy, but multicenter data are lacking. Purpose To compare two-dimensional (2D) and three-dimensional (3D) CE US diagnostic performance with that of CE MRI or CT, the current clinical standard, in the detection of residual viable HCC after transarterial chemoembolization (TACE) in a prospective multicenter trial. Materials and Methods Participants aged at least 21 years with US-visible HCC scheduled for TACE were consecutively enrolled at one of three participating academic medical centers from May 2016 to March 2022. Each underwent baseline 2D and 3D CE US before TACE, 2D and 3D CE US 1-2 weeks and/or 4-6 weeks after TACE, and CE MRI or CT 4-6 weeks after TACE. CE US and CE MRI or CT were evaluated by three fellowship-trained radiologists for the presence or absence of viable tumors and were compared with reference standards of pathology (18%), angiography on re-treatment after identification of residual disease at 1-2-month follow-up imaging (31%), 4-8-month CE MRI or CT (42%), or short-term (approximately 1-2 months) CE MRI or CT if clinically decompensated and estimated viability was greater than 50% at imaging (9%). Diagnostic performance criteria, including sensitivity and specificity, were obtained for each modality and time point with generalized estimating equation analysis. Results A total of 132 participants were included (mean age, 64 years ± 7 [SD], 87 male). Sensitivity of 2D CE US 4-6 weeks after TACE was 91% (95% CI: 84, 95), which was higher than that of CE MRI or CT (68%; 95% CI: 58, 76; P < .001). Sensitivity of 3D CE US 4-6 weeks after TACE was 89% (95% CI: 81, 94), which was higher than that of CE MRI or CT (P < .001), with no evidence of a difference from 2D CE US (P = .22). CE MRI or CT had 85% (95% CI: 76, 91) specificity, higher than that of 4-6-week 2D and 3D CE US (70% [95% CI: 56, 80] and 67% [95% CI: 53, 78], respectively; P = .046 and P = .023, respectively). No evidence of differences in any diagnostic criteria were observed between 1-2-week and 4-6-week 2D CE US (P > .21). Conclusion The 2D and 3D CE US examinations 4-6 weeks after TACE revealed higher sensitivity in the detection of residual HCC than CE MRI or CT, albeit with lower specificity. Importantly, CE US performance was independent of follow-up time. Clinical trial registration no. NCT02764801 © RSNA, 2023 Supplemental material is available for this article.

Identification of sentinel lymph nodes in esophageal cancer patients using contrast-enhanced EUS with peritumoral injections

Objectives

The objective of this pilot study was to compare the performance of contrast-enhanced EUS (CE-EUS)-guided fine-needle aspiration (FNA) with EUS-FNA for lymph node (LN) staging in esophageal cancer.

Methods

Thirty-seven subjects with esophageal cancer undergoing EUS staging were enrolled, and 30 completed this institutional review board-approved study. A Prosound F75 US system (Hitachi Medical Systems, Tokyo, Japan) with harmonic contrast imaging software and GF-UCT180 curvilinear endoscope (Olympus, Tokyo, Japan) was utilized. All LNs identified by standard EUS were first noted. Sonazoid (dose: 1 mL; GE Healthcare, Oslo, Norway) was administered peritumorally, and all enhanced LNs were recorded. Fine-needle aspiration was performed on LNs considered suspicious by EUS alone, as well as LNs enhanced on CE-EUS. Performance of each modality was compared using FNA cytology as reference standard.

Results

A total of 132 LNs were detected with EUS, of which 59 showed enhancement on CE-EUS. Fifty-three LNs underwent FNA, and 22 LNs were determined to be malignant. Among the latter, 10 were considered suspicious by EUS, whereas the other 12 LNs underwent FNA only because of CE-EUS enhancement. Contrast-enhanced EUS showed enhancement in 19 of the 22 malignant LNs. The rate of metastatic node identification from EUS was 45% (10/22), and it was 86% (19/22; P = 0.008) for CE-EUS. Eight subjects (8/30 [27% of study total]) had nodal status upgraded by the addition of CE-EUS, which influenced LN staging and clinical management.

Conclusions

Fine-needle aspiration of LNs identified by CE-EUS may increase metastasis positive rate by ruling out LNs not associated with the tumor drainage pattern. In addition, CE-EUS seems to identify more metastatic LNs that would not be biopsied under the standard EUS criteria.

Ambient Pressure Sensitivity of the Subharmonic Response of Coated Microbubbles: Effects of Acoustic Excitation Parameters

OBJECTIVE

The sensitivity of the acoustic response of microbubbles, specifically a strong correlation between their subharmonic response and the ambient pressure, has motivated development of a non-invasive subharmonic-aided pressure estimation (SHAPE) method. However, this correlation has previously been found to vary depending on the microbubble type, the acoustic excitation and the hydrostatic pressure range. In this study, the ambient pressure sensitivity of microbubble response was investigated.

METHODS

The fundamental, subharmonic, second harmonic and ultraharmonic responses from an in-house lipid-coated microbubble were measured for excitations with peak negative pressures (PNPs) of 50-700 kPa and frequencies of 2, 3 and 4 MHz in the ambient overpressure range 0-25 kPa (0-187 mmHg) in an in vitro setup.

RESULTS

The subharmonic response typically has three stages-occurrence, growth and saturation-with increasing excitation PNP. We find distinct decreasing and increasing variations of the subharmonic signal with overpressure that are closely related to the threshold of subharmonic generation in a lipid-shelled microbubble. Above the excitation threshold, that is, in the growth-saturation phase, subharmonic signals decreased linearly with slopes as high as -0.56 dB/kPa with ambient pressure increase; below the threshold excitation (at atmospheric pressure), increasing overpressure triggers subharmonic generation, indicating a lowering of subharmonic threshold, and therefore leads to an increase in subharmonic with overpressure, the maximum enhancement being ∼11 dB for 15 kPa overpressure at 2 MHz and 100 kPa PNP.

CONCLUSION

This study indicates the possible development of novel and improved SHAPE methodologies.

Comparison of Magnetic Resonance-Based Elastography and Ultrasound Shear Wave Elastography in Patients With Suspicion of Nonalcoholic Fatty Liver Disease

ABSTRACT

This study investigated the correlation between magnetic resonance elastography (MRE) and shear wave ultrasound elastography (SWE) in patients with clinically diagnosed or suspected nonalcoholic fatty liver disease (NAFLD). Subjects with or at risk of NAFLD identified by magnetic resonance imaging (MRI) proton density fat fraction (PDFF) were prospectively enrolled. For each patient, 6 valid 2-dimensional SWE measurements were acquired using a Logiq E10 scanner (GE HealthCare, Waukesha, WI). A reliability criterion of an interquartile range to median ratio of ≤15% was used for SWE to indicate quality dataset. Magnetic resonance elastography, and MR-fat quantification data were collected the same day as part of the patient's clinical standard of care. Magnetic resonance imaging PDFF was used as a reference to quantify fat with >6.4% indicating NAFLD. Pearson correlation and t-test were performed for statistical analyses. A total of 140 patients were enrolled, 112 of which met SWE reliability measurement criteria. Magnetic resonance elastography and 2-dimensional SWE showed a positive correlation across all study subjects ( r = 0.27; P = 0.004). When patients were grouped according to steatosis and fibrosis state, a positive correlation was observed between MRE and SWE in patients with fibrosis ( r = 0.30; P = 0.03), without fibrosis ( r = 0.27; P = 0.03), and with NAFLD ( r = 0.28; P = 0.02). No elastography technique correlated with liver fat quantification ( P > 0.52). Magnetic resonance elastography was significantly different between patients with and without fibrosis ( P < 0.0001). However, this difference was not apparent with SWE ( P = 0.09). In patients with suspected or known NAFLD, MRE, and SWE demonstrated a positive correlation. In addition, these noninvasive imaging modalities may be useful adjunct techniques for monitoring NAFLD.

Comparison of Gray-scale Inversion to Improve Detection of Pulmonary Nodules on Chest X-rays Between Radiologists and a Deep Convolutional Neural Network

Detection of pulmonary nodules on chest x-rays is an important task for radiologists. Previous studies have shown improved detection rates using gray-scale inversion. The purpose of our study was to compare the efficacy of gray-scale inversion in improving the detection of pulmonary nodules on chest x-rays for radiologists and machine learning models (ML). We created a mixed dataset consisting of 60, 2-view (posteroanterior view - PA and lateral view) chest x-rays with computed tomography confirmed nodule(s) and 62 normal chest x-rays. Twenty percent of the cases were separated for a testing dataset (24 total images). Data augmentation through mirroring and transfer learning was used for the remaining cases (784 total images) for supervised training of 4 ML models (grayscale PA, grayscale lateral, gray-scale inversion PA, and gray-scale inversion lateral) on Google's cloud-based AutoML platform. Three cardiothoracic radiologists analyzed the complete 2-view dataset (n=120) and, for comparison to the ML, the single-view testing subsets (12 images each). Gray-scale inversion (area under the curve (AUC) 0.80, 95% confidence interval (CI) 0.75-0.85) did not improve diagnostic performance for radiologists compared to grayscale (AUC 0.84, 95% CI 0.79-0.88). Gray-scale inversion also did not improve diagnostic performance for the ML. The ML did demonstrate higher sensitivity and negative predictive value for grayscale PA (72.7% and 75.0%), grayscale lateral (63.6% and 66.6%), and gray-scale inversion lateral views (72.7% and 76.9%), comparing favorably to the radiologists (63.9% and 72.3%, 27.8% and 58.3%, 19.5% and 50.5% respectively). In the limited testing dataset, the ML did demonstrate higher sensitivity and negative predictive value for grayscale PA (72.7% and 75.0%), grayscale lateral (63.6% and 66.6%), and gray-scale inversion lateral views (72.7% and 76.9%), comparing favorably to the radiologists (63.9% and 72.3%, 27.8% and 58.3%, 19.5% and 50.5%, respectively). Further investigation of other post-processing algorithms to improve diagnostic performance of ML is warranted.

Assessment of Transarterial Chemoembolization Using Super-resolution Ultrasound Imaging and a Rat Model of Hepatocellular Carcinoma

OBJECTIVE

Hepatocellular carcinoma (HCC) is a highly prevalent form of liver cancer diagnosed annually in 600,000 people worldwide. A common treatment is transarterial chemoembolization (TACE), which interrupts the blood supply of oxygen and nutrients to the tumor mass. The need for repeat TACE treatments may be assessed in the weeks after therapy with contrast-enhanced ultrasound (CEUS) imaging. Although the spatial resolution of traditional CEUS has been restricted by the diffraction limit of ultrasound (US), this physical barrier has been overcome by a recent innovation known as super-resolution US (SRUS) imaging. In short, SRUS enhances the visible details of smaller microvascular structures on the 10 to 100 µm scale, which unlocks a host of new clinical opportunities for US.

METHODS

In this study, a rat model of orthotopic HCC is introduced and TACE treatment response (to a doxorubicin-lipiodol emulsion) is assessed using longitudinal SRUS and magnetic resonance imaging (MRI) performed at 0, 7 and 14 d. Animals were euthanized at 14 d for histological analysis of excised tumor tissue and determination of TACE response, that is, control, partial response or complete response. CEUS imaging was performed using a pre-clinical US system (Vevo 3100, FUJIFILM VisualSonics Inc.) equipped with an MX201 linear array transducer. After administration of a microbubble contrast agent (Definity, Lantheus Medical Imaging), a series of CEUS images were collected at each tissue cross-section as the transducer was mechanically stepped at 100 μm increments. SRUS images were formed at each spatial position, and a microvascular density metric was calculated. Microscale computed tomography (microCT, OI/CT, MILabs) was used to confirm TACE procedure success, and tumor size was monitored using a small animal MRI system (BioSpec 3T, Bruker Corp.).

RESULTS

Although there were no differences at baseline (p > 0.15), both microvascular density levels and tumor size measures from the complete responder cases at 14 d were considerably lower and smaller, respectively, than those in the partial responder or control group animals. Histological analysis revealed tumor-to-necrosis levels of 8.4%, 51.1% and 100%, for the control, partial responder and complete responder groups, respectively (p < 0.005).

CONCLUSION

SRUS imaging is a promising modality for assessing early changes in microvascular networks in response to tissue perfusion-altering interventions such as TACE treatment of HCC.

Improved Tumor Control Following Radiosensitization with Ultrasound-Sensitive Oxygen Microbubbles and Tumor Mitochondrial Respiration Inhibitors in a Preclinical Model of Head and Neck Cancer

Tumor hypoxia (oxygen deficiency) is a major contributor to radiotherapy resistance. Ultrasound-sensitive microbubbles containing oxygen have been explored as a mechanism for overcoming tumor hypoxia locally prior to radiotherapy. Previously, our group demonstrated the ability to encapsulate and deliver a pharmacological inhibitor of tumor mitochondrial respiration (lonidamine (LND)), which resulted in ultrasound-sensitive microbubbles loaded with O₂ and LND providing prolonged oxygenation relative to oxygenated microbubbles alone. This follow-up study aimed to evaluate the therapeutic response to radiation following the administration of oxygen microbubbles combined with tumor mitochondrial respiration inhibitors in a head and neck squamous cell carcinoma (HNSCC) tumor model. The influences of different radiation dose rates and treatment combinations were also explored. The results demonstrated that the co-delivery of O₂ and LND successfully sensitized HNSCC tumors to radiation, and this was also enhanced with oral metformin, significantly slowing tumor growth relative to unsensitized controls (p < 0.01). Microbubble sensitization was also shown to improve overall animal survival. Importantly, effects were found to be radiation dose-rate-dependent, reflecting the transient nature of tumor oxygenation.

Microbubble cavitation restores Staphylococcus aureus antibiotic susceptibility in vitro and in a septic arthritis model

Treatment failure in joint infections is associated with fibrinous, antibiotic-resistant, floating and tissue-associated Staphylococcus aureus aggregates formed in synovial fluid (SynF). We explore whether antibiotic activity could be increased against Staphylococcus aureus aggregates using ultrasound-triggered microbubble destruction (UTMD), in vitro and in a porcine model of septic arthritis. In vitro, when bacterially laden SynF is diluted, akin to the dilution achieved clinically with lavage and local injection of antibiotics, amikacin and ultrasound application result in increased bacterial metabolism, aggregate permeabilization, and a 4-5 log decrease in colony forming units, independent of microbubble destruction. Without SynF dilution, amikacin + UTMD does not increase antibiotic activity. Importantly, in the porcine model of septic arthritis, no bacteria are recovered from the SynF after treatment with amikacin and UTMD-ultrasound without UTMD is insufficient. Our data suggest that UTMD + antibiotics may serve as an important adjunct for the treatment of septic arthritis.

Microwave ablation as a primary versus secondary treatment for hepatocellular carcinoma

PURPOSE

The purpose of this study was to analyze and compare the outcomes of percutaneous microwave ablation (MWA) when used as a primary vs. secondary treatment for hepatocellular carcinoma (HCC).

METHODS

The clinical data of 192 patients with HCC treated with MWA between January 2012 and July 2021 were reviewed retrospectively, with 152 patients being treatment naïve (primary treatment) vs. 40 who had residual or recurrent disease following previous trans-arterial chemoembolization or trans-arterial radioembolization (secondary treatment). The primary outcomes were primary technical efficacy, 1- and 3-year local recurrence-free survival (RFS) and overall survival (OS), local recurrence rates, and adverse events. Pre- and post-intervention liver function tests were compared using a Wilcoxon signed rank test. Univariate and multivariate analyses were also performed, looking at prognostic factors associated with OS and local RFS.

RESULTS

There was no significant difference in 1-year local RFS (primary 93.6% vs. secondary 93.7; P = 0.97) and 3-year local RFS (primary 80.6% vs. secondary 86.5%; P = 0.37) rates. There was no significant difference in 1-year OS (primary 82.4% vs. secondary 86.6%; P = 0.51) and 3-year OS (primary 68.3% vs. secondary 77.4%; P = 0.25) between the two groups. The local recurrence rate (primary 9.8% vs. secondary 14.6%; P = 0.37), primary technical efficacy (primary 96.2% vs. secondary 95%; P = 0.73), and adverse events (primary 8.0% vs. secondary 11.6%; P = 0.45) were also similar between the two groups.

CONCLUSION

Microwave ablation is safe and effective as a secondary treatment for patients with HCC in a clinical salvage scenario and should be utilized more frequently.

Longitudinal ultrasound imaging and network modeling in rats reveal sex-dependent suppression of liver regeneration after resection in alcoholic liver disease

Liver resection is an important surgical technique in the treatment of cancers and transplantation. We used ultrasound imaging to study the dynamics of liver regeneration following two-thirds partial hepatectomy (PHx) in male and female rats fed via Lieber-deCarli liquid diet protocol of ethanol or isocaloric control or chow for 5–7 weeks. Ethanol-fed male rats did not recover liver volume to the pre-surgery levels over the course of 2 weeks after surgery. By contrast, ethanol-fed female rats as well as controls of both sexes showed normal volume recovery. Contrary to expectations, transient increases in both portal and hepatic artery blood flow rates were seen in most animals, with ethanol-fed males showing higher peak portal flow than any other experimental group. A computational model of liver regeneration was used to evaluate the contribution of physiological stimuli and estimate the animal-specific parameter intervals. The results implicate lower metabolic load, over a wide range of cell death sensitivity, in matching the model simulations to experimental data of ethanol-fed male rats. However, in the ethanol-fed female rats and controls of both sexes, metabolic load was higher and in combination with cell death sensitivity matched the observed volume recovery dynamics. We conclude that adaptation to chronic ethanol intake has a sex-dependent impact on liver volume recovery following liver resection, likely mediated by differences in the physiological stimuli or cell death responses that govern the regeneration process. Immunohistochemical analysis of pre- and post-resection liver tissue validated the results of computational modeling by associating lack of sensitivity to cell death with lower rates of cell death in ethanol-fed male rats. Our results illustrate the potential for non-invasive ultrasound imaging to assess liver volume recovery towards supporting development of clinically relevant computational models of liver regeneration.

Ultrasound-Based Machine Learning Approach for Detection of Nonalcoholic Fatty Liver Disease

OBJECTIVES

Current diagnosis of nonalcoholic fatty liver disease (NAFLD) relies on biopsy or MR-based fat quantification. This prospective study explored the use of ultrasound with artificial intelligence for the detection of NAFLD.

METHODS

One hundred and twenty subjects with clinical suspicion of NAFLD and 10 healthy volunteers consented to participate in this institutional review board-approved study. Subjects were categorized as NAFLD and non-NAFLD according to MR proton density fat fraction (PDFF) findings. Ultrasound images from 10 different locations in the right and left hepatic lobes were collected following a standard protocol. MRI-based liver fat quantification was used as the reference standard with >6.4% indicative of NAFLD. A supervised machine learning model was developed for assessment of NAFLD. To validate model performance, a balanced testing dataset of 24 subjects was used. Sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy with 95% confidence interval were calculated.

RESULTS

A total of 1119 images from 106 participants was used for model development. The internal evaluation achieved an average precision of 0.941, recall of 88.2%, and precision of 89.0%. In the testing set AutoML achieved a sensitivity of 72.2% (63.1%-80.1%), specificity of 94.6% (88.7%-98.0%), positive predictive value (PPV) of 93.1% (86.0%-96.7%), negative predictive value of 77.3% (71.6%-82.1%), and accuracy of 83.4% (77.9%-88.0%). The average agreement for an individual subject was 92%.

CONCLUSIONS

An ultrasound-based machine learning model for identification of NAFLD showed high specificity and PPV in this prospective trial. This approach may in the future be used as an inexpensive and noninvasive screening tool for identifying NAFLD in high-risk patients.

Sonoporation for disrupting the pancreatic cancer microenvironment to enhance chemotherapy delivery and improve outcomes

TPS777

Background: Pancreatic ductal adenocarcinoma (PDAC) is 3% of cancers diagnosed in the United States and it is the fourth leading cause of cancer-related deaths. Hence, there is a considerable clinical need to develop innovative strategies for effective drug delivery and treatment. Sonoporation is a novel method that can enhance the therapeutic efficacy of co-administered chemotherapy by localized contrast-enhanced ultrasound imaging (CEUS) of gas-filled microbubbles (ultrasound contrast agent UCA), which temporarily changes tumor vascular microenvironment by increasing angiogenic vessel leakage through microstreaming, shockwaves and the activation of various intracellular signaling responses. Our Phase I clinical trial of sonoporation in 10 PDAC patients treated with Gemcitabine demonstrated no additional toxicity and an increase in median survival compared to 63 historical controls (8.9 vs 17.6 months; p = 0.011). Animal studies investigated 4 commercial UCAs under 2 different acoustic regimes and established the optimal UCA (Sonazoid, GE Healthcare, Oslo, Norway) as well as acoustic settings for sonoporation of PDAC. Methods: This Phase II clinical trial aims to improve standard of care (SoC) chemotherapy treatment by adding sonoporation (i.e., augmenting the SoC treatment with CEUS and microbubbles). Two sites (one in USA and one in Norway) will enroll 120 subjects with PDAC stage III or IIV prior to starting SoC chemotherapy. Exclusion criteria include known allergies to the UCA. The primary objective is to evaluate the safety and therapeutic efficacy of sonoporation on PDAC SoC treatment based on local progression-free and overall survival. Two groups: SoC chemotherapy or SoC chemotherapy followed by sonoporation. The optimal CEUS and microbubble conditions will be applied to a single PDAC tumor imaged by ultrasound. Treatment delivered day of SoC chemotherapeutic treatment for PDAC when the concentration of drugs is maximum. Gehan-Breslow-Wilcoxon test and Log-rank test will be used to compare survival. All clinical variables (e.g., concomitant imaging results, blood tests, etc.) will also be compared between groups with and without sonoporation. Clinical trial information: NCT04821284 .

Microvascular Flow Imaging: A State-of-the-Art Review of Clinical Use and Promise

Vascular imaging with color and power Doppler is a useful tool in the assessment of various disease processes. Assessment of blood flow, from infarction and ischemia to hyperemia, in organs, neoplasms, and vessels, is used in nearly every US investigation. Recent developments in this area are sensitive to small-vessel low velocity flow without use of intravenous contrast agents, known as microvascular flow imaging (MVFI). MVFI is more sensitive in detection of small vessels than color, power, and spectral Doppler, reducing the need for follow-up contrast-enhanced US (CEUS), CT, and MRI, except when arterial and venous wash-in and washout characteristics would be helpful in diagnosis. Varying clinical applications of MVFI are reviewed in adult and pediatric populations, including its technical underpinnings. MVFI shows promise in assessment of several conditions including benign and malignant lesions in the liver and kidney, acute pathologic abnormalities in the gallbladder and testes, and superficial lymph nodes. Future potential of MVFI in different conditions (eg, endovascular repair) is discussed. Finally, clinical cases in which MVFI correlated and potentially obviated additional CEUS, CT, or MRI are shown.

A Comprehensive Motion Compensation Method for In-Plane and Out-of-Plane Motion in Dynamic Contrast-Enhanced Ultrasound of Focal Liver Lesions

Contrast-enhanced ultrasound (CEUS) acquisitions of focal liver lesions are affected by motion, which has an impact on contrast signal quantification. We therefore developed and tested, in a large patient cohort, a motion compensation algorithm called the Iterative Local Search Algorithm (ILSA), which can correct for both periodic and non-periodic in-plane motion and can reject frames with out-of-plane motion. CEUS cines of 183 focal liver lesions in 155 patients from three hospitals were used to develop and test ILSA. Performance was evaluated through quantitative metrics, including the root mean square error and R2 in fitting time-intensity curves and standard deviation value of B-mode intensities, computed across cine frames), and qualitative evaluation, including B-mode mean intensity projection images and parametric perfusion imaging. The median root mean square error significantly decreased from 0.032 to 0.024 (p < 0.001). Median R2 significantly increased from 0.88 to 0.93 (p < 0.001). The median standard deviation value of B-mode intensities significantly decreased from 6.2 to 5.0 (p < 0.001). B-Mode mean intensity projection images revealed improved spatial resolution. Parametric perfusion imaging also exhibited improved spatial detail and better differentiation between lesion and background liver parenchyma. ILSA can compensate for all types of motion encountered during liver CEUS, potentially improving contrast signal quantification of focal liver lesions.

Volumetric Quantitative Contrast-enhanced Ultrasonography Evaluation of Complex Renal Cysts: An Adjunctive Metric to the Bosniak Classification System to Predict Malignancy

BACKGROUND

Management of complex renal cysts is guided by the Bosniak classification system, which may be inadequate for risk stratification of patients for intervention. Fractional tumor vascularity (FV) calculated from volumetric contrast-enhanced ultrasound (CEUS) images may provide additional useful information.

OBJECTIVE

To evaluate CEUS and FV calculation for risk stratification of patients with complex renal cysts.

DESIGN, SETTING, AND PARTICIPANTS

This was a pilot prospective study with institutional review board approval involving patients undergoing surgery for Bosniak IIF-IV complex renal cysts. CEUS was performed preoperatively on the day of surgery with two-dimensional (2D) and three-dimensional (3D) imaging and sulfur hexafluoride lipid-type A microspheres as the ultrasound contrast agent.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

A custom MATLAB program was used to select regions of interest on CEUS scans. FV was calculated according to FV = 1 - (total nonenhancing area/total lesion area). We assessed the ability of 2D- and 3D-derived percentage FV (2DFV%, and 3DFV%) and Bosniak classification schemes (pre-2019 [P2019B] and post-2019 [B2019]) to predict malignancy, aggressive histology, and upstaging on surgical pathology. Performance was assessed as area under the receiver operating characteristic curve (AUC).

RESULTS AND LIMITATIONS

Twenty eligible patients were included in final analysis, of whom 85% (n = 17) had Bosniak IV cysts and 85% (n = 17) had malignant disease on final pathology. Four (24%) of the malignant lesions were International Society of Urological Pathology grade 3-4. The AUC for predicting malignancy was 0.980, 0.824, 0.863, and 0.824 with P2019B, B2019, 2DFV%, and 3DFV%, respectively. When the Bosniak classification was combined with FV%, three models had an AUC of 1, while the combined 2DFV% + B2019 model had AUC of 0.980.

CONCLUSIONS

FV is a novel metric for evaluating complex cystic renal masses and enhances the ability of the Bosniak classification system to predict malignancy. This metric may serve as an adjunct in risk stratification for surgical intervention. Further prospective evaluation is warranted.

PATIENT SUMMARY

Cysts in the kidney are currently classified using a scheme called the Bosniak system. We assessed measurement of the percentage of vascular tissue (called fractional vascularity) in cysts on a special type of ultrasound scan. This promising test adds information when combined with the Bosniak system and can help in guiding appropriate treatment.

Ultrasound Pressure Estimation for Diagnosing Portal Hypertension in Patients Undergoing Dialysis for Chronic Kidney Disease

OBJECTIVES

Hepatic venous pressure gradient (HVPG) is considered the standard in quantifying portal hypertension, but can be unreliable in dialysis patients. A noninvasive ultrasound technique, subharmonic-aided pressure estimation (SHAPE), may be a valuable surrogate of these pressure estimates. This study compared SHAPE and HVPG with pathology findings for fibrosis in dialysis patients.

METHODS

This was a subgroup study from an IRB-approved trial that included 20 patients on dialysis undergoing SHAPE examinations of portal and hepatic veins using a modified Logiq 9 scanner (GE, Waukesha, WI), during infusion of Sonazoid (GE Healthcare, Oslo, Norway). SHAPE was compared to HVPG and pathology findings using the Ludwig-Batts scoring system for fibrosis. Logistic regression, ROC analysis, and t-tests were used to compare HVPG and SHAPE with pathological findings of fibrosis.

RESULTS

Of 20 cases, 5 had HVPG values corresponding to subclinical and clinical portal hypertension (≥6 and ≥10 mmHg, respectively) while 15 had normal HVPG values (≤5 mmHg). SHAPE and HVPG correlated moderately (r = 0.45; P = .047). SHAPE showed a trend toward correlating with fibrosis (r = 0.42; P = .068), while HVPG did not (r = 0.18; P = .45). SHAPE could differentiate between mild (stage 0-1) and moderate to severe (stage 2-4) fibrosis (-10.4 ± 4.9 dB versus -5.4 ± 3.2 dB; P = .035), HVPG could not (3.0 ± 0.6 mmHg versus 4.8 ± 0.7 mmHg; P = .30). ROC curves showed a diagnostic accuracy for SHAPE of 80%, while HVPG reached 76%.

CONCLUSION

Liver fibrosis staging in dialysis patients evaluated for portal hypertension appears to be more accurately predicted by SHAPE than by HVPG; albeit in a small sample size.

Material Properties, Dissolution and Time Evolution of PEGylated Lipid-Shelled Microbubbles: Effects of the Polyethylene Glycol Hydrophilic Chain Configurations

Polyethylene glycol (PEG) is often added to the lipid coating of a contrast microbubble to prevent coalescence and improve circulation. At high surface density, PEG chains are known to undergo a transition from a mushroom configuration to an extended brush configuration. We investigated the effects of PEG chain configuration on attenuation and dissolution of microbubbles by varying the molar ratio of the PEGylated lipid in the shell with three (0%, 2% and 5%) in the mushroom configuration and two (10% and 20%) in the brush configuration. We measured attenuation through the bubble suspensions and used it to obtain the characteristic rheological properties of their shells according to two interfacial rheological models. The interfacial elasticity was found to be significantly lower in the brush regime (∼0.6 N/m) than in the mushroom regime (∼1.3 N/m), but similar in value within each regime. The dissolution behavior of microbubbles under acoustic excitation inside an air-saturated medium was studied by measuring the time-dependent attenuation. Total attenuation recorded a transient increase because of growth resulting from air influx and an eventual decrease caused by dissolution. Microbubble shell composition with varying PEG concentrations had significant effects on dissolution dynamics.

Activation of Phase Change Contrast Agents Using Ionizing Radiation

The feasibility of activating phase change contrast agents (PCCA) made from Definity (Lantheus Medical Imaging) using X-rays was investigated. A 10 mL of Definity PCCA (0.1 mL PCCA/mL) were injected into gelatin phantoms and irradiated using doses of 0, 30, 50, and 100 Gy. Size distribution and PCCA activation were measured after irradiation. Definity PCCAs were activated at a threshold of 50 Gy. Changes were visible with microscopy, visual inspection of T-flasks, and ultrasound imaging of gelatin phantoms. Moreover, increasing the radiation dose above 50 Gy appeared to further activate PCCA. These results indicate Definity PCCAs may be useful for ultrasound-based radiation dosimetry.

Network Meta-Analysis: Noninvasive Imaging Modalities for Identifying Clinically Significant Portal Hypertension

BACKGROUND

Although measurement of the hepatic venous pressure gradient (HVPG) is the current reference standard for obtaining portal venous pressures, several noninvasive imaging-based modalities have been proposed as alternatives.

AIMS

We performed a systematic review and meta-analysis to compare the diagnostic accuracy of noninvasive imaging approaches for identifying clinically significant portal hypertension (CSPH).

METHODS

Two independent reviewers conducted a literature search of PubMed, SCOPUS, and the Cochrane Library from inception until January 5, 2021. The following imaging modalities were compared to HVPG: computed tomography (CT), magnetic resonance imaging (MRI), magnetic resonance elastography, ultrasound, transient elastography (TE), shear wave elastography (SWE), acoustic radiation force impulse (ARFI) imaging, contrast-enhanced ultrasound (CEUS), and subharmonic-aided pressure estimation (SHAPE). Sensitivity, specificity, diagnostic odds ratio (DOR), and area under the curve (AUC) for summary receiver operating characteristic were calculated using both frequentist random effects and Bayesian network meta-analytic approaches.

RESULTS

We analyzed 45 studies of 5678 patients. A broad overlapping confidence interval (CI) of DOR was observed among different imaging modalities: ARFI (30.5; 95% CI 12.7-73.3), CEUS and SHAPE (21.1; 95% CI 6.4-69.8), TE of liver stiffness (21.1; 95% CI 13.3-33.5), CT and MRI (13.7; 95% CI 7.40-25.4), SWE of liver stiffness (10.5; 95% CI 5.2-21.1), and ultrasound (9.5; 95% CI 4.9-18.4). The AUC of all imaging methods exceeded 0.8, indicating very good performance. At a cutoff of 80% specificity, TE, CEUS, and SHAPE exceeded 80% sensitivity.

CONCLUSION

Overall, noninvasive imaging modalities perform well for identifying CSPH. Clinicians should consider these noninvasive and cost-efficient tests when diagnosing CSPH.

3D Harmonic and Subharmonic Imaging for Characterizing Breast Lesions: A Multi-Center Clinical Trial

OBJECTIVE

Breast cancer is the most frequent type of cancer among women. This multi-center study assessed the ability of 3D contrast-enhanced ultrasound to characterize suspicious breast lesions using clinical assessments and quantitative parameters.

METHODS

Women with suspicious breast lesions scheduled for biopsy were enrolled in this prospective, study. Following 2D grayscale ultrasound and power Doppler imaging (PDI), a contrast agent (Definity; Lantheus) was administrated. Contrast-enhanced 3D harmonic imaging (HI; transmitting/receiving at 5.0/10.0 MHz), as well as 3D subharmonic imaging (SHI; transmitting/receiving at 5.8/2.9 MHz), were performed using a modified Logiq 9 scanner (GE Healthcare). Five radiologists independently scored the imaging modes (including standard-of-care imaging) using a 7-point BIRADS scale as well as lesion vascularity and diagnostic confidence. Parametric volumes were constructed from time-intensity curves for vascular heterogeneity, perfusion, and area under the curve. Diagnostic accuracy was determined relative to pathology using receiver operating characteristic (ROC) and reverse, step-wise logistical regression analyses. The κ-statistic was calculated for inter-reader agreement.

RESULTS

Data were successfully acquired in 219 cases and biopsies indicated 164 (75%) benign and 55 (25%) malignant lesions. SHI depicted more anastomoses and vascularity than HI (P < .021), but there were no differences by pathology (P > .27). Ultrasound achieved accuracies of 82 to 85%, which was significantly better than standard-of-care imaging (72%; P < .03). SHI increased diagnostic confidence by 3 to 6% (P < .05), but inter-reader agreements were medium to low (κ < 0.52). The best regression model achieved 97% accuracy by combining clinical reads and parametric SHI.

CONCLUSIONS

Combining quantitative 3D SHI parameters and clinical assessments improves the characterization of suspicious breast lesions.

Improved Sensitivity of Ultrasound-Based Subharmonic Aided Pressure Estimation Using Monodisperse Microbubbles

OBJECTIVES

Subharmonic aided pressure estimation (SHAPE) has been shown effective for noninvasively measuring hydrostatic fluid pressures in a variety of clinical applications. The objective of this study was to explore potential improvements in SHAPE sensitivity using monodisperse microbubbles.

METHODS

Populations of monodisperse microbubbles were created using a commercially available microfluidics device (Solstice Pharmaceuticals). Size distributions were assessed using a Coulter Counter and stability of the distribution following fabrication was evaluated over 24 hours. Attenuation of the microbubble populations from 1 to 10 MHz was then quantified using single element transducers to identify each formulation's resonance frequency. Frequency spectra over increasing driving amplitudes were investigated to determine the nonlinear phases of subharmonic signal generation. SHAPE sensitivity was evaluated in a hydrostatic pressure-controlled water bath using a Logiq E10 scanner (GE Healthcare).

RESULTS

Monodisperse lipid microbubble suspensions ranging from 2.4 to 5.3 μm in diameter were successfully created and they showed no discernable change in size distribution over 24 hours following activation. Calculated resonance frequencies ranged from 2.1 to 6.3 MHz and showed excellent correlation with microbubble diameter (R²  > 0.99). When investigating microbubble frequency response, subharmonic signal occurrence was shown to begin at 150 kPa peak negative pressure, grow up to 225 kPa, and saturate at approximately 250 kPa. Using the Logiq E10, monodisperse bubbles demonstrated a SHAPE sensitivity of -0.17 dB/mmHg, which was nearly twice the sensitivity of the commercial polydisperse microbubble currently being used in clinical trials.

CONCLUSIONS

Monodisperse microbubbles have the potential to greatly improve the sensitivity of SHAPE for the noninvasive measurement of hydrostatic pressures.

Development of a Dual Drug-Loaded, Surfactant-Stabilized Contrast Agent Containing Oxygen

Co-delivery of cancer therapeutics improves efficacy and encourages synergy, but delivery faces challenges, including multidrug resistance and spatiotemporal distribution of therapeutics. To address these, we added paclitaxel to previously developed acoustically labile, oxygen-core, surfactant-stabilized microbubbles encapsulating lonidamine, with the aim of developing an agent containing both a therapeutic gas and two drugs acting in combination. Upon comparison of unloaded, single-loaded, and dual-loaded microbubbles, size (~1.7 µm) and yield (~2 × 10⁹ microbubbles/mL) (~1.7) were not statistically different, nor were acoustic properties (maximum in vitro enhancements roughly 18 dB, in vitro enhancements roughly 18 dB). Both drugs encapsulated above required doses calculated for head and neck squamous cell carcinoma, the cancer of choice. Interestingly, paclitaxel encapsulation efficiency increased from 1.66% to 3.48% when lonidamine was included. During preparation, the combination of single drug-loaded micelles gave higher encapsulation (µg drug/g microbubbles) than micelles loaded with either drug alone (lonidamine, 104.85 ± 22.87 vs. 87.54 ± 16.41), paclitaxel (187.35 ± 8.38 vs. 136.51 ± 30.66). In vivo intravenous microbubbles produced prompt ultrasound enhancement within tumors lasting 3–5 min, indicating penetration into tumor vasculature. The ability to locally destroy the microbubble within the tumor vasculature was confirmed using a series of higher intensity ultrasound pulses. This ability to locally destroy microbubbles shows therapeutic promise that warrants further investigation.

Making waves: how ultrasound-targeted drug delivery is changing pharmaceutical approaches

Administration of drugs through oral and intravenous routes is a mainstay of modern medicine, but this approach suffers from limitations associated with off-target side effects and narrow therapeutic windows. It is often apparent that a controlled delivery of drugs, either localized to a specific site or during a specific time, can increase efficacy and bypass problems with systemic toxicity and insufficient local availability. To overcome some of these issues, local delivery systems have been devised, but most are still restricted in terms of elution kinetics, duration, and temporal control. Ultrasound-targeted drug delivery offers a powerful approach to increase delivery, therapeutic efficacy, and temporal release of drugs ranging from chemotherapeutics to antibiotics. The use of ultrasound can focus on increasing tissue sensitivity to the drug or actually be a critical component of the drug delivery. The high spatial and temporal resolution of ultrasound enables precise location, targeting, and timing of drug delivery and tissue sensitization. Thus, this noninvasive, non-ionizing, and relatively inexpensive modality makes the implementation of ultrasound-mediated drug delivery a powerful method that can be readily translated into the clinical arena. This review covers key concepts and areas applied in the design of different ultrasound-mediated drug delivery systems across a variety of clinical applications.

Contrast-Enhanced Ultrasound in Small Intestinal Ischemia: Proof of Concept

BACKGROUND

Small intestinal ischemia is a challenging diagnosis to make, even with the combination of imaging, laboratory analysis, and physical exam. This pilot study investigated the role of CEUS in evaluating small bowel wall vascularity in participants with suspected ischemia.

METHODS

In this IRB-approved pilot study, CEUS using perflutren lipid microspheres (DEFINITY®; Lantheus Medical Imaging Inc., N. Billerica, MA) was performed on participants determined by the clinical surgical team to have concerns for small intestinal ischemia. CEUS interpretations were performed at both the bedside and later by a blinded radiologist and compared to clinical imaging, surgical findings, or long-term clinical outcomes.

RESULTS

Fifteen CEUS examinations were performed on 14 participants. Five of the participants underwent exploratory laparotomy. Of these, one had small intestinal ischemia (without necrosis). Point of care CEUS demonstrated no evidence of bowel necrosis in any case, and delayed enhancement (indicative of intestinal ischemia) in three cases, resulting in a sensitivity of 100% (95% CI 2.5-100%) and specificity of 85.7% (95% CI 57.2-98.2%). CEUS correctly ruled out ischemia in 91.7% of cases with CT suspicion of small bowel obstruction and 60% of cases that underwent surgical intervention. Additionally, the rate of agreement between bedside interpretation and later radiologist read was high (93%).

CONCLUSIONS

CEUS is uniquely positioned for evaluating the small intestine, because of its high temporal resolution and immediacy of results. Combined with multi-sectional imaging for focal areas of ischemia and/or clinical suspicion for pan ischemia, CEUS may be a useful rule out test for small intestinal ischemia.

Evaluation of Suspected Small Bowel Ischemia Using Contrast-Enhanced Ultrasound with Computed Tomography Fusion

Small bowel ischemia can lead to fatal complications such as necrosis, perforation, and sepsis. Clinical examinations and laboratory tests are usually inconclusive in critically ill patients. The need for surgical exploration is decided based on imaging, examination, and clinical judgment. The decision to operate is time-critical and can be lifesaving, but surgical intervention has the potential to cause additional morbidity, especially in unstable patients. Contrast-enhanced computed tomography (CECT) is the study of choice in suspected small bowel ischemia but has poor specificity. Contrast-enhanced ultrasound (CEUS) provides real-time visualization of the bowel wall vascularity. In this case report, we used a CEUS with CT fusion examination to rule out small bowel ischemia in a critically ill patient with suspected closed loop small bowel obstruction on CECT and in whom surgical exploration would have not been well tolerated. The patient's condition later improved, and an abdominal CT showed no evidence of obstruction.

Diagnostic Value of TI-RADS Classification System and Next Generation Genetic Sequencing in Indeterminate Thyroid Nodules

RATIONALE AND OBJECTIVES

This study aims to evaluate the diagnostic accuracy, inter-reader, and intra-reader variability of the ACR Thyroid Imaging Reporting and Data System (TI-RADS) for risk-stratification of indeterminate thyroid nodules using next generation genetic sequencing and tissue histology as a reference standard.

MATERIALS AND METHODS

Retrospective chart review was performed on all patients who underwent thyroid ultrasound for a nodule with subsequent fine-needle aspiration ± surgical resection from January 2017 to August 2018. Four radiologists with expertise in thyroid ultrasound assessed imaging twice, ≥1 month apart. Results of cytology and next generation genetic sequencing were used as a reference standard for high versus low risk of malignancy in each nodule. Inter-reader reliability between readers and intra-reader reliability between replicate self-reads for TI-RADS categorization were assessed. Univariate analysis, kappa statistics, and receiver operating characteristic curve were calculated.

RESULTS

One hundred and thirty six nodules across 121 patients met inclusion criteria. 84.6% of patients were female and average age was 55.8 ± 14.1 years. One hundred and eighteen of 135 nodules (87%) had indeterminate cytology (Bethesda III or IV). One of 23 high-risk mutations was identified in 30.1% (42) of the nodules. Of the 52 patients who had surgery, 24 (47.1%) had confirmed malignant disease on surgical pathology. Inter-reader reliability between the four radiologists was marginal, κ = 0.293. Intra-reader reliability ranged from marginal to good, κ = 0.337 to κ = 0.560, respectively. The area under the receiver operating characteristic curve was 0.509, and no optimal TI-RADS Level for identifying high-risk nodules existed.

CONCLUSION

The ACR TI-RADS classification system performs with low inter-reader and intra-reader reliability when assessing the genetic risk of nodules with indeterminate cytology.

Assessment of Axillary Lymph Nodes for Metastasis on Ultrasound Using Artificial Intelligence

The purpose of this study was to evaluate an artificial intelligence (AI) system for the classification of axillary lymph nodes on ultrasound compared to radiologists. Ultrasound images of 317 axillary lymph nodes from patients referred for ultrasound guided fine needle aspiration or core needle biopsy and corresponding pathology findings were collected. Lymph nodes were classified into benign and malignant groups with histopathological result serving as the reference. Google Cloud AutoML Vision (Mountain View, CA) was used for AI image classification. Three experienced radiologists also classified the images and gave a level of suspicion score (1-5). To test the accuracy of AI, an external testing dataset of 64 images from 64 independent patients was evaluated by three AI models and the three readers. The diagnostic performance of AI and the humans were then quantified using receiver operating characteristics curves. In the complete set of 317 images, AutoML achieved a sensitivity of 77.1%, positive predictive value (PPV) of 77.1%, and an area under the precision recall curve of 0.78, while the three radiologists showed a sensitivity of 87.8% ± 8.5%, specificity of 50.3% ± 16.4%, PPV of 61.1% ± 5.4%, negative predictive value (NPV) of 84.1% ± 6.6%, and accuracy of 67.7% ± 5.7%. In the three external independent test sets, AI and human readers achieved sensitivity of 74.0% ± 0.14% versus 89.9% ± 0.06% (p = .25), specificity of 64.4% ± 0.11% versus 50.1 ± 0.20% (p = .22), PPV of 68.3% ± 0.04% versus 65.4 ± 0.07% (p = .50), NPV of 72.6% ± 0.11% versus 82.1% ± 0.08% (p = .33), and accuracy of 69.5% ± 0.06% versus 70.1% ± 0.07% (p = .90), respectively. These preliminary results indicate AI has comparable performance to trained radiologists and could be used to predict the presence of metastasis in ultrasound images of axillary lymph nodes.

A Noninvasive Ultrasound Based Technique to Identify Treatment Responders in Patients with Portal Hypertension

RATIONALE AND OBJECTIVE

Subharmonic aided pressure estimation (SHAPE) is based on the inverse relationship between the subharmonic amplitude of ultrasound contrast microbubbles and ambient pressure. The aim of this study was to verify if SHAPE can accurately monitor disease progression in patients identified with portal hypertension.

MATERIALS & METHODS

A modified Logiq 9 scanner with a 4C curvi-linear probe (GE, Waukesha, WI) was used to acquire SHAPE data (transmitting and receiving at 2.5 and 1.25 MHz, respectively) using Sonazoid (GE Healthcare, Oslo, Norway; FDA IND 124,465). Twenty-one (median age 59 years; 12 Males) of the 178 patients enrolled in this institutional review board approved study (14F.113) were identified as having clinically significant portal hypertension based on their hepatic venous pressure gradient results ≥ 10 mmHg. Repeat SHAPE examinations were done every 6.2 months. Liver function tests and clinical indicators were used to establish treatment response.

RESULTS

Of the 21 portal hypertensive subjects, 11 had successful follow up scans with an average follow up time of 6.2 months. There was a significantly larger SHAPE signal reduction in the group who were classified as treatment responders (n = 10; -4.01±3.61 dB) compared to the single nonresponder (2.33 dB; p < 0.001). Results for responders matched the corresponding clinical outcomes of improved model for end stage liver disease (MELD) scores, improvement in underlying cause of portal hypertension, improved liver function tests and reduced ascites.

CONCLUSION

SHAPE can potentially monitor disease progression in portal hypertensive patients and hence, may help clinicians in patient management. A larger study would further validate this claim.

Predicting Long-Term Hepatocellular Carcinoma Response to Transarterial Radioembolization Using Contrast-Enhanced Ultrasound: Initial Experiences

Conventional cross-sectional imaging done shortly after radioembolization of hepatocellular carcinoma (HCC) does not reliably predict long-term response to treatment. This study evaluated whether quantitative contrast-enhanced ultrasound (CEUS) can predict the long-term response of HCC to yttrium-90 (Y-90) treatment. Fifteen patients underwent CEUS at three time points: immediately following treatment and 1 and 2 wk post-treatment. Response 3-6 mo after treatment was categorized on contrast-enhanced magnetic resonance imaging by two experienced radiologists using the Modified Response Evaluation Criteria in Solid Tumors. CEUS data were analyzed by quantifying tumor perfusion and residual fractional vascularity using time-intensity curves. Patients with stable disease on magnetic resonance imaging had significantly greater fractional vascularity 2 wk post-treatment (65.15%) than those with partial or complete response (13.8 ± 9.9%, p = 0.007, and 14.9 ± 15.4%, p = 0.009, respectively). Complete responders had lower tumor vascularity at 2 wk than at post-operative examination (-38.3 ± 15.4%, p = 0.045). Thus, this pilot study suggests CEUS may provide an earlier indication of Y-90 treatment response than cross-sectional imaging.

Shaping the synthesis of surfactant-stabilized oxygen microbubbles to accommodate encapsulated drug

We have developed oxygen filled microbubbles, SE61_O2, for localized, ultrasound-triggered oxygen delivery to hypoxic tumors prior to radiation therapy. Microbubbles, created by sonication, have a shell composed of D-α-Tocopherol polyethylene glycol 1000 succinate (TPGS) and sorbitan monostearate. Preliminary studies in mice with breast tumor xenographs showed that increases in oxygen partial pressure levels lasted less than 3 min, which is insufficient for most clinical applications. Hence, we investigated the potential of incorporating a hydrophobic antiglycolytic drug, modeled with Nile red. A new fabrication method was developed by first creating drug-loaded TPGS micelles. The resulting microbubbles had similar shell compositions, physical size, morphology, and acoustic properties as the original method. However, microbubble yield was more than doubled, resulting in twice the encapsulation efficiency. For the TPGS micelle method these include similar shell compositions (94.4 ± 0.6 % Montane 60), physical size post freeze-drying and reconstitution (1.57 ± 0.42 μm), morphology (spherical), and acoustic properties (maximum enhancement 19.92 ± 0.55 dB). However, microbubble yield was more than doubled, resulting in twice the encapsulation efficiency (up to 10.49 %). We propose that a nonideal mixture is formed when the surfactants are combined by the standard method, resulting in the formation of mixed micelles that are more stable, making microbubble creation more difficult during the sonication step.

Contrast-enhanced ultrasound (CEUS) in HCC diagnosis and assessment of tumor response to locoregional therapies

Hepatocellular carcinoma (HCC) is a global problem constituting the second leading cause of cancer deaths worldwide, thereby necessitating an accurate and cost-effective solution for managing care. Ultrasound is well poised to address this need due to its low cost, portability, safety, and excellent temporal resolution. The role of ultrasound for HCC screening has been well established and supported by multiple international guidelines. Similarly, contrast-enhanced ultrasound (CEUS) can be used for the characterization of focal liver lesions in high-risk populations, and standardized criteria for CEUS have been established by the American College of Radiology Liver Imaging Reporting & Data System (LI-RADS). Following HCC identification, CEUS can also be highly beneficial in treatment planning, delivery, and monitoring HCC response to locoregional therapies. Specific advantages of CEUS include providing real-time treatment guidance and improved diagnostic performance for the detection of residual tumor viability or recurrence, thereby identifying patients in need of retreatment substantially earlier than contrast-enhanced CT and MRI. This review provides a primer on ultrasound and CEUS for the screening and characterization of HCC, with an emphasis on assessing tumor response to locoregional therapies.

Gemcitabine-loaded microbubble system for ultrasound imaging and therapy

Ultrasound imaging presents many positive attributes, including safety, real-time imaging, universal accessibility, and cost. However, inherent difficulties in discrimination between soft tissues and tumors prompted development of stabilized microbubble contrast agents. This presents the opportunity to develop agents in which drug is entrapped in the microbubble shell. We describe preparation and characterization of theranostic poly(lactide) (PLA) and pegylated PLA (PEG-PLA) shelled microbubbles that entrap gemcitabine, a commonly used drug for pancreatic cancer (PDAC). Entrapping 6 wt% gemcitabine did not significantly affect drug activity, microbubble morphology, or ultrasound contrast activity compared with unmodified microbubbles. In vitro microbubble concentrations yielding ≥ 500nM entrapped gemcitabine were needed for complete cell death in MIA PaCa-2 PDAC drug sensitivity assays, compared with 62.5 nM free gemcitabine. In vivo administration of gemcitabine-loaded microbubbles to xenograft MIA PaCa-2 PDAC tumors in athymic mice was well tolerated and provided substantial tumoral image enhancement before and after destructive ultrasound pulses. However, no significant differences in tumor growth were observed among treatment groups, in keeping with the in vitro observation that much higher doses of gemcitabine are required to mirror free gemcitabine activity. STATEMENT OF SIGNIFICANCE: The preliminary results shown here are encouraging and support further investigation into increased gemcitabine loading. Encapsulation of gemcitabine within polylactic acid (PLA) microbubbles does not damage its activity towards pancreatic cancer (pancreatic ductal adenocarcinoma, PDAC) cells. Excellent imaging and evidence of penetration into the highly desmoplastic PDAC tumors is demonstrated. Microbubble destruction was confirmed in vivo, showing that elevated mechanical index shatters the microbubbles for enhanced delivery. The potential to slow PDAC growth in vivo is shown, but higher gemcitabine concentrations are required. Current efforts are directed at increasing drug loading by inclusion of drug-carrying nanoparticles for effective in vivo treatment.

Emerging Applications of Ultrasound-Contrast Agents in Radiation Therapy

Radiation therapy (RT) causes DNA damage through ionization, leading to double-strand breaks. In addition, it generates reactive oxygen species (ROS), which are toxic to tumor cells and the vasculature. However, hypoxic regions in the tumor have been shown to not only decrease treatment response but also increase the likelihood of recurrence and metastasis. Ultrasound-sensitive micro-bubbles are emerging as a useful diagnostic and therapeutic tool within RT. Contrast-enhanced ultrasound (CEUS) has shown great promise in early prediction of tumor response to RT. Ultrasound-triggered micro-bubble cavitation has also been shown to induce bio-effects that can sensitize angiogenic tumor vessels to RT. Additionally, ultrasound can trigger the release of drugs from micro-bubble carriers via localized micro-bubble destruction. This approach has numerous applications in RT, including targeted oxygen delivery before radiotherapy. Furthermore, micro-bubbles can be used to locally create ROS without radiation. Sonodynamic therapy uses focused ultrasound and a sonosensitizer to selectively produce ROS in the tumor region and has been explored as a treatment option for cancer. This review summarizes emerging applications of ultrasound contrast agents in RT and ROS augmentation.

Selecting the optimal parameters for sonoporation of pancreatic cancer in a pre-clinical model

Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest cancers in the modern world, in part due to poor delivery of chemotherapeutics. Sonoporation can be used to enhance the efficacy of standard of care therapies for PDAC. Using xenograft models of PDAC we investigate sonoporation using four ifferent ultrasound contrast agents (UCAs) and two ultrasound regimens to identify the ideal parameters to increase therapeutic efficacy. MIA-PaCa2 xenografts in over 175 immunodeficient mice were treated with gemcitabine and paclitaxel and subjected to low or high power ultrasound (60 and 200 mW/cm² respectively) in conjunction with one of four different UCAs. The UCAs investigated were Definity®, SonoVue®, Optison™ or Sonazoid™. Tumor volumes, vascularity, hemoglobin, and oxygenation were measured and compared to controls. High power treatment in conjunction with Sonazoid sonoporation led to significantly smaller tumors when started early (tumors ~50mm³; p = .0105), while no UCAs significantly increased efficacy in the low power cohort. This trend was also found in larger tumors (~250mm³) where all four UCA agents significantly increased therapeutic efficacy in the high power group (p < .01), while only Definity and SonoVue increased efficacy in the low power cohort (p < .03). Overall, the higher power ultrasound treatment modality was more consistently effective at decreasing tumor volume and increasing vascularity characteristics. In conclusion, Sonazoid was the most consistently effective UCA at decreasing tumor volume and increasing vascularity. Thus, we are pursuing a larger phase II clinical trial to validate the increased efficacy of sonoporation in conjunction with chemotherapy in PDAC patients.

Effects of Contrast-Enhanced Ultrasound of Indeterminate Renal Masses on Patient Clinical Management: Retrospective Analysis From 2 Institutions

OBJECTIVES

To investigate the long-term impact of contrast-enhanced ultrasound (CEUS) on the treatment of patients with indeterminate renal masses.

METHODS

In this retrospective study, consecutive charts of all patients receiving renal CEUS at 1 of 2 academic medical centers between January 1, 2014, and December 31, 2018, were reviewed. Patients were included in the study if they had documented chronic renal disease (estimated glomerular filtration rate < 60 mL/min/1.73 m² ) or prior nephrectomy and received CEUS for a previously untreated renal mass.

RESULTS

A total of 215 lesions in 157 patients were used for analysis. Contrast-enhanced ultrasound provided a final treatment recommendation in 71.6% of lesions (154 of 215). Of these 154 lesions, 7.8% (12 of 154) were lost to follow-up despite CEUS suggesting malignancy; 15.6% (24 of 154) went directly for surgical intervention, with malignancy confirmed by pathologic results in 87.5% (21 of 24) of these cases; and the remaining 76.6% (118 of 154) were deemed benign and required no additional follow-up. Of the 118 lesions diagnosed by CEUS as benign and requiring no follow-up, none showed evidence of later renal cell carcinoma development and, only 5.1% (6 of 118) of the total population was referred for further cross-sectional imaging of the mass in question. In 28.4% of all lesions (61 of 215), CEUS resulted in a recommendation for surveillance imaging at a 6- to 12-month interval, and less than 10% (6 of 61) of these underwent additional cross-sectional imaging within the recommended 6 months after CEUS.

CONCLUSIONS

These findings highlight the impact of CEUS on clinical treatment of indeterminate renal masses, including reducing the use of the potentially nephrotoxic contrast agents and providing a direct pathway to transplant.

US-triggered Microbubble Destruction for Augmenting Hepatocellular Carcinoma Response to Transarterial Radioembolization: A Randomized Pilot Clinical Trial

Background US contrast agents are gas-filled microbubbles (MBs) that can be locally destroyed by using external US. Among other bioeffects, US-triggered MB destruction, also known as UTMD, has been shown to sensitize solid tumors to radiation in preclinical models through localized insult to the vascular endothelial cells. Purpose To evaluate the safety and preliminary efficacy of combining US-triggered MB destruction and transarterial radioembolization (TARE) in participants with hepatocellular carcinoma (HCC). Materials and Methods In this pilot clinical trial, participants with HCC scheduled for sublobar TARE were randomized to undergo either TARE or TARE with US-triggered MB destruction 1-4 hours and approximately 1 and 2 weeks after TARE. Enrollment took place between July 2017 and February 2020. Safety of US-triggered MB destruction was evaluated by physiologic monitoring, changes in liver function tests, adverse events, and radiopharmaceutical distribution. Treatment efficacy was evaluated by using modified Response Evaluation Criteria in Solid Tumors (mRECIST) on cross-sectional images, time to required next treatment, transplant rates, and overall survival. Differences across mRECIST reads were compared by using a Mann-Whitney U test, and the difference in prevalence of tumor response was evaluated by Fisher exact test, whereas differences in time to required next treatment and overall survival curves were compared by using a log-rank (Mantel-Cox) test. Results Safety results from 28 participants (mean age, 70 years ± 10 [standard deviation]; 17 men) demonstrated no significant changes in temperature (P = .31), heart rate (P = .92), diastolic pressure (P = .31), or systolic pressure (P = .06) before and after US-triggered MB destruction. No changes in liver function tests between treatment arms were observed 1 month after TARE (P > .15). Preliminary efficacy results showed a greater prevalence of tumor response (14 of 15 [93%; 95% CI: 68, 100] vs five of 10 [50%; 95% CI: 19, 81]; P = .02) in participants who underwent both US-triggered MB destruction and TARE (P = .02). Conclusion The combination of US-triggered microbubble destruction and transarterial radioembolization is feasible with an excellent safety profile in this patient population and appears to result in improved hepatocellular carcinoma treatment response. © RSNA, 2020.

Contrast-Enhanced Subharmonic Aided Pressure Estimation (SHAPE) using Ultrasound Imaging with a Focus on Identifying Portal Hypertension

Noninvasive, accurate measurement of pressures within the human body has long been an important but elusive clinical goal. Contrast agents for ultrasound imaging are gas-filled, encapsulated microbubbles (diameter < 10 μm) that traverse the entire vasculature and enhance signals by up to 30 dB. These microbubbles also produce nonlinear oscillations at frequencies ranging from the subharmonic (half of the transmit frequency) to higher harmonics. The subharmonic amplitude has an inverse linear relationship with the ambient hydrostatic pressure. Here an ultrasound system capable of performing real-time, subharmonic aided pressure estimation (SHAPE) is presented. During ultrasound contrast agent infusion, an algorithm for optimizing acoustic outputs is activated. Following this calibration, subharmonic microbubble signals (i.e., SHAPE) have the highest sensitivity to pressure changes and can be used to noninvasively quantify pressure. The utility of the SHAPE procedure for identifying portal hypertension in the liver is the emphasis here, but the technique has applicability across many clinical scenarios.

Diagnosing Portal Hypertension with Noninvasive Subharmonic Pressure Estimates from a US Contrast Agent

Background The current standard for assessing the severity of portal hypertension is the invasive acquisition of hepatic venous pressure gradient (HVPG). A noninvasive US-based technique called subharmonic-aided pressure estimation (SHAPE) could reduce risk and enable routine acquisition of these pressure estimates. Purpose To compare quantitative SHAPE to HVPG measurements to diagnose portal hypertension in participants undergoing a transjugular liver biopsy. Materials and Methods This was a prospective cross-sectional trial conducted at two hospitals between April 2015 and March 2019 (ClinicalTrials.gov identifier, NCT02489045). This trial enrolled participants who were scheduled for transjugular liver biopsy. After standard-of-care transjugular liver biopsy and HVPG pressure measurements, participants received an infusion of a US contrast agent and saline. During infusion, SHAPE data were collected from a portal vein and a hepatic vein, and the difference was compared with HVPG measurements. Correlations between data sets were determined by using the Pearson correlation coefficient, and statistical significance between groups was determined by using the Student t test. Receiver operating characteristic analysis was performed to determine the sensitivity and specificity of SHAPE. Results A total of 125 participants (mean age ± standard deviation, 59 years ± 12; 80 men) with complete data were included. Participants at increased risk for variceal hemorrhage (HVPG ≥12 mm Hg) had a higher mean SHAPE gradient compared with participants with lower HVPGs (0.79 dB ± 2.53 vs -4.95 dB ± 3.44; P < .001), which is equivalent to a sensitivity of 90% (13 of 14; 95% CI: 88, 94) and a specificity of 80% (79 of 99; 95% CI: 76, 84). The SHAPE gradient between the portal and hepatic veins was in good overall agreement with the HVPG measurements (r = 0.68). Conclusion Subharmonic-aided pressure estimation is an accurate noninvasive technique for detecting clinically significant portal hypertension. © RSNA, 2020 Online supplemental material is available for this article. See also the editorial by Kiessling in this issue.

Incorporation of a Machine Learning Algorithm With Object Detection Within the Thyroid Imaging Reporting and Data System Improves the Diagnosis of Genetic Risk

Background

The role of next generation sequencing (NGS) for identifying high risk mutations in thyroid nodules following fine needle aspiration (FNA) biopsy continues to grow. However, ultrasound diagnosis even using the American College of Radiology's Thyroid Imaging Reporting and Data System (TI-RADS) has limited ability to stratify genetic risk. The purpose of this study was to incorporate an artificial intelligence (AI) algorithm of thyroid ultrasound with object detection within the TI-RADS scoring system to improve prediction of genetic risk in these nodules.

Methods

Two hundred fifty-two nodules from 249 patients that underwent ultrasound imaging and ultrasound-guided FNA with NGS with or without resection were retrospectively selected for this study. A machine learning program (Google AutoML) was employed for both automated nodule identification and risk stratification. Two hundred one nodules were used for model training and 51 reserved for testing. Three blinded radiologists scored the images of the test set nodules using TI-RADS and assigned each nodule as high or low risk based on the presence of highly suspicious imaging features on TI-RADS (very hypoechoic, taller-than-wide, extra-thyroidal extension, punctate echogenic foci). Subsequently, the TI-RADS classification was modified to incorporate AI for T4 nodules while treating T1-3 as low risk and T5 as high risk. All diagnostic predictions were compared to the presence of a high-risk mutation and pathology when available.

Results

The AI algorithm correctly located all nodules in the test dataset (100% object detection). The model predicted the malignancy risk with a sensitivity of 73.9%, specificity of 70.8%, positive predictive value (PPV) of 70.8%, negative predictive value (NPV) of 73.9% and accuracy of 72.4% during the testing. The radiologists performed with a sensitivity of 52.1 ± 4.4%, specificity of 65.2 ± 6.4%, PPV of 59.1 ± 3.5%, NPV of 58.7 ± 1.8%, and accuracy of 58.8 ± 2.5% when using TI-RADS and sensitivity of 53.6 ± 17.6% (p=0.87), specificity of 83.3 ± 7.2% (p=0.06), PPV of 75.7 ± 8.5% (p=0.13), NPV of 66.0 ± 8.8% (p=0.31), and accuracy of 68.7 ± 7.4% (p=0.21) when using AI-modified TI-RADS.

Conclusions

Incorporation of AI into TI-RADS improved radiologist performance and showed better malignancy risk prediction than AI alone when classifying thyroid nodules. Employing AI in existing thyroid nodule classification systems may help more accurately identifying high-risk nodules.

Superb Microvascular Imaging Improves Detection of Vascularity in Indeterminate Renal Masses

OBJECTIVES

Vascular assessment of indeterminate renal masses (iRMs) remains a crucial element of diagnostic imaging, as the presence of blood flow within renal lesions suggests malignancy. We compared the utility of Superb Microvascular Imaging (SMI; Canon Medical Systems, Tustin, CA), a novel Doppler technique, to standard color Doppler imaging (CDI) and power Doppler imaging (PDI) for the detection of vascularity within iRMs.

METHODS

Patients undergoing contrast-enhanced ultrasound (CEUS) evaluations for iRMs first underwent a renal ultrasound examination with the following modes: CDI, PDI, color Superb Microvascular Imaging (cSMI), and monochrome Superb Microvascular Imaging (mSMI), using an Aplio i800 scanner with an i8CX1 transducer (Canon Medical Systems). After image randomization, each mode was assessed for iRM vascularity by 4 blinded readers on a diagnostic confidence scale of 1 to 5 (5 = most confident). The results were compared to CEUS as the reference standard.

RESULTS

Forty-one patients with 50 lesions met inclusion criteria. Relative to the other 3 modalities, mSMI had the highest sensitivity (63.3%), whereas cSMI had the highest specificity (62.1%). Both cSMI and mSMI also had the highest diagnostic accuracy (0.678 and 0.680, respectively; both P < 0.001) compared to CDI (0.568) and PDI (0.555). Although the reader-reported confidence interval of mSMI (mean ± SD, 3.6 ± 1.1) was significantly lower than CDI (4.1 ± 1.0) and PDI (4.0 ± 1.0; P < 0.001), the confidence level of cSMI (4.1 ± 0.9) was not (P > 0.173).

CONCLUSIONS

Preliminary data suggest that SMI is a potentially useful modality in detecting microvasculature in iRMs compared to standard Doppler techniques. Future studies should aim to compare the efficacy of both SMI and CEUS and to assess the ability of SMI to characterize malignancy in iRMs.

Characterization of indeterminate breast lesions on B-mode ultrasound using automated machine learning models

Purpose: While mammography has excellent sensitivity for the detection of breast lesions, its specificity is limited. Adjunct screening with ultrasound may partially alleviate this issue but also increases false positives, resulting in unnecessary biopsies. Our study investigated the use of Google AutoML Vision (Mountain View, California), a commercially available machine learning service, to both identify and characterize indeterminate breast lesions on ultrasound. Approach: B-mode images from 253 independent cases of indeterminate breast lesions scheduled for core biopsy were used for model creation and validation. The performances of two sub-models from AutoML Vision, the image classification model and object detection model, were evaluated, while also investigating training strategies to enhance model performances. Pathology from the patient's biopsy was used as a reference standard. Results: The image classification models trained under different conditions demonstrated areas under the precision-recall curve (AUC) ranging from 0.85 to 0.96 during internal validation. Once deployed, the model with highest internal performance demonstrated a sensitivity of 100% [95% confidence interval (CI) of 73.5% to 100%], specificity of 83.3% (CI = 51.6 % to 97.9%), positive predictive value (PPV) of 85.7% (CI = 62.9 % to 95.5%), and negative predictive value (NPV) of 100% (CI non-evaluable) in an independent dataset. The object detection model demonstrated lower performance internally during development (AUC = 0.67 ) and during prediction in the independent dataset [sensitivity = 75 % (CI = 42.8 to 94.5), specificity = 80 % (CI = 51.9 to 95.7), PPV = 75 % (CI = 50.8 to 90.0), and NPV = 80 % (CI = 59.3 % to 91.7%)], but was able to demonstrate the location of the lesion within the image. Conclusions: Two models appear to be useful tools for identifying and classifying suspicious areas on B-mode images of indeterminate breast lesions.

Tumor Vascular Networks Depicted in Contrast-Enhanced Ultrasound Images as a Predictor for Transarterial Chemoembolization Treatment Response

Hepatocellular carcinoma (HCC) is prevalent worldwide. Among the various therapeutic options, transarterial chemoembolization (TACE) can be applied to the tumor vascular network by restricting the nutrients and oxygen supply to the tumor. Unique morphologic properties of this network may provide information predictive of future therapeutic responses, which would be significant for decision making during treatment planning. The extraction of morphologic features from the tumor vascular network depicted in abdominal contrast-enhanced ultrasound (CEUS) images faces several challenges, such as organ motion, limited resolution caused by clutter signal and segmentation of the vascular structures at multiple scales. In this study, we present an image processing and analysis approach for the prediction of HCC response to TACE treatment using clinical CEUS images and known pathologic responses. This method focuses on addressing the challenges of CEUS by incorporating a two-stage motion correction strategy, clutter signal removal, vessel enhancement at multiple scales and machine learning for predictive modeling. The morphologic features, namely, number of vessels (NV), number of bifurcations (NB), vessel to tissue ratio (VR), mean vessel length, tortuosity and diameter, from tumor architecture were quantified from CEUS images of 36 HCC patients before TACE treatment. Our analysis revealed that NV, NB and VR are the dominant features for the prediction of long-term TACE response. The model had an accuracy of 86% with a sensitivity and specificity of 89% and 82%, respectively. Reliable prediction of the TACE therapy response using CEUS-derived image features may help to provide personalized therapy planning, which will ultimately improve patient outcomes.

Characterizing Breast Lesions Using Quantitative Parametric 3D Subharmonic Imaging: A Multicenter Study

RATIONALE AND OBJECTIVES

Breast cancer is the leading type of cancer among women. Visualization and characterization of breast lesions based on vascularity kinetics was evaluated using three-dimensional (3D) contrast-enhanced ultrasound imaging in a clinical study.

MATERIALS AND METHODS

Breast lesions (n = 219) were imaged using power Doppler imaging (PDI), 3D contrast-enhanced harmonic imaging (HI), and 3D contrast-enhanced subharmonic imaging (SHI) with a modified Logiq 9 ultrasound scanner using a 4D10L transducer. Quantitative metrics of vascularity derived from 3D parametric volumes (based on contrast perfusion; PER and area under the curve; AUC) were generated by off-line processing of contrast wash-in and wash-out. Diagnostic accuracy of these quantitative vascular parameters was assessed with biopsy results as the reference standard.

RESULTS

Vascularity was observed with PDI in 93 lesions (69 benign and 24 malignant), 3D HI in 8 lesions (5 benign and 3 malignant), and 3D SHI in 83 lesions (58 benign and 25 malignant). Diagnostic accuracy for vascular heterogeneity, PER, and AUC ranged from 0.52 to 0.75, while the best logistical regression model (vascular heterogeneity ratio, central PER, and central AUC) reached 0.90.

CONCLUSION

3D SHI successfully detects contrast agent flow in breast lesions and characterization of these lesions based on quantitative measures of vascular heterogeneity and 3D parametric volumes is promising.

Perfusion-guided sonopermeation of neuroblastoma: a novel strategy for monitoring and predicting liposomal doxorubicin uptake in vivo

Neuroblastoma (NB) is the most common extracranial solid tumor in infants and children, and imposes significant morbidity and mortality in this population. The aggressive chemoradiotherapy required to treat high-risk NB results in survival of less than 50%, yet is associated with significant long-term adverse effects in survivors. Boosting efficacy and reducing morbidity are therefore key goals of treatment for affected children. We hypothesize that these may be achieved by developing strategies that both focus and limit toxic therapies to the region of the tumor. One such strategy is the use of targeted image-guided drug delivery (IGDD), which is growing in popularity in personalized therapy to simultaneously improve on-target drug deposition and assess drug pharmacodynamics in individual patients. IGDD strategies can utilize a variety of imaging modalities and methods of actively targeting pharmaceutical drugs, however in vivo imaging in combination with focused ultrasound is one of the most promising approaches already being deployed for clinical applications. Over the last two decades, IGDD using focused ultrasound with "microbubble" ultrasound contrast agents (UCAs) has been increasingly explored as a method of targeting a wide variety of diseases, including cancer. This technique, known as sonopermeation, mechanically augments vascular permeability, enabling increased penetration of drugs into target tissue. However, to date, methods of monitoring the vascular bioeffects of sonopermeation in vivo are lacking. UCAs are excellent vascular probes in contrast-enhanced ultrasound (CEUS) imaging, and are thus uniquely suited for monitoring the effects of sonopermeation in tumors. Methods: To monitor the therapeutic efficacy of sonopermeation in vivo, we developed a novel system using 2D and 3D quantitative contrast-enhanced ultrasound imaging (qCEUS). 3D tumor volume and contrast enhancement was used to evaluate changes in blood volume during sonopermeation. 2D qCEUS-derived time-intensity curves (TICs) were used to assess reperfusion rates following sonopermeation therapy. Intratumoral doxorubicin (and liposome) uptake in NB was evalauted ex vivo along with associated vascular changes. Results: In this study, we demonstrate that combining focused ultrasound therapy with UCAs can significantly enhance chemotherapeutic payload to NB in an orthotopic xenograft model, by improving delivery and tumoral uptake of long-circulating liposomal doxorubicin (L-DOX) nanoparticles. qCEUS imaging suggests that changes in flow rates are highly sensitive to sonopermeation and could be used to monitor the efficacy of treatment in vivo. Additionally, initial tumor perfusion may be a good predictor of drug uptake during sonopermeation. Following sonopermeation treatment, vascular biomarkers show increased permeability due to reduced pericyte coverage and rapid onset of doxorubicin-induced apoptosis of NB cells but without damage to blood vessels. Conclusion: Our results suggest that significant L-DOX uptake can occur by increasing tumor vascular permeability with microbubble sonopermeation without otherwise damaging the vasculature, as confirmed by in vivo qCEUS imaging and ex vivo analysis. The use of qCEUS imaging to monitor sonopermeation efficiency and predict drug uptake could potentially provide real-time feedback to clinicians for determining treatment efficacy in tumors, leading to better and more efficient personalized therapies. Finally, we demonstrate how the IGDD strategy outlined in this study could be implemented in human patients using a single case study.