Introduction: We use a novel 4-dimensional (4D) volumetric M-mode (VMM) ultrasound (US) technique to assess intravascular volume by monitoring the inferior vena cava (IVC). The VMM method expands the spatial coverage of standard M-mode scanning (depth vs time) by including lateral image direction and adds transducer tilt to cover the region surrounding the IVC. Current ultrasound methods for volume assessment suffer from intra- and inter-operator variability. The VMM technique aims to address these limitations, aiding in early detection of hypovolemia/hemorrhage and guiding resuscitation.
Methods/technical approach: The 4D VMM technique was used on animals that underwent a swine hemorrhagic shock protocol with fluid resuscitation. 2D ultrasound images obtained were formatted in a 3D volume to capture changes over time in vessel size with respiration and volume status. Planes were then extracted from the 3D volume at multiple lateral locations to find and track the IVC. The vessel walls were manually traced on vertical planes (depth vs. time) to determine mean IVC diameter and IVC collapsibility at each measurement time point in the shock/resuscitation protocol. Planes at constant depth (lateral vs. time) were selected to extract respiratory and cardiac cycle data.
Results: Mean IVC diameter in the baseline phase was significantly greater than in the hemorrhage phase (p = 0.020). There was no significant different in mean IVC diameter between baseline and resuscitation (p = 0.064) or hemorrhage and resuscitation phases (p = 0.531). There was no statistically significant difference in mean collapsibility or ΔIVC diameter between protocol phases. The 4D VMM technique effectively measured heart and respiratory rates, consistent with monitored vitals.
Conclusion: 4D VMM identified IVC changes corresponding to blood loss and resuscitation during hemorrhagic shock as well as heart/respiratory rates. This innovative approach holds promise in reducing operator variability and providing actionable information during treatment of shock.

The Circle of Willis (CoW) is a ring-like network of blood vessels that perfuses the brain. Flow in the collateral pathways that connect major arterial inputs in the CoW change dynamically in response to vessel narrowing or occlusion. Vasospasm is an involuntary constriction of blood vessels following subarachnoid hemorrhage (SAH), which can lead to stroke. This study investigated interactions between localization of vasospasm in the CoW, vasospasm severity, anatomical variations, and changes in collateral flow directions. Patient-specific computational fluid dynamics (CFD) simulations were created for 25 vasospasm patients. Computed tomographic angiography scans were segmented capturing the anatomical variation and stenosis due to vasospasm. Transcranial Doppler ultrasound measurements of velocity were used to define boundary conditions. Digital subtraction angiography was analyzed to determine the directions and magnitudes of collateral flows as well as vasospasm severity in each vessel. Percent changes in resistance and viscous dissipation were analyzed to quantify vasospasm severity and localization of vasospasm in a specific region of the CoW. Angiographic severity correlated well with percent changes in resistance and viscous dissipation across all cerebral vessels. Changes in flow direction were observed in collateral pathways of some patients with localized vasospasm, while no significant changes in flow direction were observed in others. CFD simulations can be leveraged to quantify the localization and severity of vasospasm in SAH patients. These factors as well as anatomical variation may lead to changes in collateral flow directions. Future work could relate localization and vasospasm severity to clinical outcomes like the development of infarct.

Obesity is a common finding and a major pathogenetic factor in obstructive sleep apnea (OSA) in adults. To understand the mechanisms behind this, the present study investigated the tissue properties and respiratory kinematics of the tongue base and soft palate in the obese OSA minipig model. In 4 verified obese/OSA and 3 non-obese/non-OSA control minipigs, MRI fat-weighted images, ultrasound elastography (USE), and sleep video-fluoroscopy (SVF) were performed to quantify the fat composition, tissue stiffness, and respiratory kinematics of the tongue base and soft palate during sedated sleep. The results indicated that the fat composition gradually increased from the rostral to caudal tongue base, particularly in the posterior 1/3 of the tongue base, regardless of the presence of obesity and OSA. However, this trend was not seen in the soft palate and pharyngeal wall. The pharyngeal wall presented the highest fat composition as compared with the tongue base and soft palate. Overall, obese OSA minipigs showed stiffer tongue tissue than the controls, particularly in the rostral region of the tongue in obese Yucatan minipigs. The respiratory moving ranges of the soft palate were greater in both dorsal-ventral and rostral-caudal directions and during both respiratory and expiratory phases in OSA obese than control minipigs, and the largest moving ranges were seen in OSA obese Panepinto minipigs. The moving range of the tongue base was significantly smaller. These results suggest more fat infiltration in the caudal region of the tongue base regardless of the presence of obesity and/or OSA. The greater tissue stiffness of the tongue in obese OSA minipigs may result from altered neuromuscular drive.

Tissue pulsatility imaging is an ultrasonic technique that can be used to map regional changes in blood flow in the brain. Classification of regional differences in pulsatility signals can be optimized by restricting the analysis to brain tissue. For 2D transcranial ultrasound imaging, we have implemented an automated image analysis procedure to specify a region of interest in the field of view that corresponds to brain.

Our segmentation method applies an initial K-means clustering algorithm that incorporates both echo strength and tissue displacement to identify skull in ultrasound brain scans. The clustering step is followed by processing steps that use knowledge of the scan format and anatomy to create an image mask that designates brain tissue. Brain regions were extracted from the ultrasound data using different numbers of K-means clusters and multiple combinations of ultrasound data. Masks generated from ultrasound data were compared with reference masks derived from Computed Tomography (CT) data.

A segmentation algorithm based on ultrasound intensity with two K-means clusters achieves an accuracy better than 80% match with the CT data. Some improvement in the match is found with an algorithm that uses ultrasound intensity and displacement data, three K-means clusters, and addition of an algorithm to identify shallow sources of ultrasound shadowing.

Several segmentation algorithms achieve a match of over 80% between the ultrasound and Computed Tomography brain masks. A final tradeoff can be made between processing complexity and the best match of the two data sets.

Bacterial loads can be effectively reduced using cavitation-mediated focused ultrasound, or histotripsy. In this study, gram-negative bacteria (Escherichia coli) in suspension were used as model bacteria to evaluate the effectiveness of two regimens of histotripsy treatments: cavitation histotripsy (CH) and boiling histotripsy (BH).

The results of this study suggest that both CH and BH can be used to inactivate E. coli in suspension, with the optimal regimen depending on the attainable peak negative focal pressure at the target.

Upper urinary tract stones are increasingly prevalent in pet cats and are difficult to manage. Surgical procedures to address obstructing ureteroliths have short- and long-term complications, and medical therapies (e.g., fluid diuresis and smooth muscle relaxants) are infrequently effective. Burst wave lithotripsy is a non-invasive, ultrasound-guided, handheld focused ultrasound technology to disintegrate urinary stones, which is now undergoing human clinical trials in awake unanesthetized subjects.

In this study, we designed and performed in vitro testing of a modified burst wave lithotripsy system to noninvasively fragment stones in cats. The design accounted for differences in anatomic scale, acoustic window, skin-to-stone depth, and stone size. Prototypes were fabricated and tested in a benchtop model using 35 natural calcium oxalate monohydrate stones from cats. In an initial experiment, burst wave lithotripsy was performed using peak ultrasound pressures of 7.3 (n = 10), 8.0 (n = 5), or 8.9 MPa (n = 10) for up to 30 min. Fourteen of 25 stones fragmented to < 1 mm within the 30 min. In a second experiment, burst wave lithotripsy was performed using a second transducer and peak ultrasound pressure of 8.0 MPa (n = 10) for up to 50 min. In the second experiment, 9 of 10 stones fragmented to < 1 mm within the 50 min. Across both experiments, an average of 73–97% of stone mass could be reduced to fragments < 1 mm. A third experiment found negligible injury with in vivo exposure of kidneys and ureters in a porcine animal model.

These data support further evaluation of burst wave lithotripsy as a noninvasive intervention for obstructing ureteroliths in cats.

The Circle of Willis (CoW) is a redundant network of blood vessels that perfuses the brain. The ringlike anatomy mitigates the negative effects of stroke by activating collateral pathways that help maintain physiological perfusion. Previous studies have investigated the activation of these pathways during embolic stroke and internal carotid artery occlusion. However, the role of collateral pathways during cerebral vasospasm–an involuntary constriction of blood vessels after subarachnoid hemorrhage–is not well-documented. This study presents a novel technique to create patient-specific computational fluid dynamics (CFD) simulations of the Circle of Willis before and during vasospasm. Computed tomographic angiography (CTA) scans are segmented to model the vasculature, and transcranial Doppler ultrasound (TCD) measurements of blood flow velocity are applied as boundary conditions. Bayesian analysis leverages information about the uncertainty in the measurements of vessel diameters and velocities to find an optimized parameter set that satisfies mass conservation and that is applied in the final simulation. With this optimized parameter set, the diameters, velocities, and flow rates fall within typical literature values. Virtual angiograms modeled using passive scalar transport agree closely with clinical angiography. A sensitivity analysis quantifies the changes in collateral flow rates with respect to changes in the inlet and outlet flow rates. This analysis can be applied in the future to a cohort of patients to investigate the relationship between the locations and severities of vasospasm, the patient-to-patient anatomical variability in the Circle of Willis, and the activation of collateral pathways.

Phase aberration induced by soft tissue inhomogeneities often complicates high-intensity focused ultrasound (HIFU) therapies by distorting the field and, previously, we designed and fabricated a bilayer gel phantom to reproducibly mimic that effect. A surface pattern containing size scales relevant to inhomogeneities of a porcine body wall was introduced between gel materials with fat- and muscle-like acoustic properties — ballistic and polyvinyl alcohol gels. Here, the phantom design was refined to achieve relevant values of ultrasound absorption and scattering and make it more robust, facilitating frequent handling and use in various experimental arrangements. The fidelity of the interfacial surface of the fabricated phantom to the design was confirmed by three-dimensional ultrasound imaging. The HIFU field distortions — displacement of the focus, enlargement of the focal region, and reduction of focal pressure — produced by the phantom were characterized using hydrophone measurements with a 1.5 MHz 256-element HIFU array and found to be similar to those induced by an ex vivo porcine body wall. A phase correction approach was used to mitigate the aberration effect on nonlinear focal waveforms and enable boiling histotripsy treatments through the phantom or body wall. The refined phantom represents a practical tool to explore HIFU therapy systems capabilities.

Abscesses are walled-off collections of infected fluids that often develop as complications in the setting of surgery and trauma. Treatment is usually limited to percutaneous catheterization with a course of antibiotics. As an alternative to current treatment strategies, a histotripsy approach was developed and tested in a novel porcine animal model. The goal of this article is to use advanced ultrasound imaging modes to extract sonographic features associated with the progression of abscess development in a porcine model. Intramuscular or subcutaneous injections of a bi-microbial bacteria mixture plus dextran particles as an irritant led to identifiable abscesses over a 2 to 3 wk period. Selected abscesses were imaged at least weekly with B-mode, 3-D B-mode, shear-wave elastography and plane-wave Doppler imaging. Mature abscesses were characterized by a well-defined core of varying echogenicity surrounded by a hypoechoic capsule that was highly vascularized on Doppler imaging. 3-D imaging demonstrated the natural history of abscess morphology, with the abscess becoming less complex in shape and increasing in volume. Furthermore, shear-wave elastography demonstrated variations in stiffness as phlegmon becomes abscess and then liquefies, over time. These ultrasound features potentially provide biomarkers to aid in selection of treatment strategies for abscesses.

Infected abscesses are walled-off collections of pus and bacteria. They are a common sequela of complications in the setting of surgery, trauma, systemic infections and other disease states. Current treatment is typically limited to antibiotics with long-term catheter drainage, or surgical washout when inaccessible to percutaneous drainage or unresponsive to initial care efforts. Antibiotic resistance is also a growing concern. Although bacteria can develop drug resistance, they remain susceptible to thermal and mechanical damage. In particular, short pulses of focused ultrasound (i.e., histotripsy) generate mechanical damage through localized cavitation, representing a potential new paradigm for treating abscesses non-invasively, without the need for long-term catheterization and antibiotics. In this pilot study, boiling and cavitation histotripsy treatments were applied to subcutaneous and intramuscular abscesses developed in a novel porcine model. Ultrasound imaging was used to evaluate abscess maturity for treatment monitoring and assessment of post-treatment outcomes. Disinfection was quantified by counting bacteria colonies from samples aspirated before and after treatment. Histopathological evaluation of the abscesses was performed to identify changes resulting from histotripsy treatment and potential collateral damage. Cavitation histotripsy was more successful in reducing the bacterial load while having a smaller treatment volume compared with boiling histotripsy. The results of this pilot study suggest focused ultrasound may lead to a technology for in situ treatment of acoustically accessible abscesses.

Abscesses are walled-off collections of infected fluids containing pus and bacteria. They are often treated with percutaneous drainage in which a drainage catheter may be sutured in place for up to several weeks. Complications such as clogged drains or secondary infections require rehospitalization and wound management. Bacteria are susceptible to mechanical damage, and thus we hypothesize that histotripsy may be a potential new paradigm for treating abscesses noninvasively, without the need for long term catheterization and antibiotics. We developed a porcine animal model that recapitulates some of the features of human abscesses (including size and loculations). Boiling and cavitation histotripsy treatments were applied to subcutaneous and intramuscular abscesses in this porcine model. Ultrasound imaging was used to evaluate abscess maturity, for treatment monitoring and assessment of post-treatment outcomes. Disinfection was quantified by counting bacteria colonies from samples aspirated before and after treatment. Histopathological evaluation of the abscesses was performed to identify changes resulting from histotripsy treatment and potential collateral damage. The results of this pilot study suggest focused ultrasound may lead to a technology for in situ treatment of acoustically accessible abscesses.

Abscesses are walled-off collections of infected fluids most often treated with percutaneous drains placed under CT guidance. Complications such as clogged drains or secondary infections require rehospitalization and wound management. Histotripsy treatment has the potential to eliminate the need for long term catheterization and antibiotics. The progression of abscess development has yet to be fully described. The objective of this study was to use the latest advances in non-contrast ultrasound technologies to characterize abscess development in a porcine animal model. Intramuscular or subcutaneous injections of bacteria plus dextran particles as an irritant led to identifiable abscesses over a 2- to 3-week period. Ultrasound imaging was performed at least weekly, in some cases with a 3D tracking device that provided quantifiable size and shape measurements. Abscess progression was also measured with a plane-wave Doppler mode providing increased sensitivity to low-velocity flows, while abscess stiffness was quantified using shear wave elastography. Most of the mature abscesses were characterized by a rounded core of varying echogenicity surrounded by a hypoechoic capsule that was highly vascularized on Doppler imaging. A treatable abscess was defined by its hypervascular rim and avascular core. Stiffness varied within the abscess but generally decreased over time. Abscess echogenicity, shape, stiffness and vascularity potentially provide features to identify lesions suitable for treatment.

We developed a duplex ultrasound simulator and used it to assess accuracy of volume flow measurements in dialysis access fistula (DAF) models. The simulator consists of a mannequin, computer, and mock transducer. Each case is built from a patient's B-mode images that are used to create a 3-dimensional surface model of the DAF. Computational fluid dynamics is used to determine blood flow velocities based on model vessel geometry. The simulator displays real-time B-mode and color-flow images, and Doppler spectral waveforms are generated according to user-defined settings. Accuracy was assessed by scanning each case and measuring volume flow in the inflow artery and outflow vein for comparison with true volume flow values. Four examiners made 96 volume flow measurements on four DAF models. Measured volume flow deviated from the true value by 35 ± 36%. Mean absolute deviation from true volume flow was lower for arteries than veins (22 ± 19%, N = 48 vs. 58 ± 33%, N = 48, p < 0.0001). This finding is attributed to eccentricity of outflow veins which resulted in underestimating true cross-sectional area. Regression analysis indicated that error in measuring cross-sectional area was a predictor of error in volume flow measurement (β = 0.948, p < 0.001). Volume flow error was reduced from 35 ± 36% to 9 ± 8% (p < 0.000001) by calculating vessel area as an ellipse. Duplex volume flow measurements are based on a circular vessel shape. DAF inflow arteries are circular, but outflow veins can be elliptical. Simulation-based analysis showed that error in measuring volume flow is mainly due to assumption of a circular vessel.

We developed a duplex ultrasound simulator for training and assessment of scanning skills. We used the simulator to test examiner performance in the measurement of flow velocities in dialysis access fistulas. Test cases were created from 3-D ultrasound scans of two dialysis access fistulas by reconstructing 3-D blood vessel models and simulating blood flow velocity fields within the lumens. The simulator displays a 2-D B-mode or color Doppler image corresponding to transducer position on a mannequin; a spectral waveform is generated according to Doppler sample volume location and system settings. Examiner performance was assessed by comparing the measured peak systolic velocity (PSV) with the true PSV provided by the computational flow model. The PSV measured by four expert examiners deviated from the true value by 7.8 ± 6.1%. The results indicate the ability of the simulator to objectively assess an examiner's measurement accuracy in complex vascular targets.

Highlights

A camera-based mapping of venous oxygenation saturation S_vO₂ was further developed.

Monte Carlo method was used for modeling the imaging system.

Curvature and motion correction algorithms were included in the image process.

The spatial resolution for S_vO₂ map achieved was 1.25 mm×1.25 mm.

The measured S_vO₂ was validated by a NIRS system and in line with published data.

Duplex ultrasound scanning with B-mode imaging and both color Doppler and Doppler spectral waveforms is relied upon for diagnosis of vascular pathology and selection of patients for further evaluation and treatment. In most duplex ultrasound applications, classification of disease severity is based primarily on alterations in blood flow velocities, particularly the peak systolic velocity (PSV) obtained from Doppler spectral waveforms. We developed a duplex ultrasound simulator for training and assessment of scanning skills.

Physician-modified endovascular grafts, with fenestrations added to accommodate major branch vessels, provide a means for endovascular treatment of abdominal aortic aneurysms that are adjacent to the renal arteries. Manual measurements of vessel origin locations from computed tomography images, however, take time and can lead to errors in the positions of the fenestrations. To make the fenestration process faster and more accurate, we have developed a procedure to create custom templates that serve as patient-specific guides for graft fenestration. We use a three-dimensional printer to create a clear rigid sleeve that replicates the patient's aorta and includes holes placed precisely at the locations of the branch vessels. The sleeve is slipped over the graft, the locations of the openings are marked with a pen, and the fenestrations are created after the sleeve is removed. Custom fenestration templates can potentially save procedural costs and make minimally invasive aortic aneurysm repair available to more patients.

Arteriovenous fistulae are created surgically to provide adequate access for dialysis patients suffering from end-stage renal disease. It has long been hypothesized that the rapid blood vessel remodeling occurring after fistula creation is in part a process to restore the mechanical stresses to some preferred level, i.e., mechanical homeostasis. The current study presents fluid%u2013structure interaction (FSI) simulations of a patient-specific model of a mature arteriovenous fistula reconstructed from 3D ultrasound scans. The FSI results are compared with previously published data of the same model but with rigid walls. Ultrasound-derived wall motion measurements are also used to validate the FSI simulations of the wall motion. Very large time-averaged shear stresses, 10%u201315 Pa, are calculated at the fistula anastomosis in the FSI simulations, values which are much larger than what is typically thought to be the normal homeostatic shear stress in the peripheral vasculature. Although this result is systematically lower by as much as 50 % compared to the analogous rigid-walled simulations, the inclusion of distensible vessel walls in hemodynamic simulations does not reduce the high anastomotic shear stresses to %u201Cnormal%u201D values. Therefore, rigid-walled analyses may be acceptable for identifying high shear regions of arteriovenous fistulae.

A method of non-contact mapping of venous oxygen saturation (SvO₂) is presented. A CCD camera is used to image skin tissue illuminated alternately by a red (660 nm) and an infrared (800 nm) LED light source. Low cuff pressures of 30–40 mmHg are applied to induce a venous blood volume change with negligible change in the arterial blood volume. A hybrid model combining the Beer–Lambert law and the light diffusion model is developed and used to convert the change in the light intensity to the change in skin tissue absorption coefficient. A simulation study incorporating the full light diffusion model is used to verify the hybrid model and to correct a calculation bias. SvO₂ in the fingers, palm, and forearm for five volunteers are presented and compared with results in the published literature. Two-dimensional maps of venous oxygen saturation are given for the three anatomical regions.

Background
Ultrasonographic (US) assessment of abdominal aortic aneurysms is typically performed by measuring maximal aneurysm diameter from two-dimensional images. These measurements are prone to inaccuracies owing to image planes and interobserver variability. The purpose of this study was to compare the variability in diameter, cross-sectional area (CSA), and volume measurements of abdominal aortic aneurysms obtained using a three-dimensional (3D) US imaging system with those obtained using computed tomographic (CT) angiography, and to determine the reliability of these measures.

Methods
Seven patients in whom endovascular aneurysm repairs were performed underwent CT angiography in addition to a 3D US scan. Measurements computed using 3D surface reconstructions of CT and 3D US scans included maximum diameter, CSA, and aneurysm volume. The seven matched CT and 3D US scans were compared at baseline and 6 to 8 weeks later.

Results
The average aneurysm measured 57.2 mm on CT and 56.2 mm on US (P = 0.14). Correlation coefficients for diameter, CSA, and volume were 0.88, 0.90, and 0.93, respectively (all P values < 0.001). A Bland–Altman analysis demonstrated a strong agreement between 92% of the diameter, 96.4% of the CSA, and 100% of the volume measurements. The interrater reliability was remarkably high comparing the modalities (CT vs. US), and ranged from 0.934 to 0.997 for single measurements and 0.965 to 0.998 for all measurements together; moreover, there was a strong reliability when the tests were reviewed 6 to 8 weeks later, with a reliability of 0.962 to 0.998 for single measurements and 0.992 to 0.999 for all tests (all P values < 0.001).

Conclusions
The 3D US is an accurate and noninvasive method of determining aneurysm size and geometry that is reproducible. Volumetric measurements may represent a significant advancement in long-term follow-up after endovascular aneurysm repair.

Arteriovenous fistulae are surgically created to provide adequate access for dialysis patients suffering from end-stage renal disease. It has long been hypothesized that the rapid blood vessel remodeling occurring after fistula creation is, in part, a process to restore the mechanical stresses to some preferred level, i.e., mechanical homeostasis. We present computational hemodynamic simulations in four patient-specific models of mature arteriovenous fistulae reconstructed from 3D ultrasound scans. Our results suggest that these mature fistulae have remodeled to return to "normal" shear stresses away from the anastomoses: about 1.0 Pa in the outflow veins and about 2.5 Pa in the inflow arteries. Large parts of the anastomoses were found to be under very high shear stresses >15 Pa, over most of the cardiac cycle. These results suggest that the remodeling process works toward restoring mechanical homeostasis in the fistulae, but that the process is limited or incomplete, even in mature fistulae, as evidenced by the elevated shear at or near the anastomoses. Based on the long term clinical viability of these dialysis accesses, we hypothesize that the elevated nonhomeostatic shear stresses in some portions of the vessels were not detrimental to fistula patency.

Background: Duplex ultrasound with Doppler velocimetry is widely used to evaluate the presence and severity of internal carotid artery stenosis; however, a variety of velocity criteria are currently being applied to classify stenosis severity. The purpose of this study is to compare published Doppler velocity measurements to the severity of internal carotid artery stenosis as assessed by x-ray angiography in order to clarify the relationship between these 2 widely used approaches to assess carotid artery disease. Methods: Scatter diagrams or "scattergrams" of correlations between Doppler velocity measurements and stenosis severity as assessed by x-ray contrast angiography were obtained from published articles for native and stented internal carotid arteries. The scattergrams were graphically digitized, combined, and segmented into categories bounded by 50% and 70% diameter reduction. These data were combined and divided into 3 sets representing different velocity parameters: (1) peak systolic velocity, (2) end-diastolic velocity, and (3) the internal carotid artery to common carotid artery peak systolic velocity ratio. The horizontal axis of each scattergram was transformed to form a cumulative distribution function, and thresholds were established for the stenosis categories to assess data variability. Results: Nineteen publications with 22 data sets were identified and included in this analysis. Wide variability was apparent between all 3 velocity parameters and angiographic percent stenosis. The optimal peak systolic velocity thresholds for stenosis in stented carotid arteries were higher than those for native carotid arteries. Within each category of stenosis, the variability of all 3 velocity parameters was significantly lower in stented arteries than in native arteries. Conclusion: Although Doppler velocity criteria have been successfully used to classify the severity of stenosis in both native and stented carotid arteries, the relationship to angiographic stenosis contains significant variability. This analysis of published studies suggests that further refinements in Doppler velocity criteria will not lead to improved correlation with carotid stenosis as demonstrated by angiography.

We report a study of the role of hemodynamic shear stress in the remodeling and failure of a peripheral artery bypass graft. Three separate scans of a femoral to popliteal above-knee bypass graft were taken over the course of a 16 month period following a revision of the graft. The morphology of the lumen is reconstructed from data obtained by a custom 3D ultrasound system. Numerical simulations are performed with the patient-specific geometries and physiologically realistic flow rates. The ultrasound reconstructions reveal two significant areas of remodeling: a stenosis with over 85% reduction in area, which ultimately caused graft failure, and a poststenotic dilatation or widening of the lumen. Likewise, the simulations reveal a complicated hemodynamic environment within the graft. Preliminary comparisons with in vivo velocimetry also showed qualitative agreement with the flow dynamics observed in the simulations. Two distinct flow features are discerned and are hypothesized to directly initiate the observed in vivo remodeling. First, a flow separation occurs at the stenosis. A low shear recirculation region subsequently develops distal to the stenosis. The low shear region is thought to be conducive to smooth muscle cell proliferation and intimal growth. A poststenotic jet issues from the stenosis and subsequently impinges onto the lumen wall. The lumen dilation is thought to be a direct result of the high shear stress and high frequency pressure fluctuations associated with the jet impingement.