Shelby Kutty

University of Nebraska Medical Center, Omaha, Nebraska, United States

Are you Shelby Kutty?

Claim your profile

Publications (118)554.68 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: To assess intervendor agreement of cardiovascular magnetic resonance feature tracking (CMR-FT) and to study the impact of repeated measures on reproducibility. Ten healthy volunteers underwent cine imaging in short-axis orientation at rest and with dobutamine stimulation (10 and 20 μg/kg/min). All images were analysed three times using two types of software (TomTec, Unterschleissheim, Germany and Circle, cvi(42), Calgary, Canada) to assess global left ventricular circumferential (Ecc) and radial (Err) strains and torsion. Differences in intra- and interobserver variability within and between software types were assessed based on single and averaged measurements (two and three repetitions with subsequent averaging of results, respectively) as determined by Bland-Altman analysis, intraclass correlation coefficients (ICC), and coefficient of variation (CoV). Myocardial strains and torsion significantly increased on dobutamine stimulation with both types of software (p<0.05). Resting Ecc and torsion as well as Ecc values during dobutamine stimulation were lower measured with Circle (p<0.05). Intra- and interobserver variability between software types was lowest for Ecc (ICC 0.81 [0.63-0.91], 0.87 [0.72-0.94] and CoV 12.47% and 14.3%, respectively) irrespective of the number of analysis repetitions. Err and torsion showed higher variability that markedly improved for torsion with repeated analyses and to a lesser extent for Err. On an intravendor level TomTec showed better reproducibility for Ecc and torsion and Circle for Err. CMR-FT strain and torsion measurements are subject to considerable intervendor variability, which can be reduced using three analysis repetitions. For both vendors, Ecc qualifies as the most robust parameter with the best agreement, albeit lower Ecc values obtained using Circle, and warrants further investigation of incremental clinical merit. Copyright © 2015 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
    Clinical Radiology 06/2015; DOI:10.1016/j.crad.2015.05.006 · 1.66 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Clinical experience suggests that measurement of left ventricular (LV) ejection fraction (EF) using two-dimensional echocardiography (2DE) is often at variance with results of three-dimensional echocardiography (3DE) in patients who have undergone heart transplantation (HT). The aim of this study was to test the hypothesis that LV mechanical dyssynchrony and abnormal regional strain are present in asymptomatic pediatric HT patients and that they promote errors in the measurement of LV function when 2DE is used. HT subjects and normal volunteer children were prospectively enrolled. All had normal estimated right ventricular systolic pressure and function. LV EF, global and regional strain, and systolic dyssynchrony index (SDI) were quantified using real time 3DE. SDI was determined from volume-time curves of the 16 LV segments and expressed as the standard deviation of the heart rate-corrected time to reach minimal segmental systolic volume. Septal strain was defined as the average of five segments in the interventricular septum. In addition to 3DE, the Teichholz, biplane Simpson, and bullet (5/6 area-length) methods were used to measure EF using 2DE in each subject. Ninety-three examinations were done: 40 in the 40 normal control subjects (mean age, 14.6 ± 10.6 years; 10 male) and 53 in 36 HT subjects (mean age, 10.3 ± 6.2 years; 21 male). SDI was greater in HT patients (mean, 6.2 ± 4.3%) than in normal controls (mean, 2.2 ± 1.1%) (P < .0001). Global and septal strain was lower in HT patients than in normal controls. EF divergence (absolute difference between two- and three-dimensional EFs) was greater in HT patients (mean, 3.8 ± 2.2%) than in normal controls (mean, 0.7 ± 0.5%) (P < .0001). EF divergence had a strong positive correlation with SDI (adjusted r(2) = 0.46, P < .001) and negative correlations with all measures of strain (range of adjusted r(2) values, 0.13-0.32). SDI had no particular relation to LV mass or to QRS duration. Children after HT have abnormal LV mechanics characterized by greater dyssynchrony and lower strain. These features correlate with, and possibly contribute to differences between measurements by 2DE and 3DE. EF should be calculated using 3DE in this population and others with dyssynchrony and regional strain abnormalities. Copyright © 2015 American Society of Echocardiography. Published by Elsevier Inc. All rights reserved.
    Journal of the American Society of Echocardiography: official publication of the American Society of Echocardiography 06/2015; DOI:10.1016/j.echo.2015.05.013 · 3.99 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To determine the inter-study reproducibility of MR feature tracking (MR-FT) derived left ventricular (LV) torsion and torsion rates for a combined assessment of systolic and diastolic myocardial function. Steady-state free precession (SSFP) cine LV short-axis stacks were acquired at 9:00 (Exam A), 9:30 (Exam B), and 14:00 (Exam C) in 16 healthy volunteers at 3 Tesla. SSFP images were analyzed offline using MR-FT to assess rotational displacement in apical and basal slices. Global peak torsion, peak systolic and peak diastolic torsion rates were calculated using different definitions ("twist", "normalized twist" and "circumferential-longitudinal (CL) shear angle"). Exam A and B were compared to assess the inter-study reproducibility. Morning and afternoon scans were compared to address possible diurnal variation. The different methods showed good inter-study reproducibility for global peak torsion (intraclass correlation coefficient [ICC]: 0.90-0.92; coefficient of variation [CoV]: 19.0-20.3%) and global peak systolic torsion rate (ICC: 0.82-0.84; CoV: 25.9-29.0%). Conversely, global peak diastolic torsion rate showed little inter-study reproducibility (ICC: 0.34-0.47; CoV: 40.8-45.5%). Global peak torsion as determined by the CL shear angle showed the best inter-study reproducibility (ICC: 0.90;CoV: 19.0%). MR-FT results were not measurably affected by diurnal variation between morning and afternoon scans (CL shear angle: 4.8 ± 1.4°, 4.8 ± 1.5°, and 4.1 ± 1.6° for Exam A, B, and C, respectively; P = 0.21). MR-FT based derivation of myocardial peak torsion and peak systolic torsion rate has high inter-study reproducibility as opposed to peak diastolic torsion rate. The CL shear angle was the most reproducible parameter independently of cardiac anatomy and may develop into a robust tool to quantify cardiac rotational mechanics in longitudinal MR-FT patient studies. J. Magn. Reson. Imaging 2015. © 2015 Wiley Periodicals, Inc.
    Journal of Magnetic Resonance Imaging 06/2015; DOI:10.1002/jmri.24979 · 2.79 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To investigate technical approaches for transcatheter closure of coronary artery fistula based on anatomic type of the fistula. The variability in coronary artery fistulae (CAF) anatomy that necessitates different transcatheter closure (TCC) approaches has not been well documented. Records of patients with CAF who underwent TCC at 2 centers were reviewed for technical details and procedural outcome. CAF were classified as proximal and distal. TCC approaches employed were arterio-venous or arterio-arterial loop, retrograde arterial, and antegrade venous. Eighteen patients with CAF, mean age 12.6 years (0.07-60), 11 male (61%), underwent TCC. All CAF drained predominantly into the right side of the heart. Types of CAF were proximal in 15 and distal in 3 patients. CAF calibers were large in 7, medium in 9, and small in 2 patients. The arterio-venous loop approach was used in the majority of the cases (11 patients) and the CAF size were medium to large. The retrograde arterial approach was used in 4; of these, 3 patients had small to medium sized CAF. In 2 patients with long tortuous CAF an antegrade venous approach was employed. TCC was successful in 17 of the 18 patients (94.4%). There were no peri-procedural deaths or vascular complications. This study documents transcatheter closure approaches for CAF and device selection based on fistula origin. The choices of TCC technique and device selection vary, and are primarily determined by the heterogeneous anatomic characteristics of the fistulae. © 2015, Wiley Periodicals, Inc.
    Journal of Interventional Cardiology 06/2015; DOI:10.1111/joic.12212 · 1.32 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background: Cardiovascular magnetic resonance (CMR) offers quantification of phasic atrial functions based on volumetric assessment and more recently, on CMR feature tracking (CMR-FT) quantitative strain and strain rate (SR) deformation imaging. Inter-study reproducibility is a key requirement for longitudinal studies but has not been defined for CMR-based quantification of left atrial (LA) and right atrial (RA) dynamics. Methods: Long-axis 2-and 4-chamber cine images were acquired at 9:00 (Exam A), 9:30 (Exam B) and 14:00 (Exam C) in 16 healthy volunteers. LA and RA reservoir, conduit and contractile booster pump functions were quantified by volumetric indexes as derived from fractional volume changes and by strain and SR as derived from CMR-FT. Exam A and B were compared to assess the inter-study reproducibility. Morning and afternoon scans were compared to address possible diurnal variation of atrial function.
    Journal of Cardiovascular Magnetic Resonance 05/2015; 17(1):36. DOI:10.1186/s12968-015-0140-2 · 5.11 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Predischarge (pre-d/c) transthoracic echocardiography (TTE) is routine after surgery for congenital heart disease, but how it affects clinical care is unknown. The aim of this study was to test the hypothesis that pre-d/c TTE frequently reveals findings associated with short-term clinical course through a systematic review of findings on pre-d/c TTE and clinical events that followed. Clinical outcomes of mortality, hospitalization, catheterization, and surgery at 1 year were examined for pediatric patients undergoing pre-d/c TTE between June 2010 and June 2012. Using logistic regression, a multivariate model was generated associating clinical, pre-d/c transthoracic echocardiographic, and demographic variables with unplanned postdischarge cardiac events (UCEs) within 1 year. Of 462 patients who underwent pre-d/c TTE, there were 265 male patients (57%) and 197 female patients (43%); the median age was 0.8 years (range, 0-33 years). Two hundred thirty-seven patients (51%) had findings (valve regurgitation, hemodynamic obstruction, ventricular dysfunction, unintended shunt, or pericardial effusion) on pre-d/c TTE, 57 of which were of more than mild severity. Agreement between pre-d/c TTE and postoperative transesophageal echocardiographic findings was only fair to moderate (κ = 0.27-0.43). Sixty-four patients (14%) had UCEs. Univariate analysis revealed that UCE were more frequent in patients with diagnoses and surgical procedures of high complexity. After accounting for these confounding nonechocardiographic variables, pre-d/c transthoracic echocardiographic findings, specifically valve regurgitation of more than mild severity, and ventricular dysfunction and obstructions of any severity were independently associated with UCEs (odds ratios, 1.90, 1.99, and 1.85, respectively). Findings on pre-d/c TTE are frequent, commonly discordant with postoperative transesophageal echocardiographic results, and associated with adverse clinical events after surgery for congenital heart disease. These data would strongly support the practice of pre-d/c TTE after surgery for congenital heart disease. Copyright © 2015 American Society of Echocardiography. Published by Elsevier Inc. All rights reserved.
    Journal of the American Society of Echocardiography: official publication of the American Society of Echocardiography 05/2015; DOI:10.1016/j.echo.2015.04.009 · 3.99 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To derive and validate a multivariate stratification model for prediction of survival free from intervention (SFFI) in ventricular septal defect (VSD). A secondary aim is for this model to serve as proof of concept for derivation of a more general congenital heart disease prognostic model, of which the VSD model will be the first component. For 12 years, 2334 subjects with congenital heart disease were prospectively and consecutively enrolled. Of these, 675 had VSD and form the derivation cohort. One hundred seven other subjects with VSD followed in another practice formed the validation cohort. The derivation cohort was serially stratified based on clinical and demographic features correlating with SFFI. Six strata were defined, the most favorable predicting nearly 100% SFFI at 10 years, and the least favorable, a high likelihood of event within weeks. Strata with best SFFI had many subjects with nearly normal physiology, muscular VSD location, or prior intervention. In the validation cohort, the relation between predicted and actual SFFI at 6 months, 1 year, 2 years, and 5 years follow-up had areas under the receiver operating characteristic curves 0.800 or greater. A prediction model for SFFI in VSD has been derived and validated. It has potential for clinical application to the benefit of patients and families, medical trainees, and practicing physicians. Copyright © 2015 Elsevier Inc. All rights reserved.
    The Journal of pediatrics 04/2015; DOI:10.1016/j.jpeds.2015.04.005 · 3.74 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Our purpose was to evaluate yield of tools commonly advocated for surveillance of tetralogy of Fallot (TOF). All patients (pts) with TOF, seen at any time from 1/2008 to 9/2013 in an academic cardiology practice were studied. At the first and each subsequent outpatient visit, the use of tools including history and physical (H&P), ECG, Holter (HOL), echocardiogram (Echo), MR or CT (MR-CT) and stress testing (STR) were noted. Recommendations for intervention (INT) and for time to next follow-up were recorded; rationale for each INT with attribution to one or more tools was identified. There were 213 pts (mean 11.5 years, 130 male) who had 916 visits, 123 of which (13.4%) were associated with 138 INTs (47 surgical, 54 catheter-mediated, 37 other medical). Recommended follow-up interval was 9.44±6.5 months, actual mean follow-up interval was 11.7 months. All 916 (100%) patient visits had a H&P which contributed to 47.2% of INT decisions. Echo was performed in 652 (71.2%) of visits, and contributed to 53.7% of INTs. MR-CT was obtained in 129 (14.1%) of visits, and contributed to 30.1% of INTs. ECG was applied in 137 (15%) visits, and contributed to 1.6% of INTs. HOL was obtained in 188 (20.5%) visits, and contributed to 11.3% of INTs. STR was performed at 101 (11%) of visits, and contributed to 8.9% of INTs. INTs are common in repaired TOF, but when visits average every 11-12 months, most visits do not result in INT. H&P, Echo and HOL were the most frequently applied screens, and all frequently yielded relevant information to guide INT decisions. STR and MR/CT were applied as targeted testing and in this limited, non-screening role had high relevance for INT. There was low utilisation of ECG and major impact on INT was not demonstrated. Risk stratification in TOF may be possible, and could result in more efficient surveillance and targeted testing.
    04/2015; 2(1):e000185. DOI:10.1136/openhrt-2014-000185
  • [Show abstract] [Hide abstract]
    ABSTRACT: Background Beta-blockers contribute to treatment of heart failure. Their mechanism of action, however, is incompletely understood. Gradients in beta-blocker sensitivity of helically aligned cardiomyocytes as compared to counteracting transversely intruding cardiomyocytes seem crucial. We hypothesize that selective blockade of transversely intruding cardiomyocytes by low-dose beta-blockade unloads ventricular performance. Cardiac magnetic resonance imaging (MRI) 3D-tagging delivers parameters of myocardial performance. Methods and Results We studied 13 healthy volunteers by MRI 3D-tagging during escalated i.v.-administration of Esmolol. The circumferential, longitudinal and radial myocardial shortening was determined for each dose. The curves were analysed for peak value, time-to-peak, upslope, and area-under-the-curve. At low doses, from 5 to 25 μg/kg/min, peak contraction increased while time-to-peak decreased yielding a steeper upslope. Combining the values revealed a left shift of the curves at low doses when compared to baseline without Esmolol. At doses of 50 to 150 μg/kg/min, a right shift with flattening occurred. Conclusions In healthy volunteers we found more pronounced myocardial shortening at low compared to clinical dosage of beta-blockers. In patients with ventricular hypertrophy and higher prevalence of transversely intruding cardiomyocytes selective low-dose beta-blockade could be even more effective. MRI 3D-tagging could help to determine optimal individual beta-blocker dosing avoiding undesirable side effects. Copyright © 2014, American Journal of Physiology - Heart and Circulatory Physiology.
    AJP Heart and Circulatory Physiology 04/2015; DOI:10.1152/ajpheart.00746.2014 · 4.01 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Objectives: To determine whether quantitative wall motion assessment by CMR myocardial feature tracking (CMR-FT) would reduce the impact of observer experience as compared to visual analysis in patients with ischemic cardiomyopathy (ICM). Methods: 15 consecutive patients with ICM referred for assessment of hibernating myocardium were studied at 3 Tesla using SSFP cine images at rest and during low dose dobutamine stress (5 and 10 μg/kg/min of dobutamine). Conventional visual, qualitative analysis was per- formed independently and blinded by an experienced and an inexperienced reader, fol- lowed by post-processing of the same images by CMR-FT to quantify subendocardial and subepicardial circumferential (Eccendo and Eccepi) and radial (Err) strain. Receiver operator characteristics (ROC) were assessed for each strain parameter and operator to detect the presence of inotropic reserve as visually defined by the experienced observer. Results: 141 segments with wall motion abnormalities at rest were eligible for the analysis. Visual scoring of wall motion at rest and during dobutamine was significantly different between the experienced and the inexperienced observer (p<0.001). All strain values (Eccendo, Eccepi and Err) derived during dobutamine stress (5 and 10 μg/kg/min) showed similar diagnostic accuracy for the detection of contractile reserve for both operators with no differences in ROC (p>0.05). Eccendo was the most accurate (AUC of 0.76, 10 μg/kg/min of dobutamine) parameter. Diagnostic accuracy was worse for resting strain with differences between oper- ators for Eccendo and Eccepi (p<0.05) but not Err (p>0.05). Conclusion: Whilst visual analysis remains highly dependent on operator experience, quantitative CMR- FT analysis of myocardial wall mechanics during DS-CMR provides diagnostic accuracy for the detection of inotropic reserve regardless of operator experience and hence may improve diagnostic robustness of low-dose DS-CMR in clinical practice.
    PLoS ONE 04/2015; 10(4):e0122858. DOI:10.1371/journal.pone.0122858 · 3.53 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Aims Ebstein's anomaly (EA) involves a displaced and dysplastic tricuspid valve resulting in an atrialized portion of the right ventricle and an enlargement of the functional right ventricle and right atrium. Biomarkers targeting heart failure such as brain natriuretic peptide (BNP) or haematological parameters [haemoglobin (Hb) and haematocrit (Hct)] are upregulated in states of pulmonary hypoperfusion. We hypothesized that decreased pulmonary perfusion dependent on the stage of right heart failure is a possible mechanism in EA, and that it can be correlated with cardiac magnetic resonance (CMR) parameters. The aim of this study was to investigate the relationship between BNP and haematological parameters with functional parameters from CMR and exercise testing in patients with EA.
    European Heart Journal – Cardiovascular Imaging 03/2015; DOI:10.1093/ehjci/jeu312 · 2.65 Impact Factor
  • Journal of the American College of Cardiology 03/2015; 65(10):A1291. DOI:10.1016/S0735-1097(15)61291-8 · 15.34 Impact Factor
  • Journal of the American College of Cardiology 03/2015; 65(10):A1292. DOI:10.1016/S0735-1097(15)61292-X · 15.34 Impact Factor
  • Journal of the American College of Cardiology 03/2015; 65(10):A517. DOI:10.1016/S0735-1097(15)60517-4 · 15.34 Impact Factor
  • Journal of the American College of Cardiology 03/2015; 65(10):A487. DOI:10.1016/S0735-1097(15)60487-9 · 15.34 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Pulmonary arterial hypertension (PAH) is an obstructive pulmonary vasculopathy, characterized by excess proliferation, apoptosis resistance, inflammation, fibrosis, and vasoconstriction. Although PAH therapies target some of these vascular abnormalities (primarily vasoconstriction), most do not directly benefit the right ventricle (RV). This is suboptimal because a patient's functional state and prognosis are largely determined by the success of the adaptation of the RV to the increased afterload. The RV initially hypertrophies but might ultimately decompensate, becoming dilated, hypokinetic, and fibrotic. A number of pathophysiologic abnormalities have been identified in the PAH RV, including: ischemia and hibernation (partially reflecting RV capillary rarefaction), autonomic activation (due to G protein receptor kinase 2-mediated downregulation and desensitization of β-adrenergic receptors), mitochondrial-metabolic abnormalities (notably increased uncoupled glycolysis and glutaminolysis), and fibrosis. Many RV abnormalities are detectable using molecular imaging and might serve as biomarkers. Some molecular pathways, such as those regulating angiogenesis, metabolism, and mitochondrial dynamics, are similarly deranged in the RV and pulmonary vasculature, offering the possibility of therapies that treat the RV and pulmonary circulation. An important paradigm in PAH is that the RV and pulmonary circulation constitute a unified cardiopulmonary unit. Clinical trials of PAH pharmacotherapies should assess both components of the cardiopulmonary unit. Copyright © 2015 Canadian Cardiovascular Society. Published by Elsevier Inc. All rights reserved.
    The Canadian journal of cardiology 01/2015; 31(4). DOI:10.1016/j.cjca.2015.01.023 · 3.94 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Palliative shunts in congenital heart disease patients are vulnerable to thrombotic occlusion. High mechanical index (MI) impulses from a modified diagnostic ultrasound (US) transducer during a systemic microbubble (MB) infusion have been used to dissolve intravascular thrombi without anticoagulation, and we sought to determine whether this technique could be used prophylactically to reduce thrombus burden and prevent occlusion of surgically placed extracardiac shunts. Heparin-bonded ePTFE tubular vascular shunts of 4 mm×2.5 cm (Propaten; W.L Gore) were surgically placed in 18 pigs: a right-sided side-to-side arteriovenous (AV, carotid-jugular) shunt, and a left-sided arterio-arterial (AA, carotid-carotid) interposition shunt in each animal. After shunt implantation, animals were randomly assigned to one of 3 groups. Transcutaneous, weekly 30-minute treatments (total of 4 treatments) of either guided high MI US+MB (Group 1; n=6) using a 3% MRX-801 MB infusion, or US alone (Group 2; n=6) were given separately to each shunt. The third group of 6 pigs received no treatments. The shunts were explanted after 4 weeks and analyzed by histopathology to quantify luminal thrombus area (mm(2)) for the length of each shunt. No pigs received antiplatelet agents or anticoagulants during the treatment period. The median overall thrombus burden in the 3 groups for AV shunts was 5.10 mm(2) compared with 4.05 mm(2) in AA (P=0.199). Group 1 pigs had significantly less thrombus burden in the AV shunts (median 2.5 mm(2)) compared with Group 2 (median 5.6 mm(2)) and Group 3 (median 7.5 mm(2)) pigs (P=0.006). No difference in thrombus burden was seen between groups for AA shunts. Transcutaneous US with intravenous MB is capable of preventing thrombus accumulation in arteriovenous shunts without the need for antiplatelet agents, and may be a method of preventing progressive occlusion of palliative shunts.
    Journal of the American Heart Association 12/2014; 3(1):e000689. DOI:10.1161/JAHA.113.000689 · 2.88 Impact Factor
  • European Heart Journal – Cardiovascular Imaging 12/2014; 15 Suppl 2:ii52-ii54. DOI:10.1093/ehjci/jeu247 · 2.65 Impact Factor
  • American Heart Association, Chicago , IL 2014; 11/2014
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: No expert consensus guides practice for intensity of ongoing pediatric cardiology surveillance of hemodynamically insignificant small and moderate muscular ventricular septal defect (mVSD). Therefore, despite the well-established benign natural history of mVSD, there is potential for widely divergent follow up practices. The purpose of this investigation was to evaluate (1) variations in follow up of mVSD within an academic children's hospital based pediatric cardiology practice, and (2) the frequency of active medical or surgical management resulting from follow up of mVSD.
    BMC Pediatrics 11/2014; 14(1):282. DOI:10.1186/1471-2431-14-282 · 1.92 Impact Factor

Publication Stats

630 Citations
554.68 Total Impact Points

Institutions

  • 2010–2015
    • University of Nebraska Medical Center
      • Division of Pediatric Cardiology
      Omaha, Nebraska, United States
  • 2008–2015
    • University of Nebraska at Omaha
      • Division of Cardiology
      Omaha, Nebraska, United States
    • Medical College of Wisconsin
      Milwaukee, Wisconsin, United States
  • 2008–2014
    • Creighton University
      Omaha, Nebraska, United States
  • 2013
    • Deutsches Herzzentrum Berlin
      • The Department of Congenital Heart Disease / Pediatric Cardiology
      Berlín, Berlin, Germany
  • 2011
    • University of Massachusetts Boston
      Boston, Massachusetts, United States
  • 2007–2008
    • Children's Hospital of Wisconsin
      Madison, Wisconsin, United States