Stuart Crozier

University of Queensland, Brisbane, Queensland, Australia

Are you Stuart Crozier?

Claim your profile

Publications (464)576.4 Total impact

  • Elliot Smith · Fabio Freschi · Maurizio Repetto · Stuart Crozier
    IEEE transactions on bio-medical engineering 09/2015; DOI:10.1109/TBME.2015.2479230 · 2.35 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To develop an automated approach for 3D quantitative assessment and measurement of alpha angles from the femoral head-neck (FHN) junction using bone models derived from magnetic resonance (MR) images of the hip joint.Bilateral MR images of the hip joints were acquired from 30 male volunteers (healthy active individuals and high-performance athletes, aged 18-49 years) using a water-excited 3D dual echo steady state (DESS) sequence. In a subset of these subjects (18 water-polo players), additional True Fast Imaging with Steady-state Precession (TrueFISP) images were acquired from the right hip joint. For both MR image sets, an active shape model based algorithm was used to generate automated 3D bone reconstructions of the proximal femur. Subsequently, a local coordinate system of the femur was constructed to compute a 2D shape map to project femoral head sphericity for calculation of alpha angles around the FHN junction. To evaluate automated alpha angle measures, manual analyses were performed on anterosuperior and anterior radial MR slices from the FHN junction that were automatically reformatted using the constructed coordinate system.High intra- and inter-rater reliability (intra-class correlation coefficients > 0.95) was found for manual alpha angle measurements from the auto-extracted anterosuperior and anterior radial slices. Strong correlations were observed between manual and automatic measures of alpha angles for anterosuperior (r = 0.84) and anterior (r = 0.92) FHN positions. For matched DESS and TrueFISP images, there were no significant differences between automated alpha angle measures obtained from the upper anterior quadrant of the FHN junction (two-way repeated measures ANOVA, F < 0.01, p = 0.98).Our automatic 3D method analysed MR images of the hip joints to generate alpha angle measures around the FHN junction circumference with very good reliability and reproducibility. This work has the potential to improve analyses of cam-type lesions of the FHN junction for large-scale morphometric and clinical MR investigations of the human hip region.
    Physics in Medicine and Biology 09/2015; 60(19):7601-7616. DOI:10.1088/0031-9155/60/19/7601 · 2.76 Impact Factor
  • P.T. Nguyen · Amin Abbosh · Stuart Crozier
    [Show abstract] [Hide abstract]
    ABSTRACT: A technique for noninvasive microwave hyperthermia treatment for breast cancer is presented. In the proposed technique, electromagnetic focusing on patient-specific breast models is implemented to concentrate the power at the tumor position while keeping the power levels at other positions (healthy tissues) at minimum values. This step is realized by optimizing phase excitations of the utilized antenna elements. In the next step, a thermal analysis is used to determine the scaling factor of the antennas’ excitation amplitudes to realize the required temperature at the tumor position. A closed-loop procedure ensures that there are no hot spots in any of the healthy tissues. The technique is tested in a realistic environment which includes three-dimensional breast models and antenna arrays. The presented results on fatty and dense breast models using two types of directional antennas validate the proposed technique. The investigations across wide frequency bands indicate that the frequencies around 4.2 GHz and 4.5 GHz are optimum values for the hyperthermia treatment of dense and fatty breasts, respectively.
    IEEE Transactions on Antennas and Propagation 08/2015; DOI:10.1109/TAP.2015.2463681 · 2.18 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In some MRI-LINAC (Magnetic Resonance Imaging and Linear Accelerator) hybrid systems, the MRI scanner is split into two parts to form a central gap for the accommodation of the patient or a LINAC. Little is known about the acoustic characteristics of the split gradient coil structure needed for this system; however, it is believed to be very different from its typical configurations. It is important to develop dedicated numerical methods for the characterization of the unique acoustic properties, to provide engineering solutions for the noise attenuation for such a new system. In this article, we modeled the acoustic fields of a split MRI system and traditional gradient structures using the finite element method. The models were validated against acoustic experimental results obtained from a conventional MRI scanner. The acoustic field distribution analysis showed that the average sound pressure levels in the central gap were lower than those in the cylindrical tunnels of the split MRI system at most frequencies. This was also true when both the x coils or z coils were energized independently. Thus, if the patient bed is placed perpendicular to the axis of the main magnet of the split MRI system, the patient will be subjected to relatively lower acoustic intensities compared with that if the patient bed is placed parallel to the axis of the main magnet. Further work is planned to reduce the sound level in the central gap where the patient bed may be placed for this split system. © 2015 Wiley Periodicals, Inc. Concepts Magn Reson Part B (Magn Reson Engineering), 2015
    Concepts in Magnetic Resonance Part B Magnetic Resonance Engineering 06/2015; DOI:10.1002/cmr.b.21283 · 0.69 Impact Factor
  • Xia Kong · Minhua Zhu · Ling Xia · Stuart Crozier · Qiuliang Wang · Zhipeng Ni · Feng Liu
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents a novel passive shimming method for the effective correction of static magnetic field (B0) inhomogeneities in Magnetic Resonance Imaging (MRI) systems. Passive shimming is used to find an optimum configuration for the placement of iron pieces applied to improve the B0 uniformity in the predefined imaging region referred to as the diameter of spherical volume (DSV). However, most passive shimming methods neglect to recognize that the space under the patient bed is not in use for imaging. In this work, we present a new algorithm that attempts to avoid the unnecessary shimming of the space under the patient bed. During implementation, the B0 field is still measured over the DSV surface and then mapped onto the effective imaging volume surface; a dedicated sensitivity matrix is generated only for the imaging area above the patient bed. A linear programming optimization procedure is performed for the determination of thicknesses and locations the shim pieces. Our experimental results showed that by revising the shimming target area, the new method provides superior optimization solutions. Compared to a conventional approach, the new method requires smaller amount of iron to correct the B0 inhomogeneities in the imaging area which has the effect of improving thermal stability to the B0 field. It also reduces the complexity of the optimization problem. Our new shimming strategy helps to improve the magnetic field homogeneity within the realistic imaging space, and ultimately improve image quality. Copyright © 2015 Elsevier Inc. All rights reserved.
    Journal of Magnetic Resonance 05/2015; 257. DOI:10.1016/j.jmr.2015.05.004 · 2.51 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Many medical image processing techniques rely on accurate shape modeling of anatomical features. The presence of shape abnormalities challenges traditional processing algorithms based on strong morphological priors. In this work, a sparse shape reconstruction from a statistical shape model is presented. It combines the advantages of traditional statistical shape models (defining a 'normal' shape space) and previously presented sparse shape composition (providing localized descriptors of anomalies). The algorithm was incorporated into our image segmentation and classification software. Evaluation was performed on simulated and clinical MRI data from 22 sciatica patients with intervertebral disc herniation, containing 35 herniated and 97 normal discs. Moderate to high correlation (R=0.73) was achieved between simulated and detected herniations. The sparse reconstruction provided novel quantitative features describing the herniation morphology and MRI signal appearance in three dimensions (3D). The proposed descriptors of local disc morphology resulted to the 3D segmentation accuracy of 1.07±1.00mm (mean absolute vertex-to-vertex mesh distance over the posterior disc region), and improved the intervertebral disc classification from 0.888 to 0.931 (area under receiver operating curve). The results show that the sparse shape reconstruction may improve computer-aided diagnosis of pathological conditions presenting local morphological alterations, as seen in intervertebral disc herniation. Crown Copyright © 2015. Published by Elsevier Ltd. All rights reserved.
    Computerized Medical Imaging and Graphics 05/2015; DOI:10.1016/j.compmedimag.2015.05.002 · 1.22 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To simulate the delivery of proton beams to the treatment zone inside a split-bore MRI-guided proton therapy system. Field maps from a split-bore 1 T MRI system are used as input to Monte Carlo simulations which model the trajectory of proton beams towards isocentre. Both inline (along the MRI bore) and perpendicular (through the split-bore gap) orientations are simulated. Monoenergetic diverging beams of energy 90 MeV, 195 MeV and 300 MeV starting from 1.5 m above isocentre were modelled. A phase space file detailing a 2D calibration pattern is used to set the particle starting positions, and their spatial location as they cross isocentre recorded. Inline Orientation: The radial symmetry of the solenoidal style fringe field acts to rotate the protons around the beam's central axis. For protons starting at 1.5 m from isocentre this rotation is 17° (90 MeV), and 8° (300 MeV). Perpendicular Orientation: Isocentre shifts of 135 mm (90 MeV) and 65 mm (300 MeV) were observed in the direction perpendicular to the main imaging field. Off-axis protons are also slightly deflected towards or away from the central axis in the direction perpendicular to the main deflection direction. This leads to a distortion of the phase space pattern, not just a shift. The distortion increases from zero at the central axis to 10 mm (90 MeV) and 5 mm (300 MeV) for a proton 150 mm off-axis. The complexity and energy-dependent nature of the magnetic deflection and distortion indicates the pencil beam scanning method will be the only choice for delivering a therapeutic proton beam inside a potential MRI-guided proton therapy system. Significant correction strategies will be required to account for the MRI fringe fields. The authors acknowledge funding from NHMRC Program Grant No. 1036078 and ARC Discovery Grant No. DP120100821.
    Medical Physics 05/2015; 42(5):2113-2124. DOI:10.1118/1.4916661 · 2.64 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The paper presents a computational study for the estimation of the temperature elevation occurring in a human subject carrying metallic hip prostheses when exposed to the magnetic field produced by gradient coils. The simulations are performed through validated numerical codes, which solve the electromagnetic and thermal equations applied to a high-resolution anatomical human model. Three different sets of gradient coils (traditional, split and uniplanar) are considered to evaluate the maximum steady-state temperature elevation in the human body. This result is then rescaled to take into account the waveform of the signal, the duty-cycle and the duration of the scan. Several exposure situations obtained by changing the patient's position are analyzed, finding temperature elevations on the order of some degrees. The results are of possible concern and provide evidence of the need for further specific investigations aimed at assuring the safety of potential patients carrying metallic hip implants. Magn Reson Med, 2015. © 2015 Wiley Periodicals, Inc. © 2015 Wiley Periodicals, Inc.
    Magnetic Resonance in Medicine 04/2015; 74(1). DOI:10.1002/mrm.25687 · 3.57 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: IntroductionAccurate knowledge of O6-methylguanine methyltransferase (MGMT) gene promoter subtype in patients with glioblastoma (GBM) is important for treatment. However, this test is not always available. Pre-operative diffusion MRI (dMRI) can be used to probe tumour biology using the apparent diffusion coefficient (ADC); however, its ability to act as a surrogate to predict MGMT status has shown mixed results. We investigated whether this was due to variations in the method used to analyse ADC.Methods We undertook a retrospective study of 32 patients with GBM who had MGMT status measured. Matching pre-operative MRI data were used to calculate the ADC within contrast enhancing regions of tumour. The relationship between ADC and MGMT was examined using two published ADC methods.ResultsA strong trend between a measure of ‘minimum ADC’ and methylation status was seen. An elevated minimum ADC was more likely in the methylated compared to the unmethylated MGMT group (U = 56, P = 0.0561). In contrast, utilising a two-mixture model histogram approach, a significant reduction in mean measure of the ‘low ADC’ component within the histogram was associated with an MGMT promoter methylation subtype (P < 0.0246).Conclusion This study shows that within the same patient cohort, the method selected to analyse ADC measures has a significant bearing on the use of that metric as a surrogate marker of MGMT status. Thus for dMRI data to be clinically useful, consistent methods of data analysis need to be established prior to establishing any relationship with genetic or epigenetic profiling.
    04/2015; 62(2). DOI:10.1002/jmrs.103
  • [Show abstract] [Hide abstract]
    ABSTRACT: Recent studies have consistently shown that amongst staff working with MRI, transient symptoms directly attributable to the MRI system including dizziness, nausea, tinnitus, and concentration problems are reported. This study assessed symptom prevalence and incidence in radiographers and other staff working with MRI in healthcare in the UK. One hundred and four volunteer staff from eight sites completed a questionnaire and kept a diary to obtain information on subjective symptoms and work practices, and wore a magnetic field dosimeter during one to three randomly selected working days. Incidence of MRI-related symptoms was obtained for all shifts and prevalence of MRI-related and reference symptoms was associated to explanatory factors using ordinal regression. Incident symptoms related to working with MRI were reported in 4 % of shifts. Prevalence of MRI-related, but not reference symptoms were associated with number of hours per week working with MRI, shift length, and stress, but not with magnetic field strength (1.5 and 3 T) or measured magnetic field exposure. Reporting of prevalent symptoms was associated with longer duration of working in MRI departments, but not with measured field strength of exposure. Other factors related to organisation and stress seem to contribute to increased reporting of MRI-related symptoms. • Routine work with MRI is associated with increased reporting of transient symptoms • No link to the strength of the magnetic field was demonstrated. • Organisational factors and stress additionally contribute to reporting of MRI-related symptoms.
    European Radiology 03/2015; 25(9). DOI:10.1007/s00330-015-3629-z · 4.01 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: PurposeTo validate a fully automated scheme to extract biochemical information from the hip joint cartilages using MR T2 mapping images incorporating segmentation of co-registered three-dimensional Fast-Spin-Echo (3D-SPACE) images.Methods Manual analyses of unilateral hip (3 Tesla) MR images of 24 asymptomatic volunteers were used to validate a 3D deformable model method for automated cartilage segmentation of SPACE scans, partitioning of the individual femoral and acetabular cartilage plates into clinically defined sub-regions and propagating these results to T2 maps to calculate region-wise T2 value statistics. Analyses were completed on a desktop computer (∼10 min per case).ResultsThe mean voxel overlap between automated A and manual M segmentations of the cartilage volumes in the (clinically based) SPACE images was 73% . The automated and manual analyses demonstrated a relative difference error <10% in the median “T2 average signal” for each cartilage plate. The automated and manual analyses showed consistent patterns between significant differences in T2 data across the hip cartilage sub-regions.Conclusion The good agreement between the manual and automatic analyses of T2 values indicates the use of structural 3D-SPACE MR images with the proposed method provides a promising approach for automated quantitative T2 assessment of hip joint cartilages. Magn Reson Med, 2015. © 2015 Wiley Periodicals, Inc.
    Magnetic Resonance in Medicine 01/2015; DOI:10.1002/mrm.25598 · 3.57 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: We present a statistical shape model approach for automated segmentation of the proximal humerus and scapula with subsequent bone-cartilage interface (BCI) extraction from 3D magnetic resonance (MR) images of the shoulder region. Manual and automated bone segmentations from shoulder MR examinations from 25 healthy subjects acquired using steady-state free precession sequences were compared with the Dice similarity coefficient (DSC). The mean DSC scores between the manual and automated segmentations of the humerus and scapula bone volumes surrounding the BCI region were 0.926 ± 0.050 and 0.837 ± 0.059, respectively. The mean DSC values obtained for BCI extraction were 0.806 ± 0.133 for the humerus and 0.795 ± 0.117 for the scapula. The current model-based approach successfully provided automated bone segmentation and BCI extraction from MR images of the shoulder. In future work, this framework appears to provide a promising avenue for automated segmentation and quantitative analysis of cartilage in the glenohumeral joint.
    Physics in Medicine and Biology 01/2015; 60(4):1441-1459. DOI:10.1088/0031-9155/60/4/1441 · 2.76 Impact Factor
  • IEEE Transactions on Magnetics 01/2015; DOI:10.1109/TMAG.2015.2485778 · 1.39 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Using a new rotating SENSitivity Encoding (rotating-SENSE) algorithm, we have successfully demonstrated that the rotating radiofrequency coil array (RRFCA) is capable of achieving a significant reduction in scan time and a uniform image reconstruction for a homogeneous phantom at 7 Tesla. However, at 7 Tesla the in vivo sensitivity profiles ( ) become distinct at various angular positions. Therefore, sensitivity at other angular positions cannot be obtained by numerically rotating the acquired sensitivity. In this work, a novel sensitivity estimation method for the RRFCA was developed and validated with in vivo human brain imaging. The method employed a library database and registration techniques to estimate coil sensitivity at an arbitrary angular position. The estimated sensitivity maps were then compared to the acquired sensitivity maps. The results indicate that the proposed method is capable of accurately estimating both the magnitude and phase of the sensitivity maps at an arbitrary angular position, which enables us to employ the rotating-SENSE method to perform acceleration and image reconstruction. Compared to a stationary coil array with the same number of coil elements, the RRFCA was able to reconstruct good quality images at a high reduction factor. It is hoped that the proposed sensitivity estimation algorithm and the acceleration ability of the RRFCA will be particularly useful for ultra high field MRI.
    Journal of Magnetic Resonance 12/2014; 252C. DOI:10.1016/j.jmr.2014.12.004 · 2.51 Impact Factor
  • Jin Jin · Feng Liu · Stuart Crozier
    [Show abstract] [Hide abstract]
    ABSTRACT: It is generally a challenging task to reconstruct dynamic magnetic resonance (MR) images with high spatial and high temporal resolutions, especially with highly incomplete k-space sampling. In this work, a novel method that combines a non-rigid image registration technique with sparsity-constrained image reconstruction is introduced. Employing a multi-resolution free-form deformation technique with B-spline interpolations, the non-rigid image registration accurately models the complex deformations of the physiological dynamics, and provides artefact-suppressed high spatial-resolution predictions. Based on these prediction images, the sparsity-constrained data fidelity-enforced image reconstruction further improves the reconstruction accuracy. When compared with the k-t FOCUSS with motion estimation/motion compensation (MEMC) technique on volunteer scans, the proposed method consistently outperforms in both the spatial and the temporal accuracy with variously accelerated k-space sampling. High fidelity reconstructions for dynamic systolic phases with reduction factor of 10 and cardiac perfusion series with reduction factor of 3 are presented.
    Magnetic Resonance Imaging 12/2014; 32(10). DOI:10.1016/j.mri.2014.08.006 · 2.09 Impact Factor
  • JB Golding · V Sarafis · S Crozier · SE Rose
    HortScience: a publication of the American Society for Horticultural Science 11/2014; 32:112-113. · 0.90 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Accurate segmentation of hip joint cartilage from magnetic resonance (MR) images offers opportunities for quantitative investigations of pathoanatomical conditions such as osteoarthritis. In this paper, we present a fully automatic scheme for the segmentation of the individual femoral and acetabular cartilage plates in the human hip joint from high-resolution 3D MR images. The developed scheme uses an improved optimal multi-object multi-surface graph search framework with an arc-weighted graph representation that incorporates prior morphological knowledge as a basis for segmentation of the individual femoral and acetabular cartilage plates despite weak or incomplete boundary interfaces. This automated scheme was validated against manual segmentations from 3D true fast imaging with steady-state precession (TrueFISP) MR examinations of the right hip joints in 52 asymptomatic volunteers. Compared with expert manual segmentations of the combined, femoral and acetabular cartilage volumes, the automatic scheme obtained mean (± standard deviation) Dice's similarity coefficients of 0.81 (± 0.03), 0.79 (± 0.03) and 0.72 (± 0.05). The corresponding mean absolute volume difference errors were 8.44% (± 6.36), 9.44% (± 7.19) and 9.05% (± 8.02). The mean absolute differences between manual and automated measures of cartilage thickness for femoral and acetabular cartilage plates were 0.13 mm (± 0.12) and 0.11 mm (± 0.11), respectively.
    Physics in Medicine and Biology 11/2014; 59(23):7245-7266. DOI:10.1088/0031-9155/59/23/7245 · 2.76 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: This paper aims at evaluating the thermal effects produced by the gradient coils on patients carrying metallic hip prostheses when undergoing check-up in combined magnetic resonance imaging with a linear accelerator systems. The computations are performed by two noncommercial codes specifically developed by the authors for voxel based human models. The electromagnetic field problem is solved by a hybrid finite element-boundary element technique implemented in a massively parallelized GPU system. The temperature elevation due to the electromagnetic exposure is then evaluated through a finite element code. The computations are performed for a large number of situations, considering both radial and axial arrangements of the patient.
    IEEE Transactions on Magnetics 11/2014; 50(11):1-4. DOI:10.1109/TMAG.2014.2323119 · 1.39 Impact Factor
  • Zhen Feng · Feng Liu · Mingfeng Jiang · Stuart Crozier · He Guo · Yuxin Wang
    [Show abstract] [Hide abstract]
    ABSTRACT: l1-SPIRiT is a fast magnetic resonance imaging (MRI) method which combines parallel imaging (PI) with compressed sensing (CS) by performing a joint l1-norm and l2-norm optimization procedure. The original l1-SPIRiT method uses two-dimensional (2D) Wavelet transform to exploit the intra-coil data redundancies and a joint sparsity model to exploit the inter-coil data redundancies. In this work, we propose to stack all the coil images into a three-dimensional (3D) matrix, and then a novel 3D Walsh transform-based sparsity basis is applied to simultaneously reduce the intra-coil and inter-coil data redundancies. Both the 2D Wavelet transform-based and the proposed 3D Walsh transform-based sparsity bases were investigated in the l1-SPIRiT method. The experimental results show that the proposed 3D Walsh transform-based l1-SPIRiT method outperformed the original l1-SPIRiT in terms of image quality and computational efficiency.
    Magnetic Resonance Imaging 09/2014; 32(7). DOI:10.1016/j.mri.2014.04.008 · 2.09 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Objective To validate an automatic scheme for the segmentation and quantitative analysis of the medial (MM) and lateral meniscus (LM) in magnetic resonance (MR) images of the knee. Method We analysed sagittal water-excited dual-echo steady-state MR images of the knee from a subset of the Osteoarthritis Initiative cohort. The MM and LM were automatically segmented in the MR images based on a deformable model approach. Quantitative parameters including volume, subluxation and tibial-coverage were automatically calculated for comparison (Wilcoxon tests) between knees with variable radiographic osteoarthritis (rOA), medial and lateral joint space narrowing (mJSN, lJSN) and pain. Automatic segmentations and estimated parameters were evaluated for accuracy using manual delineations of the menisci in 88 pathological knee MR examinations at baseline and 12 months time-points. Results The median (95% confidence-interval) Dice similarity index ( 2*|Auto ∩Manual|/(|Auto|+|Manual|)*100 2*|Auto∩ Manual|/(|Auto|+|Manual|)*100) between manual and automated segmentations for the MM and LM were 78.3%(75.0—78.7), 83.9%(82.1—83.9) at baseline and 75.3%(72.8—76.9), 83.0%(81.6—83.5) at 12 months. Pearson coefficients between automatic and manual segmentation parameters ranged from r=0.70 to r=0.92. MM in rOA/mJSN knees had significantly greater subluxation and smaller tibial-coverage than no-rOA/no-mJSN knees. LM in rOA knees had significantly greater volumes and tibial-coverage than no-rOA knees. Conclusion Our automated method successfully segmented the menisci in normal and osteoarthritic knee MR images and detected meaningful morphological differences with respect to rOA and JSN. Our approach will facilitate analyses of the menisci in prospective MR cohorts such as the OAI for investigations into pathophysiological changes occurring in early OA development.
    Osteoarthritis and Cartilage 09/2014; 22(9). DOI:10.1016/j.joca.2014.06.029 · 4.17 Impact Factor

Publication Stats

3k Citations
576.40 Total Impact Points


  • 1991–2015
    • University of Queensland
      • • School of Information Technology and Electrical Engineering
      • • School of Mathematics and Physics
      Brisbane, Queensland, Australia
  • 2012
    • The Florey Institute of Neuroscience and Mental Health
      • Stroke Division
      Melbourne, Victoria, Australia
  • 2010
    • Qingdao University
      Tsingtao, Shandong Sheng, China
  • 2004–2009
    • University of Tasmania
      • School of Mathematics & Physics
      Newnham, Tasmania, Australia
  • 2004–2008
    • Zhejiang University
      • Department of Biomedical Engineering
      Hangzhou, Zhejiang Sheng, China
  • 2005
    • National High Magnetic Field Laboratory
      Tallahassee, Florida, United States
    • Harvard Medical School
      Boston, Massachusetts, United States