2084420 Real-Time Elastography in Inflammatory Skin Diseases: A Primer

S82

Ultrasound in Medicine and Biology

surgical specimen immediately after resection. Young’s modulus of each region was measured ex vivo under constant pre-load compression. The relation between image variations of EI in vivo and magnitude of stress was investigated as well. This study was approved by the IRB of University of Tsukuba and Tsukuba Medical Center. Results: Young’s moduli were directly measured ex vivo in surgical specimens ranging from 2.60 kPa (fat) to 16.08 kPa (invasive carcinoma) under the weak-stress condition (0.2–0.4 kPa), which corresponds to the manipulation with appropriate pre-load compression in EI acquisition. In this manner, EIs reveal specific features of ‘‘redand green-striped fat’’ and ‘‘blue pectoral muscle’’ on RTE which are easily recognized. The contrast (ratio) of lesion to fat in elasticity ex vivo gradually decreased as the stress applied increased (around 1.0 kPa) on the background of significant non-linearity of the breast tissue. Conclusions: Our results from ex vivo and in vivo assessment indicate that the differences in non-linearity in elasticity under minimal stress conditions are closely related to the variation in EI quality. In terms of the manner of manipulation, pre-load compression is a fundamental and substantial factor in evaluating accurately with elastography. A proposed simple, easily recognized features in EI can be used to ensure the appropriate amount of pre-load compression applied clinically. 2091372 Assessment Of Cutaneous Sclerotic Disorders Using ARFI/SWEI Seung Yun Lee,1 Adela Cardones,2 Kathryn Nightingale1 Mark Palmeri1 1Biomedical Engineering, Duke University, Durham, NC, United States; 2 Dermatology, Duke University Medical Center, Durham, NC, United States Objectives: Acoustic Radiation Force Impulse and Shear Wave Elasticity Imaging methods were applied to the skin to investigate their feasibility in assessing sclerotic skin diseases. The clinical motivation was to develop a non-invasive imaging technology that can be used in the clinical setting for realtime sclerotic skin disease diagnosis and longitudinal monitoring. We present results from an ongoing clinical study recruiting subjects with and without sclerosis, imaging multiple body sites at 3-month intervals. Methods: In an IRB-approved study of 22 patients with normal and sclerotic skin, the stiffness of each imaged skin site was evaluated using two metrics: (1) mean ARFI displacement magnitude, and (2) group shear wave speed inside the region of excitation (ROE). Imaging was performed using a modified Siemens ACUSON S2000TM scanner and a Siemens 14L5 linear array transducer operating at 7.3 MHz, focused at 5.5 mm with an F/1 focal configuration. Harmonic tracking (6,15/12.3 MHz) was used to improve B-mode image quality with reduced clutter and higher resolution using a pulse inversion technique. Shear wave speeds were estimated using Radon sum transformation on time-to-peak-slope data. Results: In a subject with localized sclerosis due to Graft Versus Host Disease (GVHD), mean ARFI displacement magnitude inside sclerotic skin was 66% lower (p,0.1) and shear wave speed 108% higher (p,0.001) than normal skin. Both of these metrics indicate that the sclerotic skin was stiffer than the normal skin. Similarly, in a subject with morphea, we found the mean ARFI displacement magnitude inside sclerotic skin was 19% lower (p,0.01) and shear wave speed 27% higher (p,0.05) than contrlateral normal skin. Across all subjects, we

Volume 41, Number 4S, 2015

were able to successfully estimate ARFI displacements and shear wave speeds in 70.2% of the acquisitions, with failures mostly due to the ROE extending outside the thin dermis. This success rate rose to 94.3% when the ROE was completely contained within the dermis. Conclusions: ARFI and SWEI can differentiate sclerotic from normal skin and quantify the severity of the sclerotic disease with changes in stiffness relative to the contralateral normal skin. 2088979 A New Quality Control Program for Strain Elastography (Full Automatic Strain Ratio) Kazutaka Nakashima1 1General Surgery, Kawasaki Medical School, Okayama, Japan Objectives: Elastography is a most important program for breast ultrasound. There are two category, one is ‘‘Strain Elastography’’, and other is ‘‘Shear Wave Elastography’’. These application are usefull by clinical trials, but someone often say ‘‘Strain Elastography have some intrinsic bias in handling’’. We think the bias is made from the selection of time phase of image to evaluate, and the unbalance of the pressure, ROI setting. To exclude all bias, we made a full automatic strain ratio(FASR) and we show the preliminary results in this presentation. Methods: The program have three important point. After more than 2 second elastography examination, it select the most balanced phase on pressure in all phase of examination automatically. And the user set a point mark on the any area in target mass to know the rigidity, the target circle (to be calculate strain) gets wide enough within the target. And the fat area circle is made in subcutaneous fat tissue to compare the target strain, then FASR is indicate on the monitor automatically. To evaluate the system, we have two step preliminary study. At first, we made a phantom some mass target put into the breast gland like material by ultrasound. The target was calculated its rigidity by industrial machine before. And we use the full automatic strain ratio program to these target and evaluate the strain ratio and the calculated rigidity on each target. Next, we use the program stored by the strain movie data in usual elastography examination to compare the program with the usual strain ratio gotten by skilled doctor for preliminary clinical study. Results: On a phantom study, there is a high correlation between strain ratio and the target rigidity. We think the program will be able to calculate the target rigidity with high quality. At clinical date study, the three is correlative not only the mass target ROI, but also the fat ROI. And we think the program is useful, instead of manual process. Conclusions: We think the easier way made a higher quality control, because of the bias will decrease. So we want to say that a FASR has a high quality, because the most way is automatic. The strain elastography will be a new high quality application by FASR. 2084420 Real-Time Elastography in Inflammatory Skin Diseases: A Primer Fernando Alfageme,1 Eugenio Cerezo,2 1 1 Gaston Roustan Dermatology, Hospital Universitario Puerta de Hierro, Majadahonda, Madrid, Spain; 2Ecografias America, Madridi, Spain

Abstracts

Objectives: Elastography is a non-invasive ultrasonographical method to assess tissue stiffness widely used in the evaluation of inflammation of several organs (liver, thyroid, tendons and muscle). Scarce reports are available about the application of this technique to assess inflammatory skin diseases in comparation with normal skin and B mode images. Methods: We reviewed our experience in the evaluation of several inflammatory skin diseases (psoriasis, subacute and acute lupus, eccema, warts, foreign body reactions to fillers) with real time elastography (RTE) in two systems different equipments and elastography software ( Esaote Class C 13 18 lineal probe ElaxtoÓ , General Electric S8 6-13 MHz lineal probe, Elasto Ó ) B mode scans of diseased and clinical healthy adjacent skin were acquired. Inflammatory areas were defined as hypoechoic dermal subdermal areas in comparation with clinical healthy areas. Simultaneously qualitative soft -hard images were acquired in both diseased and clinical adjacent healthy skin. Correlation of inflammation in B mode scans with soft-hard images of these areas was stablished as coincident or extended. Results: Inflamed skin areas showed a decrease in dermal-subdermal stiffness in comparation with normal adjacent areas. Inflammatory skin areas tended to correlate with soft areas and extend these limits even in absence of dermal hypoechogenicity. Conclusions: Inflammatory skin diseases, in our experience, tend to show decreased stiffness compared with clinical healthy skin. The extent of softness areas is greater than inflamed areas what may indicate presence of inflammatory areas not detected by B Mode scans or changes in stiffness due to inflammation. 2075089 The Microchannel Flow Model for Soft Tissues Kevin Parker1 1Electrical & Computer Engineering, University of Rochester, Rochester, NY, United States Objectives: A number of advances in elastography have increased our ability to make measurements of viscoelastic properties of tissues. Accordingly, the question of an appropriate viscoelastic model becomes paramount. This study examines normal soft tissues and develops a model of behavior based on the flow of viscous fluids through the extensive network of tissue microchannels in response to applied stress. This behavior can be captured in a 2-parameter model, and the model appears to predict the stress-relaxation behavior and the dispersive shear wave behavior of bovine liver specimens and other soft tissues and phantoms over a frequency range of 60 – 400 Hz. The relationship of the microchannel flow model (MFM) to more traditional models is also examined. Methods: We consider a block of liver tissue, comprised of a fine-scale interlocking of hepatic cells, connective tissue, and a variety of fluid channels including billary, capillary, and lymphatic. We then consider a small fluid microchannel as a fluid-filled inclusion and derive a characteristic time constant for stress relaxation. Assuming there are multiple microchannels of unequal radius rn and therefore unequal flow rates Qn, and that each contributes to the stress relaxation at their respective time constant tn, then the simplest model for this looks like a generalized Maxwell-Weichert model. Next, a relaxation spectrum is defined by a power law, linked to the fractal branching behavior of vessels. Combining these yields the constitutive equations for soft tissue.

S83

Results: Time domain measurements of stress relaxation in liver tissue speciments, and shear wave measurements between 60 – 400 Hz, demonstrate the accuracy of the MFM and the relevance of its parameters. The key power law parameter ‘‘a’’ is in the range of 0.15 – 0.25 for normal liver and can indicate the branching microvasculature and microvessel distribution. Conclusions: The MFM is capable of describing the viscoelastic properties of soft tissues such as the liver, and predicts a power law form of dispersion of shear waves. 2080299 Local Lung Ventilation Estimation Using Ultrasound Strain Measurements Jonathan Rubin,1 Jeffrey Horowitz,2 2 3 Thomas Sisson, Kang Kim, James Hamilton4 1 Radiology, Univ Michigan, Ann Arbor, MI, United States; 2Internal Medicine, University of Michigan, Ann Arbor, MI, United States; 3Bioengineering, University of PIttsburgh, Pittsburgh, PA, United States; 4Epsilon Imaging, Inc., Ann Arbor, MI, United States Objectives: To estimate local lung ventilation using strain measurements from the moving lung surface during respiration. Methods: The surfaces of the right lungs in 2 human volunteers (IRB approved study) and both lungs in 6 mice, 2 controls and 4 with pulmonary fibrosis induced by either transtracheal bleomycin injection or intraperitoneal diphtheria toxin (DTR:DT) injection in transgenetic mice expressing the diphtheria toxin receptor (DTR)on their type II alveolar epithelial cells. (One additional bleomycin mouse died before the second lung was scanned.) Mice were scanned with a Vevo 2100 (FujiFilm Visualsonics, Toronto, CN) using a 55 MHz linear array and a 6 MHz linear array (Epsilon Imaging, Ann Arbor, MI) was used in humans. Real-time radiofrequency signals were analyzed with speckle tracking software (EchoInsight, Epsilon Imaging, Ann Arbor, MI). Mean peak-to-trough strains on longitudinal scans for each lung were calculated and pooled by group. A hydroxyproline assay was used to assess pulmonary fibrosis. A two-tailed t-test was used for comparison; p , 0.05 was considered significant. Results: Mouse results are shown in the Table. Strains in controls and treated mice were significantly different. Hydroxyproline differences were also significant or nearly significant. Human lung strains were consistent with the literature. Conclusions: Local strain estimates can be made in real-time using ultrasound speckle tracking of the moving lung surface in humans and normal and abnormal mice. Ultrasound may be able to assess local ventilation. Mouse Lung Strain and Hydroxyproline Concentration Estimates Longitudinal Strain Estimates (%) control

DTR:DT

mean 17.4 7.3 standard deviation 2.4 2.9 Hydroxyproline Concentration (micrograms/ml) mean 86.1 158.7 standard deviation 18.0 6.4

bleomycin 8.8 2.0 134.8 9.5

control strain vs DTR:DT, p,0.0006, control strain vs. bleomycin, p,0.004;