Nicholas Rubert

Electrode displacement strain imaging of thermally-ablated liver tissue in an in vivo animal model.

PURPOSE Percutaneous thermal ablation is increasingly being used to destroy hepatic tumors in situ. The success of ablative techniques is highly dependent on adequate ablation zone monitoring, and ultrasound-based strain imaging could become a convenient and cost-effective means to delineate ablation zone boundaries. This study investigates in vivo electrode displacement-based strain imaging for monitoring hepatic ablation procedures that are difficult to perform with conventional elastography. METHODS a In our method, minute displacements (less than a millimeter) are applied to the unconstrained end of the ablation electrode, resulting in localized tissue deformation within the ablation zone that provides the mechanical stimuli required for strain imaging. This article presents electrode displacement strain images of radiofrequency ablation zones created in porcine liver in vivo (n = 13). RESULTS Cross-sectional area measurements from strain images of these ablation zones were obtained using manual and automated segmentation. Area measurements from strain images were highly correlated with areas measured on histopathology images, quantitated using linear regression (R = 0.894, P < 0.001 and R = 0.828, P < 0.001, respectively). CONCLUSIONS This study further demonstrates that electrode displacement elastography is capable of providing high-contrast images using widely available commercial ultrasound systems which may potentially be used to assess the extent of thermal ablation zones.

Bayesian Regularization Applied to Ultrasound Strain Imaging

Noise artifacts due to signal decorrelation and reverberation are a considerable problem in ultrasound strain imaging. For block-matching methods, information from neighboring matching blocks has been utilized to regularize the estimated displacements. We apply a recursive Bayesian regularization algorithm developed by Hayton et al. [Artif. Intell., vol. 114, pp. 125-156, 1999] to phase-sensitive ultrasound RF signals to improve displacement estimation. The parameter of regularization is reformulated, and its meaning examined in the context of strain imaging. Tissue-mimicking experimental phantoms and RF data incorporating finite-element models for the tissue deformation and frequency-domain ultrasound simulations are used to compute the optimal parameter with respect to nominal strain and algorithmic iterations. The optimal strain regularization parameter was found to be twice the nominal strain and did not vary significantly with algorithmic iterations. The technique demonstrates superior performance over median filtering in noise reduction at strains 5% and higher for all quantitative experiments performed. For example, the strain SNR was 11 dB higher than that obtained using a median filter at 7% strain. It has to be noted that for applied deformations lower than 1%, since signal decorrelation errors are minimal, using this approach may degrade the displacement image.

Ultrasound-based relative elastic modulus imaging for visualizing thermal ablation zones in a porcine model

The feasibility of using ultrasound-based elastic modulus imaging to visualize thermal ablation zones in an in vivo porcine model is reported. Elastic modulus images of soft tissues are estimated as an inverse optimization problem. Ultrasonically measured displacement data are utilized as inputs to determine an elastic modulus distribution that provides the best match to this displacement field. A total of 14 in vivo thermal ablation zones were investigated in this study. To determine the accuracy of delineation of each thermal ablation zone using elastic modulus imaging, the dimensions (lengths of long and short axes) and the area of each thermal ablation zone obtained from an elastic modulus image were compared to the corresponding gross pathology photograph of the same ablation zone. Comparison of elastic modulus imaging measurements and gross pathology measurements showed high correlation with respect to the area of thermal ablation zones (Pearson coefficient = 0.950 and p < 0.0001). The radiological-pathological correlation was slightly lower (correlation = 0.853, p < 0.0001) for strain imaging among these 14 in vivo ablation zones. We also found that, on average, elastic modulus imaging can more accurately depict thermal ablation zones, when compared to strain imaging (14.7% versus 22.3% absolute percent error in area measurements, respectively). Furthermore, elastic modulus imaging also provides higher (more than a factor of 2) contrast-to-noise ratios for evaluating these thermal ablation zones than those on corresponding strain images, thereby reducing inter-observer variability. Our preliminary results suggest that elastic modulus imaging might potentially enhance the ability to visualize thermal ablation zones, thereby improving assessment of ablative therapies.

Ultrasound Attenuation Measurements Using a Reference Phantom with Sound Speed Mismatch

Ultrasonic attenuation may be measured accurately with clinical systems and array transducers by using reference phantom methods (RPM) to account for diffraction and other system dependencies on echo signals. Assumptions with the RPM are that the speeds of sound in the sample (csam ) and in the reference medium (cref ) are the same and that they match the speed assumed in the system beamformer (cbf ). This work assesses the accuracy of attenuation measurements by the RPM when these assumptions are not met. Attenuation was measured for two homogeneous phantoms, one with a speed of sound of 1500 m/s and the other with a sound speed of 1580 m/s. Both have an attenuation coefficient approximately equal to that of the reference, in which the speed of sound is 1540 m/s. Echo signals from the samples and the reference were acquired from a Siemens S2000 scanner with a 9L4 linear array transducer. Separate acquisitions were obtained with cbf at its default value of 1540 m/s and when it was set at values matching the speeds of sound of the phantoms. Simulations were also performed using conditions matching those of the experiment. RPM-measured attenuation coefficients exhibited spatially-dependent biases when csam differed from cbf and cref Mean errors of 19% were seen for simulated data, with the maximum errors in attenuation measurements occurring for regions of interest near the transmit focus. Biases were minimized (mean error with simulated data was 5.6%) using cbf that matched csam and assuring that power spectra used for attenuation computations in the sample are from precisely the same depth as those from the reference. Setting the transmit focus well beyond the depth range used to compute attenuation values minimized the bias.

Mean scatterer spacing estimation using multi-taper coherence

It has been hypothesized that estimates of mean scatterer spacing are useful indicators for pathological changes to the liver. A commonly employed estimator of the mean scatterer spacing is the location of the maximum of the collapsed average of coherence of the ultrasound radio-frequency signal. To date, in ultrasound, estimators for this quantity have been calculated with a single taper. Using frequency-domain Monte Carlo simulations, we demonstrate that multi-taper estimates of coherence are superior to single-taper estimates for predicting mean scatterer spacing. Scattering distributions were modeled with Gamma-distributed scatterers for fractional standard deviations in scatterer spacings of 5, 10, and 15% at a mean scatterer spacing of 1 mm. Additionally, we demonstrate that we can distinguish between ablated liver tissue and unablated liver tissue based on signal coherence. We find that, on the average, signal coherence is elevated in the liver relative to signal coherence of received echoes from thermally ablated tissue. Additionally, our analysis indicates that a tissue classifier utilizing the multi-taper estimate of coherence has the potential to distinguish between ablated and unablated tissue types better than a single-taper estimate of coherence. For a gate length of 5 mm, we achieved an error rate of only 8.7% when sorting 23 ablated and 23 unablated regions of interest (ROIs) into classes based on multi-taper calculations of coherence.

CT protocol management: simplifying the process by using a master protocol concept

This article explains a method for creating CT protocols for a wide range of patient body sizes and clinical indications, using detailed tube current information from a small set of commonly used protocols. Analytical expressions were created relating CT technical acquisition parameters which can be used to create new CT protocols on a given scanner or customize protocols from one scanner to another. Plots of mA as a function of patient size for specific anatomical regions were generated and used to identify the tube output needs for patients as a function of size for a single master protocol. Tube output data were obtained from the DICOM header of clinical images from our PACS and patient size was measured from CT localizer radiographs under IRB approval. This master protocol was then used to create 11 additional master protocols. The 12 master protocols were further combined to create 39 single and multiphase clinical protocols. Radiologist acceptance rate of exams scanned using the clinical protocols was monitored for 12,857 patients to analyze the effectiveness of the presented protocol management methods using a two‐tailed Fishers exact test. A single routine adult abdominal protocol was used as the master protocol to create 11 additional master abdominal protocols of varying dose and beam energy. Situations in which the maximum tube current would have been exceeded are presented, and the trade‐offs between increasing the effective tube output via 1) decreasing pitch, 2) increasing the scan time, or 3) increasing the kV are discussed. Out of 12 master protocols customized across three different scanners, only one had a statistically significant acceptance rate that differed from the scanner it was customized from. The difference, however, was only 1% and was judged to be negligible. All other master protocols differed in acceptance rate insignificantly between scanners. The methodology described in this paper allows a small set of master protocols to be adapted among different clinical indications on a single scanner and among different CT scanners. PACS number: 87.57.QThis article explains a method for creating CT protocols for a wide range of patient body sizes and clinical indications, using detailed tube current information from a small set of commonly used protocols. Analytical expressions were created relating CT technical acquisition parameters which can be used to create new CT protocols on a given scanner or customize protocols from one scanner to another. Plots of mA as a function of patient size for specific anatomical regions were generated and used to identify the tube output needs for patients as a function of size for a single master protocol. Tube output data were obtained from the DICOM header of clinical images from our PACS and patient size was measured from CT localizer radiographs under IRB approval. This master protocol was then used to create 11 additional master protocols. The 12 master protocols were further combined to create 39 single and multiphase clinical protocols. Radiologist acceptance rate of exams scanned using the clinical protocols was monitored for 12,857 patients to analyze the effectiveness of the presented protocol management methods using a two-tailed Fishers exact test. A single routine adult abdominal protocol was used as the master protocol to create 11 additional master abdominal protocols of varying dose and beam energy. Situations in which the maximum tube current would have been exceeded are presented, and the trade-offs between increasing the effective tube output via 1) decreasing pitch, 2) increasing the scan time, or 3) increasing the kV are discussed. Out of 12 master protocols customized across three different scanners, only one had a statistically significant acceptance rate that differed from the scanner it was customized from. The difference, however, was only 1% and was judged to be negligible. All other master protocols differed in acceptance rate insignificantly between scanners. The methodology described in this paper allows a small set of master protocols to be adapted among different clinical indications on a single scanner and among different CT scanners. PACS number: 87.57.Q.

Monitoring microwave ablation for liver tumors with electrode displacement strain imaging

Minimally invasive ablative therapies have become important alternatives to surgical treatment of both hepatocellular carcinoma (HCC) and liver metastases. Image based guidance and monitoring are therefore essential. Although ultrasound (US) imaging suffers from inadequate echogenic contrast between ablated and normal tissue, US based elasticity imaging has shown remarkable ability to depict ablated regions and delineate margins. The purpose of this study is to apply “electrode displacement elastography,” or EDE for monitoring clinical microwave ablation (MWA) treatments for HCC and liver metastases. EDE images were acquired from 10 patients who underwent MWA for their liver tumors. The MWA system used was a Neuwave Medical Certus 140 (Madison, WI, USA) operating at 2.45 GHz. The MWA power and duration was adjusted for each patient, with typical values of 65 watts and 5 minutes. A Siemens S2000 scanner equipped with a curvilinear array transducer (VFX 6C1) pulsed at 4 MHz was used to acquire radiofrequency echo data. Electrode displacement was applied manually by the physician. A multi-seed two-dimensional tracking algorithm, with kernel dimensions of 0.096 mm × 3 A-lines was used to estimate local displacements between consecutive data frames. Strain images were computed as the gradient of the local displacement estimates. The average contrast of the ablated region was 0.23±0.07 (0.14-0.35) on B-mode images and 0.73±0.08 (0.56-0.82) on EDE. The average contrast improvement with EDE over B mode was about 230%. The average tumor size was 2.2±0.8 (0.7-3.5) cm on pre-treatment diagnostic images (CT or MRI). The average size of the ablated region was 3.8±0.7 (2.6-4.9) cm on EDE, with an average ablation margin of 1.6 cm which is within the clinically suggested ablated margin (>0.5 cm).

Mean Scatterer Spacing Estimation in Normal and Thermally Coagulated Ex Vivo Bovine Liver

The liver has been hypothesized to have a unique arrangement of microvasculature that presents as an arrangement of quasiperiodic scatterers to an interrogating ultrasound pulse. The mean scatterer spacing (MSS) of these quasiperiodic scatterers has been proposed as a useful quantitative ultrasound biomarker for characterizing liver tissue. Thermal ablation is an increasingly popular method for treating hepatic tumors, and ultrasonic imaging approaches for delineating the extent of thermal ablation are in high demand. In this work, we examine the distribution of estimated MSS in thermally coagulated bovine liver and normal untreated bovine liver ex vivo. We estimate MSS by detecting local maxima in the spectral coherence function of radio frequency echoes from a clinical transducer, the Siemens VFX 9L4 transducer operating on an S2000 scanner. We find that normal untreated bovine liver was characterized by an MSS of approximately 1.3 mm. We examined regions of interest 12 mm wide laterally, and ranging from 12 mm to 18 mm axially, in 2 mm increments. Over these parameters, the mode of the MSS estimates was between 1.25 and 1.37 mm. On the other hand, estimation of MSS in thermally coagulated liver tissue yields a distribution of MSS estimates whose mode varied between 0.45 and 1.0 mm when examining regions of interest over the same sizes. We demonstrate that the estimated MSS in thermally coagulated liver favors small spacings because the randomly positioned scatterers in this tissue are better modeled as aperiodic scatterers. The submillimeter spacings result from the fact that this was the most probable spacing to be estimated if the discretely sampled spectral coherence function was a uniformly random two-dimensional function.

In vivo ultrasound electrode displacement strain imaging

The incidence of primary and secondary liver tumors has increased significantly over the last two decades. Due to poor liver function and/or distribution of lesions, surgery is often not an option for affected patients. An alternative to surgical resection, percutaneous radiofrequency ablation is often used to thermally destroy the tumor in situ. Radiofrequency ablation is evolving into one of the more popular minimally-invasive treatments for hepatic tumors. Monitoring the treated region is an important factor in the success of radiofrequency ablation, and ultrasound elastography could become a convenient, cost-effective means to delineate the thermal lesion boundaries for clinical personnel during the procedure. This study assesses “electrode displacement” strain imaging for monitoring abdominal radiofrequency ablation procedures. We present results utilizing a novel approach of displacing the ablation electrode itself to introduce the mechanical stimuli required for strain imaging. Utilizing a Siemens Antares clinical ultrasound scanner equipped with a research interface, ultrasound radiofrequency data were acquired immediately following radiofrequency ablation of porcine liver. The porcine liver was excised following the procedure, and the dimensions of the thermal lesion in the imaging plane measured by slicing though the lesion. Strain images of the lesions were produced offline using axial guidance based block-matching and multi-level motion tracking algorithms. The area of the lesion on the strain image was compared to the area of the lesion in pathology images obtained from eight separate lesions. The estimated linear correlation coefficient between the pathology image and the strain image was r = 0.961 (p ≪ .001) for manual segmentation using 4 observers. The area of the lesion in the strain image slightly underestimates the area of the lesion in the pathology image for all slices, agreeing with earlier ex-vivo experiments.

A Wiki-Based Solution to Managing Your Institution's Imaging Protocols

At our institution, we have roughly 300 different CT protocols in routine clinical use. Of these 300 protocols, we have 149 distinct clinical indications that are not broken up into different patient sizes [1]. A single-phase examination with three reconstructions has roughly 50 editable fields on the scanner. These parameters need to be maintained across 12 scanner models, which translates to 180,000 (300 50 12) unique scanner parameters needed to define our complete CT protocol set. This is a low estimate because many of our examinations are multiphasic and have more than three reconstructions. This estimate also does not include any clinical parameters (eg, patient preparation, contrast instructions), which are important to maximize the diagnostic utility and workflow of our CT practice. Almost immediately, problems arise when so many protocols and protocol parameters are in clinical use. Maintaining the same protocol name and number identifier for the same indication on all scanners has proved challenging at our institution, requiring a dedicated effort by our lead CT technologist. Furthermore, many dose monitoring software packages use protocol number identifiers. Therefore, managing the