2010 Medical Imaging Seminars


Tuesday, October 26, 4:30pm, in EMS 715.

Evaluation of the Accuracy and Robustness of a Motion Correction Algorithm for 4D PET Using a Novel Phantom Measurement Approach

S. D. Wollenweber, Molecular Imaging (GE Healthcare), G. Gopalakrishnan (GE Global Research), A. Roy, R. Manjeshwar (GE Global Research), K. Thielemans (Hammersmith Imanet)

Use of a global, non-rigid image registration algorithm applied to 4D gated PET data has been investigated with phantom data including realistic feature motion. The goal of the study was to assess PET-PET image registration and sum the registered images to form a motion-corrected high-statistics PET image, a procedure we called Reconstruct, Register and Add (RRA). The phantom was constructed with 5 Ge-68 filled spheres suspended in a water-filled tank and driven by a periodic motion. Comparison of feature quantification of SUVmax and volume was made for several imaging protocols. Images from a gated acquisitions were sent to a multi-resolution level-sets non-rigid registration (NRR). The NRR images were then summed to form a motion-corrected, high-statistics image set. Spheres from the images were segmented and compared across imaging conditions. Results: The average range of center-of-mass motion was 7.35, 5.83 and 2.66 mm for the spheres over the three motion periods (8, 6, 4-second) measured. The center-of-mass for all spheres in all conditions was corrected to within 1mm for most cases of the NRR as compared to the gated data. For the RRA data, the sphere maximum activity concentration (MAC) was on average 38.9% higher (range 4.0% to 116.7%) and sphere volume was on average 10.7% smaller (range -8.2% to 28.1%) as compared to the 3-minute static scan that included sphere motion. The RRA results for MAC were on average 70% more accurate and for sphere volume 80% more accurate as compared to the 3-minute static scan with sphere motion, each case as measured against the 3-minute no-motion case. The results show that the RRA technique with the level sets algorithm significantly improves the quantitation and volume estimation as compared to un-gated data in situations where the motion is large as compared to the sphere size. The NRR algorithm was robust to the non-rigid feature motion as well as motion in the vicinity of a non-moving object.


Wednesday October 13, 4:30pm, in Physics 133.

Cardiopulmonary Structure and Function

Robert Molthen, Medical College of Wisconsin, Marquette University and Zablocki VA Medical Center

We use novel imaging, telemetry, and isolated lung perfusion techniques to examine the physiologic context of structure and function relationships that are modified by pulmonary vascular and airway remodeling. Utilizing various small animal models of disease, our objectives help us to develop a more complete understanding of cardiopulmonary etiology. The presentation combines an overview of various imaging and analysis techniques as well as data from rodent studies, including work on pulmonary hypertension and mathematical models and methods for quantifying morphology and biomechanics in the pulmonary vascular tree.


Tuesday, May 11, 4:30pm, in Physics 143.

Radiofrequency Heating of Single Walled Carbon Nanotubes and Gold Nanoparticles

Paul Cherukuri, Rice University, Dept of Chemistry & Smalley Inst. of Nanoscale Science & Technology

Nanoscale metals exhibit unique electrodynamic properties that can be utilized in numerous applications. This talk will focus on RF heating of metallic nanoparticles and discuss their potential as cancer therapeutic agents. Recently, aqueous metallic nanoparticle suspensions (carbon nanotubes, gold nanoparticles) have been shown to absorb shortwave radiofrequency energy (13.56 MHz) and boil water in less than 30 seconds. This curiously strong thermalization of RF energy by metallic nanoparticles far exceeds techniques that use magnetic fields or optical excitation. The heat released under RF absorption has been ascribed to Joule heating of nanoparticles, however a complete physical understanding of the thermalization mechanism remains.


Monday, April 26, 4:30pm, in Physics 481.

New horizons in CT and ultrasound for hyperthermic ablation monitoring

Chris Brace, Asst. Prof UW-Madison, BME & Radiology

Hyperthermic tumor ablation is an increasingly common option for the treatment of many tumors in the liver, lung, kidney and bone, especially for the large percentage of patients for whom surgery is not an option. Ablations are performed using imaging – typically ultrasound or CT – to locate and place a needle-like applicator into the tumor. One placed, energy is applied through the applicator to elevate temperatures inside the tumor to lethal levels.

To date, CT and ultrasound have been unreliable as a means to monitor the growth of ablations intraprocedurally. Bubbles formed during hyperthermic ablations tend to create a diffuse hyperechoic cloud on conventional ultrasound images, preventing adequate delineation of the ablation boundary. On CT, imaging contrast is minimal between the ablation zone and background parenchyma in the liver and kidney without contrast-enhancement. The objective of this presentation will be to outline emerging technologies for monitoring thermal ablation, including CT and ultrasound thermometry, sequential contrast-enhanced CT, and ultrasound elastography. The basic principles of each technique will be discussed and specific examples of their application for monitoring thermal ablations presented.


Thursday, February 25, 4:30pm, in Physics 142.

Full-field acousto-mammography using an acousto-optic sensor

Patrick LaRiviere, U Chicago
Jas Sandhu, Santec Systems

We have been working to develop a wide-field transmission ultrasound approach to breast imaging based on the use of a large-area acousto-optic (AO) sensor. Accompanied by a suitable acoustic source, such a detector could be mounted on a traditional mammography system and provide a mammography-like ultrasound projection image of the compressed breast in registration with the x-ray mammogram. We call the approach acoustography. The hope is that this additional information could improve the sensitivity and specificity of screening mammography.

The AO sensor converts ultrasound directly into a visual image by virtue of the acousto-optic effect of the liquid crystal layer contained in the AO sensor. The image is captured with a digital video camera for processing, analysis, and storage. Geometric resolution analysis suggests that the technique could readily detect tumors of diameter 3 mm using 8.5 MHz ultrasound, with smaller tumors detectable with higher frequency ultrasound, though depth penetration might then become a limiting factor. Preliminary phantom images show high contrast and compare favorably to digital mammograms of the same phantom.