Home

Compressive Coded Absorber for Single Photon Emission Computed Tomography

Wisdom Ibole - Jakub Waligorski - Daniel Westerman - Sean Woodfield

Sponsored by: Dr. Sebastian Obrzut

Spring 2017 MAE 156B Sponsored Project

University of California, San Diego

An estimated 12 million patients undergo Single Photon Emission Tomography (SPECT) imaging annually in the US on over 14,000 SPECT cameras (Figure 1). There is no theoretical limit on image resolution in SPECT and approximately thirty SPECT radiotracers routinely used clinically are reimbursed by insurance for diagnostic imaging. In addition, two radiotracers can be simultaneously imaged at different energy levels. This makes the use of SPECT cameras very attractive to industry professionals. However, despite SPECT cameras being very sensitive, they are inefficient, requiring long imaging times. Currently, it takes about 30-40 minutes for SPECT imaging; during this time the patient has to lie completely still. To address this challenge, we will develop an innovative SPECT camera architecture and image processing algorithm based on compressive coded imaging framework that accelerates SPECT image acquisition.

Background:

Scope:

The purpose of this project is to provide a testbed for continued research into the possible improvement of SPECT imaging efficiency and the reduction of imaging time using compressive coding. The goal of this project is thus to construct and provide Dr. Obrzut with mechanical equipment that would enable him and his team to assess the feasibility of using compressive coding to improve SPECT imaging. If successful, imaging time would significantly be decreased and more patients could be tended to in the same amount of time as before. Also, it would open up this live-saving technology to patients who might not have been able to use it due their inability to remain still for such a long time. Finally, with imaging time reduced, this would lessen associated imaging costs for hospitals and patients.

Design Solution:

Features:

1. 3-D (XYZ) PLOTTER

Figure 2. XYZ Plotter (collapsed configuration)

The Plotter (Figures 2 and 3) should be capable of motion in a 2-D plane and in the vertical direction. These two requirements were broken up with the combination of a 2-D Plotter and a Lifting Table. The grids should be variable to accommodate various patterns (e.g. 10 x 10 x 10 or 64 x 64 x 64).

Figure 3. XYZ Plotter (extended configuration)

2. PSEUDO-RANDOM ATTENUATING SCREEN

The Attenuating Screen (Figure 4) will serve in lieu of the (inefficient) collimator and will be placed over the camera head to enable compressive coding analysis. It should randomize the radiation intensities picked up by the camera at varying positions

Figure 4. Attenuating Screen (made of ballistic gel & ball bearings)

Video 3. CAD Rendering of Camera Imaging using Attenuating Screen

3. 2-D PHANTOM

The 2-D Phantom (Figure 5) should be modeled after the 3-D Jazsczak Phantom and would help assess whether the compressive coding worked as expected by comparing reconstructed image to known Phantom dimensions

Figure 5. Initial iterations of the 2-D Phantom

Video 4. CAD Rendering of 2-D Phantom being used

Additional Resources:

We highly encourage you to check out these resources to learn more about SPECT and Compressive Coding.

Page updated

Report abuse