Office of Technology Transfer – University of Michigan

ASPIRE = A sparse precomputed iterative reconstruction environment

Technology #1517

Background

Various medical image modalities including Single-Photon Electron Computed Tomography (SPECT) require image reconstruction. As images have finite sampling and statistical noise, images to accurately represent the information of interest (such as the radioactivity distribution) requires reconstruction via various algorithms.

Technology

Researchers at the University of Michigan developed ASPIRE (A sparse precomputed iterative reconstruction library), a collection of ANSI C programs for tomographic image reconstruction. The software allows for rapid calculation of the local noise and resolution properties of penalized-likelihood image estimates. These techniques are appropriate for 2-D or 3-D SPECT systems with nonuniform attenuation and are based on precomputing portions of the predictors that are independent of the object being scanned.

Applications and Advantages

Applications

  • Image reconstruction

Advantages

  • Appropriate for both 2-D and 3-D SPECT systems