Reconstruction algorithm for high-quality positron emission tomography (PET) images

This technology is an iterative reconstruction algorithm that improves resolution and contrast in positron emission tomography (PET) image data.

Unmet Need: Algorithm for reducing artifacts in PET image data

Quantification of nuclear imaging and tomographic brain images, such as PET or single-photon emission computed tomography (SPECT), acquired with the use of injected contrast agents or tracers requires the measurement of tracer blood concentration over the course of the scan. Typically, sampling is done via an arterial catheter, which is costly, risky to patients, time consuming, and involves error prone blood chemistry analysis. Iterative reconstruction with point spread function (PSF) modeling is a non-invasive method for improving contrast recovery in PET images, but also introduces ringing artifacts and over enhancement that is contrast and object size dependent. Currently, there are no available methods to enhance PET images that are both non-invasive and address the artifacts introduced with PSF modeling.

The Technology: Effective, non-invasive method for improving image data

This technology is a regularized reconstruction method that incorporates locally-weighted total variation denoising to suppress image artifacts induced by PSF modeling. This method is capable of suppressing ringing artifacts, while effectively maintaining contrast recovery. Additionally, this non-invasive approach is also capable of quantifying radioligand binding in patients without the need to collect arterial blood. For this, the imaging data and electronic health records can be used to predict one or more anchors to generate an arterial input function (AIF) for the radioligand. As a result, this technology provides an effective, non-invasive method for improving image data.

This reconstruction method has been validated on a simulated cylindrical phantom, where it demonstrated that ringing effects were effectively suppressed while maintaining contrast recovery.

Applications:

• Reconstructing positron PET image data
• Reconstructing single-photon emission computed tomography (SPECT) image data
• Quantification of PET signal from small brain structures
• Tumor detection with PET
• Quantification of amyloid plaques as detected with PET tracers

Advantages:

• 3D reconstruction
• High resolution for detection of small-sized abnormalities such as tumors
• Reduces noise level and artifacts

Lead Inventor:

Andrew Laine, Ph.D.

Patent Information:

Patent Pending (US 20170039706)

Related Publications:

• Roccia E, Mikhno A, Zanderigo F, Angelini ED, Ogden RT, Mann JJ, Laine AF. “Non-invasive quantification of brain [¹⁸F]-FDG uptake by combining medical health records and dynamic PET imaging data” Conf Proc IEEE Eng Med Biol Soc. 2015;2015:2243-6.

• Mikhno A, Zanderigo F, Todd Ogden R, John Mann J, Angelini ED, Laine AF, Parsey RV. “Toward noninvasive quantification of brain radioligand binding by combining electronic health records and dynamic PET imaging data” IEEE J Biomed Health Inform. 2015 Jul;19(4):1271-82.

• Mikhno A, Angelini ED, Bai B, Laine AF. “Locally weighted total variation denoising for ringing artifact suppression in PET reconstruction using PSF modeling” IEEE Int Symp Biomed Imaging. 2013 Dec 31; 2013: 1252–1255.

Tech Ventures Reference:

• IR CU15180, CU14251

• Licensing Contact: Jerry Kokoshka