Multiparametric PET Suite AI

Multiparametric PET AI

The first fully integrated, AI-powered solution to provide additional parameters and enable absolute quantification, compared to SUV alone


One of the most important aspects of PET imaging is the ability to quantify the uptake of tracer in a specific tissue, and the implementation of standard uptake values (SUV) based on static images was a key milestone to facilitate the utilization of semi-quantitative metrics in the clinic. However, PET tracer distribution — a dynamic process altered by a number of factors specific to each organ and region of interest — cannot always be accurately accounted for with SUV.

Features & Benefits

Multiparametric PET AI1 is completely automated and integrated into the PET/CT workflow, based on the Patlak model. It allows users to obtain images of SUV, metabolic glucose rate (Ki) and distribution volume (DV) all at once for more reproducible images and more efficient information for planning and therapy strategy.

The implementation of Multiparametric PET AI leverages several unique innovations to provide a robust solution: FlowMotion™ continuous bed motion technology, whole-body dynamic imaging, ALPHA technology, direct reconstruction, and syngo.via.

  • Fully automated, AI-powered solution obtains whole-body images of metabolic glucose rate and distribution volume, in addition to SUV images
  • Flexible acquisition protocols enable use in clinical research settings
  • Reduces the variability of quantitative measurements due to differences in uptake time, patient body size and blood glucose levels, leading to better characterization of tracer uptake compared to SUV alone
  • Provides additional information to evaluate therapy effectiveness based on metabolic glucose rates compared to SUV alone
Multiparametric PET Suite<br />
Data courtesy of Yale University, New Haven, USA.

Dynamic imaging has been limited to a single-bed position, with the organ of interest positioned in the center of the detectors. With FlowMotion, you can obtain whole-body dynamic images by continually moving the bed through the gantry in a bi-directional motion along the entire body so you can visualize the tracer distribution over time.

Parametric modeling requires the acquisition of an arterial input function (AIF). This is the amount of tracer available in the blood pool over time.

The gold standard is to perform multiple arterial blood samples over time &ndash; an operator-intensive process with a small risk for the patient. ALPHA is a Siemens, proprietary machine-learning technology that identifies anatomical landmarks, so PET images can obtain the arterial input function (AIF) by automatically measuring the uptake in a region of interest located in the descending aorta or in the left ventricle without the need for an invasive measurement.

The gold standard is to perform multiple arterial blood samples over time – an operator-intensive process with a small risk for the patient. ALPHA is Siemens Healthineers’ proprietary machine-learning technology that identifies anatomical landmarks, so PET images can obtain the arterial input function (AIF) by automatically measuring the uptake in a region of interest located in the descending aorta or in the left ventricle without the need for an invasive measurement.

Technical Specifications

General Requirements

  • Biograph Vision
  • Biograph Horizon
  • Biograph mCT

Biograph mCT HW/SW platform must be running PETsyngo VG70 or higher with MARS, ICS T13 and IRSmx3 hardware or higher.

FlowMotion and PET Dynamic option are required to enable this option.

Please Note: Additional technical pre-requisites may apply. Upon receiving your request, your local Siemens representative will clarify whether your system meets the requirements.

Learn more about Multiparametric PET AI

1