OpenKMI Workshop at IEEE Medical Imaging Conference 2023

This talk will describe the motivation and current progress of the IEEE NPSS-supported Open Kinetic Modeling Initiative. It will also discuss the next possible plans for this initiative.

Speaker Bio: Guobao Wang is a Professor in the Department of Radiology, University of California Davis Health with 18 years of experience in PET parametric imaging.  His lab develops both organ-specific (e.g. liver) and total-body  PET kinetic modeling and parametric imaging methods to advance molecular imaging in health and disease. He serves as Director of the Kinetic Modeling Core at the EXPLORER Molecular Imaging Center, UC Davis. 

The Yale PET Center has been actively involved in the development and application of models for new tracers and new modeling methods.   The modeling infrastructure is built around a SQL database which tracks all modeling steps and the parameters used in each step. The strengths and weaknesses of this approach in an academic research environment will be presented.

Speaker Bio: Dr. Carson is Professor of Radiology and Biomedical Engineering at Yale University with over 40 years of experience in new algorithms for PET image recon­struction, mathematical models for novel radiopharmaceuticals, and applications of PET tracers in clinical populations and preclinical models of disease.

This talk will review our recent experience in building a kinetic modeling research program at the MGH Gordon Center. I will summarize foundational work done before my arrival in 2011; review the development of infrastructure, tools, and training; and highlight examples of projects leveraging these resources.

Speaker Bio: Marc Normandin, PhD, is an Associate Professor of Radiology at Harvard Medical School, Associate Director of the Gordon Center for Medical Imaging at Massachusetts General Hospital, and Associate Director of the MGH PET Core. His research focuses on the development and application of functional imaging methods, especially quantification of PET data through tracer kinetic modeling.

This talk will highlight the ongoing kinetic modeling research in Thailand, with a primary focus on Chulalongkorn University's work. Clinical implementations like epileptogenic zone localization, bone turnover prediction using non-invasive biomarkers Ki-Patlak NaF in CKD patients, and dual-time-point parametric imaging feasibility for prostate cancer will be presented.

Speaker Bio: Dr. Kitiwat Khamwan is an Assistant Professor at the Department of Radiology, Faculty of Medicine, Chulalongkorn University. Over the past 15 years, he has made significant contributions to education and training in nuclear medicine across Southeast Asia and the Asia-Oceania regions. His research interests focus on radionuclide therapy dosimetry, PET performance characteristics, theranostics, kinetic modeling, and parametric imaging.

Open-source software (OSS) has the potential to accelerate research into new methods as well as their translation into practice. In this talk, I will give an overview of existing OSS for kinetic modelling of dynamic PET data. However, developing high quality and sustainable OSS can be daunting. Therefore, I will discuss some software development practices learned from our experience with various OSS packages, including STIR and SIRF.

Speaker Bio: Prof. Kris Thielemans is Professor in Medical Imaging Physics at the Institute of Nuclear Medicine (INM), University College London, UK. Prof. Thielemans worked in industry from 2001 to 2011. His research covers PET-CT, PET-MR and SPECT, including motion detection and correction, quantitative image reconstruction and synergistic (joint) reconstruction of multi-modality data and kinetic modelling, with an aim towards translation to practice, as exemplified in various projects with industry. Dr. Thielemans is the lead designer and maintainer of the open source project Software for Tomographic Image Reconstruction (stir.sourceforge.net), which allows quantitative image reconstruction for PET and SPECT data, and the Synergistic Image Reconstruction Framework (SIRF) for PET, SPECT and MR data (www.ccpsynerbi.ac.uk). He now co-leads the Emission Tomography Standardization Initiative (ETSInitiative.org) which aims to develop a standard for PET Raw data. Dr. Thielemans has co-authored 108 peer-reviewed publications, 250+ conference contributions and holds 17 patents (1 pending).

In this talk, I will introduce the audience to the brain imaging data structure (BIDS, https://bids.neuroimaging.io/) and its extension aimed at PET data. Additionally, I will showcase an example of PET data sharing by giving a status update about the OpenNeuroPET project (https://openneuropet.github.io/).

Speaker Bio: Melanie Ganz is an Associate Professor at the University of Copenhagen, Denmark. Her PhD in computer science is from the University of Copenhagen, following which she held postdocs at the A.A. Martinos Center for Biomedical Imaging at Massachusetts General Hospital and Harvard Medical School in Boston, USA, and at the Neurobiology Research Unit at the Copenhagen University Hospital, Denmark. Melanie’s research focuses on the application of medical image processing and machine learning to clinical medical image data, mostly neuroscientific data. Additionally, she aims to embrace Open Science principles in her research and is hence involved in data sharing initiatives such as the Brain Imaging Data Structure.

While access to total body and long axial field-of-view PET scanners is growing, they remain rare and data from them are not widely available. This hinders methods development and the ability to gain new insights from the data. This talk will discuss ideas and challenges around creating a  communal data and software sharing platform for total-body PET.

Speaker Bio: Ramsey Badawi, PhD is a PET physicist with approximately 30 years of experience. His major contribution to the field has been the development of total-body PET in partnership with Simon Cherry, PhD. Ramsey currently serves as co-Director of the EXPLORER Molecular Imaging Center and vice-chair for Research in the Radiology Department at UC Davis.

Siemens Healthineers mantra when implementing the whole body (WB) FDG Patlak PET feature was to make it “wickedly automated” to facilitate its use in the clinic. This talk will summarize this clinical use and highlight how it adds more information and confidence versus static standard of care PET FDG imaging alone. Feature enhancements and moving beyond the Patlak FDG model will also be discussed.

Speaker Bio: Anne M. Smith, PhD, works for Siemens Healthineers Molecular Imaging in Knoxville, TN as a Senior Staff Systems Engineer.  She has worked in the medical imaging field for thirty years and has a passion for commercializing mature scientific discoveries for broad clinical use.  Her scientific interests focus on PET image quality, image analysis, kinetic modeling and translational imaging of new biomarkers.   

In this presentation we highlight the data acquisition, image reconstruction and kinetic analysis online/offline tools enabling WB dynamic imaging for clinical and research applications. Example cases are shown and future perspectives are also discussed. 

Speaker Bio: Fotis Kotasidis PhD, works at GE HealthCare, as PET scientist for the last 6 years and is based in Zurich, Switzerland. Prior to that he had spent 10 years in academia in UK and Switzerland as PET Scientist. His interests focus on PET quantification, kinetic modelling, data corrections and image reconstruction amongst others.

United Imaging Healthcare has launched a toolkit (uKinetics) for kinetic modeling for both research and clinical purposes. uKinetics provides time-activity-curve-based analysis and parametric imaging with multiple options of input function. Both graphic analyses and compartmental models are supplied for the parametric estimation. As part of a newly developed United Imaging PET Reconstruction Toolbox (URT), the research version of uKinetics is supported by a variety of features in URT, i.e. configurable iterative reconstruction platform, patient motion correction, AI-based image analysis, and Monte-Carlo simulation. The clinical version of uKinetics has been approved by FDA on the uMI Panorama PET/CT systems.

Speaker Bio: Dr. Qing Ye received her Ph.D. from Tsinghua University in 2020. She is a Senior Engineer in the Department of Application and Advanced Algorithms, Molecular Imaging, Shanghai United Imaging Healthcare. She is in charge of algorithm development and research of uKinetics.

Speaker Bio: Geoff Warnock leads the PMOD team in Bruker Preclinical Imaging and for much of his career balanced academic and industry projects. He was directly involved in the development of 18F-UCB-H for quantification of synaptic density, first-in-man studies for tracer translation, and has advised many PMOD customers on their quantitative PET projects.