“Algorithm Development Lifecycle in Medical Imaging:
Current State and Considerations for the Future”
Luciano M. Prevedello, MD, MPH
Vice-Chair for Medical Informatics and Augmented Intelligence in Imaging
Division Chief, Medical Imaging Informatics
Director, 3D and Advanced Visualization Lab
Associate Professor, Division of Neuroradiology,
Department of Radiology
Ohio State University Wexner Medical Center
Join via Zoom: https://stanford.zoom.us/j/267814863
Abstract:
This presentation will describe some of the most important considerations involved in creating algorithms in medical imaging from inception to deployment as well as continued model improvement and/or monitoring. Examples of experience to date from the OSU laboratory for augmented intelligence in imaging will be provided. New paradigms in model creation and the role of image challenge competitions will also be covered. Current issues with model validation and generalizability will also be introduced as well as considerations for future work in this area.
Refreshments will be provided.
“A Deep Learning Framework for Efficient Registration of MRI and Histopathology Images of the Prostate”
Wei Shao, PhD
Postdoctoral Research Fellow
Department of Radiology
Stanford University
“Applications of Generative Adversarial Networks (GANs) in Medical Imaging”
Saeed Seyyedi, PhD
Paustenbach Research Fellow
Department of Radiology
Stanford University
Join via Zoom: https://stanford.zoom.us/j/593016899
Refreshments will be provided
ABSTRACT (Shao)
Magnetic resonance imaging (MRI) is an increasingly important tool for the diagnosis and treatment of prostate cancer. However, MRI interpretation suffers from high interobserver variability and often misses clinically significant cancers. Registration of histopathology images from patients who have undergone surgical resection of the prostate onto pre-operative MRI images allows direct mapping of cancer location onto MR images. This is essential for the discovery and validation of novel prostate cancer signatures on MRI. Traditional registration approaches can be computationally expensive and require a careful choice of registration hyperparameters. We present a deep learning-based pipeline to accelerate and simplify MRI-histopathology image registration in prostate cancer. Our pipeline consists of preprocessing, transform estimation by deep neural networks, and postprocessing. We refined the registration neural networks, originally trained with 19,642 natural images, by adding 17,821 medical images of the prostate to the training set. The pipeline was evaluated using 99 prostate cancer patients. The addition of the images to the training set significantly (p < 0.001) improved the Dice coefficient and reduced the Hausdorff distance. Our pipeline also achieved comparable accuracy to an existing state-of-the-art algorithm while reducing the computation time from 4.4 minutes to less than 2 seconds.
ABSTRACT (Seyyedi)
Generative adversarial networks (GANs) are advanced types of neural networks where two networks are trained simultaneously to perform two tasks of generation and discrimination. GANs have gained a lot of attention to tackle well known and challenging problems in computer vision applications including medical image analysis tasks such as medical image de-noising, detection and classification, segmentation and reconstruction.In this talk, we will introduce some of the recent advancements of GANs in medical imaging applications and will discuss the recent developments of GAN models to resolve real world imaging challenges.
MIPS Seminar: “Tiny Bubbles, Big Impact: Exploring applications of nanobubbles in ultrasound molecular imaging and therapy”
Agata A. Exner, Ph.D.
Professor of Radiology and Biomedical Engineering
Department of Radiology
Case Western Reserve
Location: Beckman Center, B230
2:00pm – 3:00pm Seminar & Discussion
ABSTRACT
Sub-micron shell stabilized gas bubbles (aka nanobubbles (NB) or ultrafine bubbles) have gained momentum as a robust contrast agent for molecular imaging and therapy using ultrasound. The small size, extended stability and high concentration of nanobubbles make them an ideal tool for new applications of contrast enhanced ultrasound and ultra-
sound-mediated therapy, especially in oncology-related problems. Compared to microbub-bles, nanobubbles can provide superior tumor delineation, identify biomarkers on the vascu-lature and on tumors cells and facilitate drug and gene delivery into tumor tissue. The pat-terns of tissue enhancement under nonlinear ultrasound imaging of nanobubbles are distinct from conventional microbubbles especially in tissues exhibiting vascular hyperper-meability. Specifically, NB kinetics, quantified via time intensity curve analysis, typically show a marked delay in the washout rate and significantly increased area under the curve compared to larger bubbles. This effect is further enhanced by molecular targeting to cellular biomarkers, such as the prostate specific membrane antigen (PSMA) or the receptor protein tyrosine phosphatase, PTPmu. The unique contrast enhancement dynamics of nanobubbles are likely to be a result of direct bubble extravasation and prolonged retention of intact bubbles in target tissue. Thus, understanding the underlying mechanisms behind the unique nanobubble behavior can be the driver of significant future innovations in contrast enhanced ultrasound imaging applications. This presentation will discuss the fundamental challenges with nanobubble formulation and characterization and will showcase how the unique fea-tures of nanobubbles can be leveraged to improve disease detection and treatment using ultrasound.
Ron Kikinis, MD
Director of the Surgical Planning Laboratory
Professor of Radiology
Department of Radiology
Brigham and Women’s Hospital
Harvard Medical School
Title: Evolving Health Care from an Artisanal Organization into an Industrial Enterprise
Refreshments will be provided
Join via Zoom: https://stanford.zoom.us/j/996417088
Abstract: During the last decade, results from basic research in the fields of genetics and immunology have begun to impact treatment in a variety of diseases. Checkpoint therapy, for instance has fundamentally changed the treatment and survival of some patients with melanoma. The medical workplace has transformed from an artisanal organization into an industrial enterprise environment. Workflows in the clinic are increasingly standardized. Their timing and execution are monitored through omnipresent software systems. This has resulted in an acceleration of the pace of care delivery. Imaging and image post-processing have rapidly evolved as well, enabled by ever-increasing computational power, novel sensor systems and novel mathematical approaches. Organizing the data and making it findable and accessible is an ongoing challenge and is investigated through a variety of research efforts. These topics will be reviewed and discussed during the lecture.
About:
Dr. Kikinis is the founding Director of the Surgical Planning Laboratory, Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, and a Professor of Radiology at Harvard Medical School. This laboratory was founded in 1990. Before joining Brigham & Women’s Hospital in 1988, he trained as a resident in radiology at the University Hospital in Zurich, and as a researcher in computer vision at the ETH in Zurich, Switzerland. He received his M.D. degree from the University of Zurich, Switzerland, in 1982. In 2004 he was appointed Professor of Radiology at Harvard Medical School. In 2009 he was the inaugural recipient of the MICCAI Society “Enduring Impact Award”. On February 24, 2010 he was appointed the Robert Greenes Distinguished Director of Biomedical Informatics in the Department of Radiology at Brigham and Women’s Hospital. On January 1, 2014, he was appointed “Institutsleiter” of Fraunhofer MEVIS and Professor of Medical Image Computing at the University of Bremen. Since then he is commuting every two months between Bremen and Boston.
During the mid-80’s, Dr. Kikinis developed a scientific interest in image processing algorithms and their use for extracting relevant information from medical imaging data. Due to the explosive increase of both the quantity and complexity of imaging data this area of research is of ever-increasing importance. Dr. Kikinis has led and has participated in research in different areas of science. His activities include technological research (segmentation, registration, visualization, high performance computing), software system development, and biomedical research in a variety of biomedical specialties. The majority of his research is interdisciplinary in nature and is conducted by multidisciplinary teams. The results of his research have been reported in a variety of peer-reviewed journal articles. He is an author and co-author of over 350 peer-reviewed articles.
Follow us on Twitter: @StanfordIBIIS
MIPS Seminar
2:00-2:45 PM | Prof. Pawel Moskal
“Positronium Imaging with the J-PET Scanner”
Head of the Department of Experimental Particle Physics and Applications
Marian Smoluchowski Institute of Physics
Jagiellonian University, 30-348 Krakow, Poland
2:45-3:30 PM | Prof. Ewa Stepien
“Preclinical studies of positronium and extracellular vesicles biomarkers”
Head of the Department of Medical Physics
Marian Smoluchowski Institute of Physics
Jagiellonian University, 30-348 Krakow, Poland
ABSTRACT
As modern medicine develops towards personalized treatment of patients, there is a need for highly specific and sensitive tests to diagnose disease. Our research aims at improvement of specificity of positron emission tomography (PET) in assessment of cancer by use of positronium as a theranostic agent. During PET scanning about 40% of positron annihilations occur through the creation of positronium. “Positronium,” which may be formed in human tissues in the intramolecular spaces, is an exotic atom composed of an electron from tissue and the positron emitted by the radioinuclide. Positronium decay in the patient body is sensitive to the nanostructure and metabolism of human tissues. This phenomenon is not used in present PET diagnostics, yet it is in principle possible to exploit such environment modified properties of positronium as diagnostic biomarkers for cancer assessment. Our first in-vitro studies have shown differences of the positronium mean lifetime and production probability in healthy and cancerous tissues, indicating that they may be used as indicators for in-vivo cancer classification. For the application in medical diagnostics, the properties of positronium atoms need to be determined in a spatially resolved manner. For that purpose we have developed a method of positronium lifetime imaging in which the lifetime and position of positronium atoms are determined on an event-by-event basis. This method requires application of β+ decaying isotope that also emits a prompt gamma ray. We will argue that with total-body PET scanners, the sensitivity of positronium lifetime imaging, which requires coincident registration of the back-to-back annihilation photons and the prompt gamma, is comparable to the sensitivities for metabolic imaging with standard PET scanners.
Our research involves also development of diagnostic methods based on the extracellular vesicles (EVs), which are micro and nano-sized, closed membrane fragments. They are produced by native cells to facilitate the transfer of different signaling factors, structural proteins, nucleic acids or lipids even to distant cells. They are present in all body fluids and they are specific to their parental cells.
Our presentation will be divided into two parts. In the first, the method of positronium imaging and the pilot positronium images obtained with the J-PET detector (the first PET system built based on plastic scintillators) will be reported. This part of the presentation will include also description and perspectives of development of the J-PET technology in view of total-body PET imaging. The second part will concern preliminary results of the preclinical studies of positronium properties in cancerous and healthy tissues sampled from patients as well as in the frozen and living healthy and cancer skin cells in-vitro. The second part will include also description of the novel method for the diagnosis of diabetes and melanoma based on EVs used as biomarkers and drug delivery systems.
References:
P. Moskal, …. E. Ł. Stępień et al., Phys. Med. Biol. 64 (2019) 055017
- Moskal, B. Jasinska, E. Ł. Stępień, S. Bass, Nature Reviews Physics 1 (2019) 527
- Roman M… .E. Ł. Stępień, Nanomedicine 17 (2019) 137
- Ł. Stępień et al., Theranostics 8 (2018) 3874
Hosted by: Craig Levin, Ph.D.
Sponsored by the Molecular Imaging Program at Stanford and the Department of Radiology
Tessa Cook, MD, PhD
Assistant Professor of Radiology
Perelman School of Medicine
University of Pennsylvania
Title: Deploying AI in the Clinical Radiology Workflow: Challenges, Opportunities, and Examples
Abstract: Although many radiology AI efforts are focused on pixel-based tasks, there is great potential for AI to impact radiology care delivery and workflow when applied to reports, EMR data, and workflow data. Radiology-pathology correlation, identification of follow-up recommendations, and report segmentation can be used to increase meaningful feedback to radiologists as well as to automate tasks that are currently manual and time-consuming. When deploying AI within the clinical workflow, there are many challenges that may slow down or otherwise affect the integration. Careful consideration of the way in which radiologists may expect to interact with AI results should be undertaken to meaningfully deploy radiology AI in a safe and effective way.
Please note this seminar is now cancelled and will be rescheduled for a later date.
MIPS Seminar: Investigating and Imaging key molecular switches associated with Acquirement of Platinum-Taxol resistance in Epithelial Ovarian Cancer
Pritha Ray, Ph.D.
Principal Investigator & Scientific Officer F
Imaging Cell Signaling & Therapeutics Lab
ACTREC, Tata Memorial Center
Navi Mumbai, India
Radiomics and Radio-Genomics: Opportunities for Precision Medicine
Zoom: https://stanford.zoom.us/j/99904033216?pwd=U2tTdUp0YWtneTNUb1E4V2x0OTFMQT09
Pallavi Tiwari, PhD
Assistant Professor of Biomedical Engineering
Associate Member, Case Comprehensive Cancer Center
Director of Brain Image Computing Laboratory
School of Medicine | Case Western Reserve University
Abstract:
In this talk, Dr. Tiwari will focus on her lab’s recent efforts in developing radiomic (extracting computerized sub-visual features from radiologic imaging), radiogenomic (identifying radiologic features associated with molecular phenotypes), and radiopathomic (radiologic features associated with pathologic phenotypes) techniques to capture insights into the underlying tumor biology as observed on non-invasive routine imaging. She will focus on clinical applications of this work for predicting disease outcome, recurrence, progression and response to therapy specifically in the context of brain tumors. She will also discuss current efforts in developing new radiomic features for post-treatment evaluation and predicting response to chemo-radiation treatment. Dr. Tiwari will conclude with a discussion on her lab’s findings in AI + experts, in the context of a clinically challenging problem of post-treatment response assessment on routine MRI scans.
Ge Wang, PhD
Clark & Crossan Endowed Chair Professor
Director of the Biomedical Imaging Center
Rensselaer Polytechnic Institute
Troy, New York
Abstract:
AI-based tomography is an important application and a new frontier of machine learning. AI, especially deep learning, has been widely used in computer vision and image analysis, which deal with existing images, improve them, and produce features. Since 2016, deep learning techniques are actively researched for tomography in the context of medicine. Tomographic reconstruction produces images of multi-dimensional structures from externally measured “encoded” data in the form of various transforms (integrals, harmonics, and so on). In this presentation, we provide a general background, highlight representative results, and discuss key issues that need to be addressed in this emerging field.
About:
AI-based X-ray Imaging System (AXIS) lab is led by Dr. Ge Wang, affiliated with the Department of Biomedical Engineering at Rensselaer Polytechnic Institute and the Center for Biotechnology and Interdisciplinary Studies in the Biomedical Imaging Center. AXIS lab focuses on innovation and translation of x-ray computed tomography, optical molecular tomography, multi-scale and multi-modality imaging, and AI/machine learning for image reconstruction and analysis, and has been continuously well funded by federal agencies and leading companies. AXIS group collaborates with Stanford, Harvard, Cornell, MSK, UTSW, Yale, GE, Hologic, and others, to develop theories, methods, software, systems, applications, and workflows.
MIPS Seminar Series: Translational Opportunities in Glycoscience
Carolyn Bertozzi, PhD
Director, ChEM-H
Anne T. and Robert M. Bass Professor in the School of Humanities and Sciences
Professor, by courtesy, of Chemical and Systems Biology
Stanford University
Location: Zoom
Webinar URL: . https://stanford.zoom.us/j/94010708043
Dial: US: +1 650 724 9799 or +1 833 302 1536 (Toll Free)
Webinar ID: 940 1070 8043
Passcode: 659236
12:00pm – 12:45pm Seminar & Discussion
RSVP Here
ABSTRACT
Cell surface glycans constitute a rich biomolecular dataset that drives both normal and pathological processes. Their “readers” are glycan-binding receptors that can engage in cell-cell interactions and cell signaling. Our research focuses on mechanistic studies of glycan/receptor biology and applications of this knowledge to new therapeutic strategies. Our recent efforts center on pathogenic glycans in the tumor microenvironment and new therapeutic modalities based on the concept of targeted degradation.
ABOUT
Carolyn Bertozzi is the Baker Family Director of Stanford ChEM-H and the Anne T. and Robert M. Bass Professor of Humanities and Sciences in the Department of Chemistry at Stanford University. She is also an Investigator of the Howard Hughes Medical Institute. Her research focuses on profiling changes in cell surface glycosylation associated with cancer, inflammation and infection, and exploiting this information for development of diagnostic and therapeutic approaches, most recently in the area of immuno-oncology. She is an elected member of the National Academy of Medicine, the National Academy of Sciences, and the American Academy of Arts and Sciences. She also has been awarded the Lemelson-MIT Prize, a MacArthur Foundation Fellowship, the Chemistry for the Future Solvay Prize, among many others.
Hosted by: Katherine Ferrara, PhD
Sponsored by: Molecular Imaging Program at Stanford & the Department of Radiology