Print PageEmail Page

SPIE Medical Imaging 2017 news and photos

Browse below for updates from throughout the week.

 

Ending on a sunny note

SPIE Medical Imaging 2017

Like the weather, SPIE Medical Imaging 2017 ended on a sunny note,
with attendees already looking forward to next year:
see you in Houston, Texas, USA, 10–15 February 2018!

 

Getting social — enjoying your photos!

Duke University contingent at SPIE Medical Imaging 2017 Poster detail, SPIE Medical Imaging 2017

We've enjoyed your photos on the SPIE #MedicalImaging Twitter feed — thanks so much for sharing with everyone! Among them, @JosephLo16 posted the photo at left of the Duke University contingent, and @ooliverdiaz from the University of Girona posted detail (at right) from one of his group's papers in the Wednesday evening poster session. To see more, follow the Twitter feed below.

 

Off to a great start!

Medical Imaging 2017 working group on Deep Learning in Medical Imaging Analysis

Attendance was strong during the first day of technical sessions, with an audience of
nearly 350 attending the workshop on "Deep Learning in Medical Imaging Analysis."

SPIE’s Medical Imaging conference, the premier event covering acquisition, display, processing, analysis, perception, and decision support, got off to a great start this week in Orlando, Florida. Activities began with a series of education courses on Saturday 11 February offering topics ranging from Fundamentals of Medical Image Processing and Analysis to High-Performance Computing and Writing for Publications in Medical Imaging.

Join the conversation:
#MedicalImaging

Twitter @ SPIEtweets Instagram @SPIEphotonics Facebook SPIE, the international society for optics and photonics SPIE, the international society for optics and photonics, on LinkedIn

 

Breast cancer imaging: considering the patient

Conferences began on Sunday, with opening keynotes in three.

Margarita Zuley at SPIE Medical Imaging

Margarita Zuley

Margarita Zuley, University of Pittsburgh Medical Center and Magee-Womens Hospital, one of the top radiologists in the U.S., spoke to a packed room in the Image Perception, Observer Performance, and Technology Assessment conference.

Zuley provided a comprehensive overview of advances in imaging for breast cancer over the last two decades. Providing her perspective as a breast radiologist in an effort to help motivate the research community to develop new solutions, Zuley covered many examples of the challenges in making diagnosis using today’s technologies.

She noted the necessity of taking into consideration the physiology of the patient, and showed multiple examples in which tumors were identifiable using one imaging modality, but not with others.

However, the multimodal imaging provided by ultrasound, tomosynthesis, and MRI can help increase the success in detecting cancer and reduce false positives. While additional imaging modalities provide expanded capability, Zuley illustrated how much more complicated the radiologist’s task has become over the years with a picture of the bank of computers and equipment used by today’s practitioners.

 

Managing data costs — and privacy

Anne LeGrand at SPIE Medical Imaging

Anne LeGrand
(Ken Hanson photo)

Anne LeGrand, Global General Manager for Watson Health Imaging at IBM, gave the day's second keynote, as part of the Image Processing conference.

LeGrand opened with an overview of grand challenges pursued by IBM leading up to the launch of IBM Watson Healthcare. Advances in image acquisition technologies and precision have led to an exponential growth in medical imaging data. The opportunity is to leverage this data to improve care, control costs, and enhance health care globally, LeGrand said.

She discussed some of the opportunities and partnerships IBM is currently pursing. Extensive discussion session ensued where members of the standing-room-only audience had the opportunity to ask questions ranging from data privacy concerns to IBM’s business model.

 

Neuroimaging and human behavior

Bruce Rosen at SPIE Medical Imaging

Bruce Rosen

Bruce Rosen, Harvard Medical School, gave the day’s final keynote talk, on advances in neuroimaging in the Biomedical Applications in Molecular, Structural, and Functional Imaging conference. Rosen, a world-leading expert in functional neuroimaging, covered a wide array of techniques and topics during his captivating lecture.

We do imaging to better understand diseases and patients to guide treatments, Rosen noted. Advances in imaging techniques have vastly contributed to our understanding of neural circuits and their role in human behavior. Rosen ended his talk discussing tools for viewing and mapping connections in the human brain.

 

Intensifiying the focus: technical workshops

Kyle Myers at SPIE Medical Imaging

Kyle Myers

Technical workshops closed the first day.

“Virtual Clinical Trials for Breast Imaging” drew a wide audience to hear the latest in breast phantoms, computational tools, reliable and computationally efficient mathematical model observers, and industry and U.S. Food and Drug Administration (FD) perspectives. (Among the speakers was Kyle Myers of the FDA, at right).

“Deep Learning in Medical Image Analysis” saw in excess of 350 participants who gathered to learn about the newest research and opportunities to leverage deep learning in their research. Deep learning is increasingly used in medical imaging, but brings new challenges and requires new perspectives on existing problems.

 

Enjoy some networking with your coffee

SPIE Medical Imaging 2017
SPIE Medical Imaging 2017 SPIE Medical Imaging 2017

 

Prognostications and predictions: implications for precision medicine

Anant Madabhushi at SPIE Medical Imaging 2017

Anant Madabhushi

Anant Madabhushi, Case Western Reserve University, gave Monday morning's keynote talk, in the conference on Digital Pathology.

Madabhushi, an original co-founder of the Digital Pathology conference, spoke to a standing room only audience on "Predictive Radiomics and Pathomics: Implications for Precision Medicine."

Anant reviewed the statistics of cancer deaths underpinning the need for new tools to aid radiologists in the diagnosis, prognosis, and treatment of cancers, and detailed various current and emerging techniques and their application to a wide range of cancers.

He closed by contrasting the capabilities of techniques based on deep learning for object detection and classification versus those using emerging handcrafted features approaches for improved predictive modeling of tissue appearance.

Elsewhere, conferences on Physics of Medical Imaging and Image Perception, Observer Performance, and Technology Assessment held a well-attended joint session on "Task-based Assessment in CT."

 

Lunch on the lawn

SPIE Medical Imaging 2017

The grounds of the Renaissance Hotel provided the perfect site for a lunch break on Monday.

SPIE Medical Imaging 2017 SPIE Medical Imaging 2017

 

Wagner Award winners announced

Adrian Homansky, Leonard Berliner at SPIE Medical Imaging Ahmet Cakir, Leonard Berliner at SPIE Medical Imaging

After a welcome to attendees Monday afternoon, symposium chair Leonard Berliner (Weill Cornell Medical College and New York Methodist Hospital) announced first-place and runner-up winners of the Robert F. Wagner All-Conference Student Paper Awards.

Adrian Howansky of Stony Brook University (at left, with Berliner) won first place for “Direct measurement of Lubberts effect in CsI:Tl scintillators using single x-ray photon imaging” (10132-8).

Ahmet Cakir of Vanderbilt University (at right) was named runner-up, for “Evaluation of a high-resolution patient-specific model of the electrically stimulated cochlea” (10135-21).

The full group of finalists (below) was congratulated for their exceptional work.

Robert F. Wagner All-Conference Student Paper Award finalists, SPIE Medical Imaging 2017

 

Deep learning and medical applications: Greg Corrado plenary talk

Berkman Sahiner, Greg Corrado, Leonard Berliner at SPIE Medical Imaging 2017

From left, Berkman Sahiner, Greg Corrado,
and Leonard Berliner

Plenary speaker Greg Corrado, Director of Augmented Intelligence Research at Google, gave a captivating talk titled “Applying Deep Learning to Medical Imaging” on Monday afternoon to a capacity audience.

After an overview of deep learning, Corrado provided examples where deep neural networks have been successfully employed in areas such as voice and image recognition as well as applications in computer vision.

He closed with a discussion of the potential to leverage the power of deep learning for medical applications commenting that in order for it to be successful it will required close collaboration between doctors and technologists.

Corrado was presented with an SPIE biophotonics tie by symposium chairs Berkman Sahiner (U.S. Food and Drug Administration) and Leonard Berliner (Weill Cornell Medical College and New York Methodist Hospital).

 

Poster session: getting it straight from the authors

SPIE Medical Imaging 2017 poster session  SPIE Medical Imaging 2017 poster session
 SPIE Medical Imaging 2017 poster session  SPIE Medical Imaging 2017 poster session

 

Dessert draws students and experts

Dessert with the Experts at SPIE Medical Imaging 2017

Students were treated to a double dip Monday evening, in a networking event with experts over desserts.

 

Making connections: Women's Networking Lunch

SPIE Medical Imaging 2017 Women's Networking Lunch

SPIE Fellow Kyle Myers, U.S. Food and Drug Administration, welcomed colleagues to a Women's Networking Lunch Tuesday afternoon, for informal discussions and making connections.

Women's Networking Lunch at SPIE Medical Imaging 2017

 

Going the last mile in clinical translation

Aaron Sodickson at SPIE Medical Imaging 2017

Aaron Sodickson

Physics of Medical Imaging keynote speaker Aaron Sodickson of Brigham and Women's Hospital, Harvard Medical School, on Tuesday afternoon highlighted the basic needs for technology innovations to successfully translate into clinical practice. Through several case studies, he gave real-life examples of improvements that offered well-defined clinical benefits yet had varying degrees of success. Roadblocks included ease of use and integration into existing workflows.

Sodickson closed by encouraging all innovators to "think about the last mile". He stressed the need to consider clinical integration and workflow early in the development of any new technology, and to engage with collaborators to get end user input as early as possible.

 

Keen interest in 3D printing in medical imaging

Around 200 people turned out for a workshop highlighting presentations from leaders in developing and applying 3D printing technologies to medical applications. This rapidly evolving technology is sure to find many more applications as new tools and materials are developed.

Speakers were:

  • Andrew Maidment,  Hospital of the University of Pennsylvania (United States)
  • Joseph Lo, Duke University
  • Jonathan Morris, Mayo Clinic, Rochester
  • Shaochen Chen, University of California, San Diego.

 

Meeting the challenge: Ultrasound Computed Tomography Data

USTC panel at SPIE Medical Imaging 2017

Nicole Ruiter, Karlsruhe Institute of Technology, and Koen van Dongen, Technische Universiteit Delft, moderated a panel discussion among participants in the Ultrasound Computed Tomography Data Challenge (USCT), a new technology aimed primarily at breast cancer imaging.

This challenge was established to bring together experts to discuss the need for a USCT database to help with theoretical and practical aspects of this field. Topics discussed included the standardization of data and restrictions, how to develop phantoms and what characteristics are required, IP issues, and longer-term plans.

Panelists, above from left, were Thomas Mattews (Washington University in St. Louis), Marylin Pérez-Liva (Universidad Complutense de Madrid), Ken-ichi Kawabata (Hitachi, Ltd.), Roberto Janniel Lavarello Montero (Pontificia Universidad Catolica del Peru), Ivana Balic (SonoView Acoustic Sensing Technologies), and Neb Duric (Delphinus Medical Technologies).

 

So that's how you do it! Live Demonstrations in the ballroom

Live Demonstrations session at SPIE Medical Imaging 2017

Tuesday evening’s Live Demonstrations workshop was well received by the enthusiastic participants who were able to see and experience first-hand the many advanced developments on show.

“A Multimodal Workstation for Analysis of Retinal Images” was selected as the winning demonstration. Team members from Radboud University Medical Center are Bart Liefers, Freerk Venhuizen, Thomas Theelen, Carol Hoyng, and Clara Sanchez; from Fraunhofer MEVIS are Nadine Traulsen and Stefan Heldmann; and Bram van Ginneken is affiliated with both organizations.

Live Demonstrations session at SPIE Medical Imaging 2017 Live Demonstrations session at SPIE Medical Imaging 2017
Live Demonstrations session at SPIE Medical Imaging 2017 Live Demonstrations session at SPIE Medical Imaging 2017

 

Regulatory role in innovation and new solutions

Kyle Myers at SPIE Medical Imaging

Kyle Myers

Kyle Myers, U.S. Food and Drug Administration, gave the keynote in the Computer-Aided Diagnosis (CAD) conference Wednesday morning, speaking on the FDA's role in the innovation and evaluation of evolving CAD solutions.

After reviewing the agency’s mission, Myers provided background and examples of the process for medical device premarket submission for clearance.

She also discussed research activities in support of the mission and closed talking about new FDA initiatives.

 

Significant strides in liver cancer treatment

William Jarnagin at SPIE Medical Imaging

William Jarnagin

William Jarnagin, Memorial Sloan-Kettering Cancer Center, provided the keynote Wednesday morning for the Image-guided Procedures, Robotic Interventions, and Modeling conference.

His talk concentrated on the application of new imaging tools to inform the care, treatment, and cure of cancer patients with a specific focus on liver cancer.

Jarnagin provided a historical review of liver surgery, citing examples of challenges that surgeons have faced over the years. Significant strides have been made in the last two decades, he noted, with new image guidance tools and more sophisticated patient selection protocols offering promise for further improvements.

 

When the weather is blustery in Orlando ...

Indoor lunch at SPIE Medical Imaging 2017

... lunch comes indoors! Above, Wednesday's mid-day meal in the Renaissance Hotel.

 

PROSTATEx Grand Challenge

Samuel Armstrong at SPIE Medical Imaging

Samuel Armato

A Wednesday afternoon session chaired by Samuel Armato (at right) and Karen Drucker of the University of Chicago celebrated the success of the PROSTATEx Grand Challenge, evidenced by more than 70 submissions from more than 30 groups around the world — including many from group not previously involved in SPIE Medical Imaging.

SPIE, along with the support of the American Association of Physicists in Medicine (AAPM) and the National Cancer Institute (NCI), conducted the challenge to provide participants from academia, industry, and government a structured means to compare quantitative image analysis methods for the diagnostic classification of clinically significant prostate lesions.

The two top-performing teams presented their work Wednesday afternoon, highlighting the methods used and potential for future improvements.

Winners (below) were announced in Monday’s plenary session.

PROSTATEx Grand Challenge winners from Hospital Albert Einstein

PROSTATEx Grand Challenge winners from the Hospital Albert Einstein
with symposium chair Leonard Berliner
PROSTATEx Grand Challenge runner-up Jarrel Chen Yi Seah with Leonard Berliner

PROSTATEx Grand Challenge runner-up
Jarrel Chen Yi Seah (Alfred Health)
with symposium chair
Leonard Berliner
PROSTATEx Grand Challenge runner-up Saifeng Liu with Leonard Berliner

PROSTATEx Grand Challenge runner-up
Saifeng Liu (MRI Institute for Biomedical
Health Research) with
symposium chair Leonard Berliner

 

Digital Mamography DREAM Challenge

Symposium chair Berkman Sahiner (U.S. Food and Drug Adminstration) introduced the Digital Mammography DREAM Challenge being conducted by Sage Bionetworks, in partnership with the open science DREAM Challenges community.

This Coding4Cancer Challenge is aimed at improving the accuracy of breast cancer detection on full-field digital screening mammograms, offers up to $1,000,000 in cash prizes, and is expected to be completed by 9 May.

 

Ultrafast ultrasound blood flow imaging

Lasse Løvstakken at SPIE Medical Imaging

Lasse Løvstakken

Lasse Løvstakken, Norwegian University of Science and Technology, gave a keynote talk Wednesday afternoon in the Ultrasonic Imaging and Tomography conference, on ultrafast ultrasound blood flow imaging.

Løvstakken gave a thorough and far-reaching overview of advances in ultrasound imaging starting with a historical perspective of the pioneering work conducted at his institute. He highlighted current state-of- the-art methods including the many advantages of real-time color Doppler imaging first introduced in the mid ’80s.

He closed his excellent presentation with a discussion of future opportunities in the field such as high-frame-rate imaging.

 

Getting into the detail in the poster session

Poster presenters were kept busy throughout the evening Wednesday talking with attendees and explaining the details of their research.

Meeting up at the event were University of Chicago colleagues, from left below, Karen Drukker, Natasha Antropova, Benjamin Huynh, and SPIE President-Elect Maryellen Giger.

Karen Drukker, Natasha Antropova, Benjamin Huynh, and SPIE President-Elect Maryellen Giger at SPIE Medical Imaging Poster session at SPIE Medical Imaging
Poster session at SPIE Medical Imaging Poster session at SPIE Medical Imaging

 

3D medical printing: multiple applications, dramatic growth

William Weadock at SPIE Medical Imaging 2017

William Weadock

William Weadock, University of Michigan, gave the week's final keynote Thursday morning, in the conference on Imaging Informatics for Healthcare, Research, and Applications.

In an introduction to 3D medical printing, he noted the field's dramatic increase in attention to 3D printing in the past few years. Applications are many, such as customized, personalized implants, life-saving tracheal stents for infants, and cutting guides for surgical procedures, with even more envisioned in this rapidly growing field. Multiple printing technologies allow objects to be printed in plastic, metals, and even cells, Weadock noted, helping inspire and drive new capabilities.

 

Physics of Medical Imaging: best paper awards

Physics of Medical Imaging 2017 best paper awards

Chairs Taly Gilat Schmidt (Marquette University) and Joseph Lo (Duke University Medical Center) presented awards Thursday morning for outstanding work in the Physics of Medical Imaging conference.

First-place winner of best student paper awards sponsored by Carestream was Adrian Howansky, Stony Brook University, et al., for “Direct measurement of Lubberts effect in CsI:Tl scintillators using single x-ray photon imaging” (10132-8).

Runners-up were:

  • Martin Sjöli, KTH Royal Institute of Technology, et al., for “Optimal sinogram sampling with temporally offset pixels in continuous rotation CT” (10132-90)
  • Lucas Borges, Escola de Engenharia de São Carlos, Universidade de São Paulo, et al., for “Pipeline for effective denoising of digital mammography and digital breast tomosynthesis” (10132-5).

The cum laude poster presentation award, sponsored by Siemens Healthineers, was awarded to Kazuki Shigeta, et al., of Toray Industries, Inc., for “High spatial resolution performance of pixelated scintillators” (10132-145).

Seven honorable mention awards were also presented.

 

Image-Guided Procedures, Robotics, Modeling: best paper awards

Image-Guided Procedures best paper awards

From left, Robert Webster, Haichong Zhang,
Michael Ketcha, Ahmet Cakir, Baowei Fei

Young Scientist Awards sponsored by Siemens Healthineers in the Image-Guided Procedures, Robotic Interventions, and Modeling conference were presented Thursday afternoon by conference chairs Robert Webster (Vanderbilt University) and Baowei Fei (Emory University).

Michael Ketcha (Johns Hopkins University) won first prize, for “Fundamental limits of image registration performance: effects of image noise and resolution in CT-guided interventions” (10135-7).

Runners-up were:

  • Ahmet Cakir (Vanderbilt University), et al., for “Evaluation of a high-resolution patient-specific model of the electrically stimulated cochlea” (10135-21)
  • Haichong Zhang (Johns Hopkins University), et al., for “Toward dynamic lumbar punctures guidance based on single element synthetic tracked aperture ultrasound imaging” 10135-18).
Image-Guided Procedures best paper awards

From left, Andrew Wiles, Ahmet Cakir
(accepting on behalf of Xiaochen Yang),
Robert Webster, Baowei Fei

Best Poster Awards sponsored by Northern Digital, Inc., (NDI) also were presented, by Andrew Wiles, NDI Manager of Advanced Research, along with Webster and Fei.

Cum laude winner was Xiaochen Yang (Vanderbilt University), et al., for “Straight trajectory planning for keyhole neurosurgery in sheep with automatic brain structure segmentation” (10135-87).

Runners-up were:

  • Pooneh Roshani (Children’s National Medical Center), et al., for “Temporal bone dissection simulator for training pediatric otolaryngology surgereons” (10135-95)
  • Christina Yan (Queen’s University), et al., for “Study into needle displacement during navigated breast cancer surgery” (10135-79).

 


Join the conversation:
#MedicalImaging

Twitter @SPIEtweets SPIE, the international society for optics and photonics, on LinkedIn Facebook SPIE, the international society for optics and photonics Instagram @SPIEphotonics

SPIE Medical Imaging

11 – 16 February 2017
Orlando, Florida, USA

 


 

See reports, photos, and video
from last year's event
here!