BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
“I did insert my Boodles™ Blue Chip card but the sequence did not download!” an
impatient voice continues in frustration with a call center customer service agent
speaking through an accent-adjusted avatar, “it is almost a week since you received
my sample! I can see the receipt online. I did exactly as it states on your site. I gently
scraped the inner surface of my lower lip with a new tongue cleaner and mailed the
plastic cleaner in sterile tube. How can you say that the sequence is still under review!
My doctor’s appointment in tomorrow and Aleve does not work for my migraines! I
§
want a drug that matches my genotypic profile
and I cannot get that without my
genome sequence! Who is your supervisor?”
§
In this hypothetical (fictional) scenario, it is just after New Year’s Day (2010) and www.YourPersonalGenome.Info
is a service of Personal Genomes Inc (PGI). PGI Chief Scientist, George Church* has reduced DNA sequencing cost
to a penny for a million bases (fictional at present but feasible). Sequencing a human genome is a “commodity”
service. Send a few of your epidermal (skin) cells to PGI and in a few days they will download your sequence (6
billion base pairs) on your global security id (GSI) which is essentially a SIM card. You can insert it in your PDA or
mobile phone. When you activate your phone with GSI password you can select to transmit your personal genome
data for analysis. bMDs Medical Semantic Run Engine uses M-Language based Repository to aggregate relevant
data. Your symptoms and biomedical data are matched with drugs or drug variants that will offer the optimal
efficacy and bioavailability. Recommendations based on your unique pattern of polymorphisms are sent to your
point-of-care physician or nurse practitioner or mobile phone with concomitant authorization for prescriptions.
When you first read about the $10,000 cost for a personal genome sequence you were bit apprehensive but
compared to the chronic agony from unsuitable medication, it seemed quite feasible. Your medical insurance
premium will also be reduced once PGI confirms that your personal genome was sequenced for your personal use.
The secure platform of BioMedical Decision Systems (bMDs) catalogs your sequence data and you are assured that
the quality of medical treatment in the future will be specifically designed for you.
*This information is
real. George Church worked with Walter Gilbert and completed his PhD in 1984 from Harvard University. His
thesis included the first direct genomic sequencing method. The two seminal papers were both published in 1977 in the same
journal but Maxam-Gilbert technique is older (Maxam, A. and Gilbert, W. A new method of sequencing DNA. PNAS 74 560-564)
than the other (Sanger, F., Nicholsen, S. and Coulson, A.R. DNA sequencing with chain-terminating inhibitors. PNAS 74 5463-5467).
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 1 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
BioMedical Decision Systems Initiative
PURPOSE
Healthcare – Extend Access and Improve Quality
VISION
Service oriented platform
♦
with embedded intelligence
♦
linked to distributed dynamic knowledge
♦
to aggregate medical services and diagnostics
1
2
POC Objective
Demonstrate template using biomedical imaging (radiology) as a service and then add electrocardiography (ECG)
POC Outcome
Healthcare providers (users) connect online with experts (radiologists, cardiologists) for diagnosis or 2nd opinion
POC Building Blocks
ƒ
Consortia – build leadership group (credibility, direction, funding)
ƒ
User group and expert panel – build ecosystem of relationships
ƒ
Visualization (users and experts) – digitization technology, hardware, software
ƒ
Data capture (users) – tools and equipment
ƒ
Template (system) – service oriented, secure data, protocol-agnostic edge servers, intuitive interface
POC Success
1
2
3
ƒ
Adoption – nucleate critical mass (scalability as a function of demand)
ƒ
Replication – add electrocardiography (ECG) as a service (scalability as a function of service aggregation)
ƒ
Evaluation – invite critics (systems, content, financial )
ƒ
Dissemination – demonstrate to medical associations, governments, global organizations
ƒ
Expansion – benefit from lessons and work toward the vision including that of a Medical Google equivalent
3
In reverse order of difficulty (most difficult to build true embedded intelligence but least difficult to stitch a canopy of services).
Proof of Concept
Convene independent study headed by an economist (Sir Clive Granger) to determine macroeconomic value of bMDs Initiative
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 2 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
Broad Steps
[-6] Obtain support to establish credibility of BioMedical Decision Systems (bMDs) Initiative
[-5] Obtain commitment to financially support individual(s) to build bMDs Initiative
[-4] Enlist Harvard-MIT Center for Bioimaging
4
as a partner
[-3] Commitment from founding sponsor to support project plan in order to proceed with bMDs Initiative
[-2] Create a mechanism to situate bMDs Initiative (for example, MIT- Harvard Medical School joint context)
[-1] Explore Satellife to map overlaps / exchange ideas
[00] Commence Phase 1 of bMDs Initiative with founding group
[01] Create proof of concept targeted content advisory (bioimaging and cardiology)
[02] Organize experts panel (for example, liaise with Harvard-MIT group and UCSF-Stanford group)
[03] Convene tools and technology working group to sketch
[04] Liaise with select potential sponsors
6
5
platform architecture / components / interface
and global knowledge groups
7
(discuss content and technology)
[05] Complete beta “platform” and run field test with experts panel (improve/upgrade/modify)
[06] Create user affiliates (for example, in China, India, Taiwan, Turkey, UK)
[07] Commence “global go live”
[08] Add ECG functionality
[09] Scale users and experts
[10] Complete Phase 1 (pilot). Ramp dissemination to government, medical associations and global organizations
4
www.nmr.mgh.harvard.edu/martinos/aboutUs/index.php
5
http://ieeexplore.ieee.org/Xplore/login.jsp?url=/iel5/8532/26947/01197568.pdf
6
IBM, Microsoft, General Electric (USA); Olympus, Hitachi (Japan); Phillips (NL); SAP, Siemens (DE); Chi-Mei (ROC)
7
http://www.cinj.org
http://www.visu.uwlax.edu/NGI/RSV.html
http://summit.stanford.edu/
http://www.uphs.upenn.edu/path
http://surgery.uc.edu/csi
http://www.cis.rit.edu/content/view/102/166/
http://www.ee.surrey.ac.uk/CVSSP/index.php
http://www.miralab.unige.ch
http://www.meduniwien.ac.at/maxillo-facial/bio.htm
http://www.nei.nih.gov
http://www.landsteiner.org/
http://www.umbc.edu/rssipl/research_publication/cad.php
http://www.nlm.nih.gov/research/visible/animations.html
http://archive.nlm.nih.gov/proj/bita/trans-pacific.php
Istituto di Radiologia, Universita Cattolica del S. Cuore, Policlinico A. Gemelli, Roma, Italy
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 3 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
DESCRIPTION of BIOMEDICAL DECISION SYSTEMS INITIATIVE
BACKGROUND
Scenarios
1
that combine remote diagnostics with that of SATELLIFE
2
form a key part of the thinking that may
germinate the BioMedical Decision Systems (bMDs) Initiative. Medical diagnostics and dissemination of medical
information (“the last mile”) through SATELLIFE may be viewed as a “data-dependent decision system” and “rulebook or workflow” for decision making, respectively. If the data and the rule can functionally converge, without
human intervention, it may be the foundation for an “intelligent” decision system. Hence, in the vein of artificial
intelligence (AI), platforms with embedded knowledge (intelligence) can analyse
3
data flow and render decisions.
The above idea exists in many flavours (http://esd.mit.edu/WPS/esd-wp-2006-10.pdf). Various biomedical decision
systems and services
4
operational but often in narrowly defined fields. The medium of the internet could (but has
not) enable the advances in these systems (albeit in narrowly defined fields) to benefit those who may need such
services. The scenario is exacerbated because these defined systems (data, information, decision support silos)
may not be interoperable and even among related fields it may be difficult to exchange information or knowledge.
For such “system of systems” to be functional and beneficial, we need platforms that are agnostic to proprietary
protocols but aggregates these systems (SoS) in a service oriented architecture (SOA). Distributed architecture
enables these systems to be accessed (as services) by those who need it (at the point of care or at the “edge” of
processes). In the real world, access to such services will be payment dependent but at present even for-payment
services may not be available because the platform for aggregation is still immature, at best. The business world
and the medical industry may view such platform(s) to emerge as a bridge between the haves and the have-nots.
It is unclear whether a single business or academic group can nucleate such multi-disciplinary approach necessary
to have a functional system.
PROPOSAL
BioMedical Decision Systems (bMDs) Initiative is a proposal to create such a platform. It is unlikely to be unique
or revolutionary and certainly not an invention but an innovative aggregation
6
5
to catalyse convergence in order to
extend healthcare access and improve the quality of healthcare through intelligent decision support systems.
As a proof of concept for the “vision” embedded in this endeavour, execution of a pilot for the proposed bMDs
platform is, therefore, only a slice of the idea that may be transformed to reality. Focus first on the “narrow field”
of bio-imaging followed by electrocardiography (EKG) to demonstrate functional scalability of the platform template.
7
The joint MIT-Harvard effort (Athinoula A. Martinos Center for Biomedical Imaging) and many other endeavours
8
along those lines may help in this effort by contributing their professional tools and expertise . Bio-imaging is
9
probably easier in terms of data digitization and remote visualization tools are well published . With respect to
cardiology, the advice of Dr Bernard Lown
10
(Harvard University) may be helpful.
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 4 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
POTENTIAL OPERATION & PROSPECTS
In actual operational terms, this model demonstration may be, in its basic form, an xray image from Uganda or
India that can be viewed by an expert in US or EU and feedback sent to the local healthcare provider. In essence it
mimics SATELLIFE, in this phase of the bMDs Initiative. Remote biomedical imaging and biotelematics
new
12
but research groups
13
11
is nothing
are generally engaged in stand-alone point-to-point systems for special purposes
14
.
This platform is expected to have an open format where any image (data) may be uploaded and viewed (analysed)
by expert(s) who will share their observations. Tools of visualization will be important and significant advances
15
may be resourced. In terms of software, this may be of “publish and subscribe” modus operandi. For medical data
the issue of security and privacy must be considered.
In the AI vision of autonomic computing, this platform may evolve to possess embedded intelligence which may
instantly (or in real-time) generate opinions or recommendations or diagnosis or referrals (exception management)
without active human intervention. The latter should be welcome news to the aging population in Europe and Japan
due to its potential to reduce basic healthcare cost. Hence, affluent nations may view this as medical automation to
reduce cost while this bMDs may enable increased access to healthcare for developing nations.
In stages, actual medical decision systems will depend upon adoption of medical ontologies and semantics which
will make the knowledge base machine-readable for such platforms to spew decisions. But, first, platforms must be
created that are intuitive and, for now, can benefit from expert-dependent decision support. These platforms and
its connected ecosystem must gain diffusion (adoption by local hospitals, government healthcare agencies, global
organizations). Establishment of an open platform, even if it mimics an existing service, must be created with the
purpose to serve as a template to add and aggregate services, which may, by being on the “same page” could
communicate with each other or autonomically offer recommendations (alerts). AI agents may share information
and communicate in a system of systems (SoS) environment to improve the holistic view necessary for treating the
patient, not merely the symptom. The bMDs Initiative offers such potential and may eventually help diagnostic
tools and expertise to reach the less fortunate people in the world.
BUSINESS MODEL & FUNDING
Altruism alone is unlikely to provide the resources necessary to sustain this effort. Incentives are necessary to
gather the momentum. This platform offers commercial potential and may be an emerging source of revenue. The
obvious “low hanging fruit” may be the bi-directional revenue stream created by (1) demand for expert opinion
(from experts in the West) from people who are able to pay (patients from the East) and (2) outsourcing of
analysis to low-cost professionals (radiologists, pathologists) from the West (US, EU) to the East (India, China).
NIH/NSF type grants
16
are steeped in quagmire and may not be the preferred source for support. bMDs Initiative
may develop a consortium of intellectual resources and invite corporate thought leaders in software/hardware with
systems expertise (IBM, Microsoft, SAP) and manufacturers of innovative biomedical tools (GE, Siemens, Olympus,
Phillips). The business development opportunities for the corporate sponsors are obvious and the platform may
unleash new channels to increase their potential for sales and services (for example, IBM for software/hardware,
Olympus
17
for remote visualization tool, GE for bioimaging
18
).
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 5 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
CAVEAT
Without tools in the hands of the user community (the last mile), the bMDs platform will be impotent irrespective of
the caliber of its experts or embedded intelligence for analytical decision support. Hence, the bMDs Initiative may
act as a catalyst for users to invest in the tools and technologies in order to extend the reach of healthcare services.
For affluent communities where access is not a problem, these tools may enable them to receive improved quality
of healthcare by taking advantage of the panel of global experts. However, organization of medical information for
effective use by point-of-care physicians still suffer from disarray. The need for a “Google MD” or medical Google in
conjunction with MOU or Medical Open University may partially bridge the chasm.
MEDICAL GOOGLE ?
The synthesis and presentation of a proper medical syllabus with didactic and interactive arms may serve as a
Medical Google equivalent. There is a desperate cry for a credible platform and an unlikely shortage of experts. The
strength of this vision could bring them together on a web based platform and given the right infrastructure it may
grow on its own with modest regulatory oversight, as in Wikipedia. For example, problems with joints are common
with an aging population. Rheumatology may be old hat and scores of websites exist but it still needs an umbrella.
Medical students as well as established practitioners may be able to access this from anywhere in the world, get an
expert opinion and better help patients, with a focus on medical information arbitrage and interrelationships.
Joints:
Examples:
Anatomy
Arthroscopy of normal joints
Physiology
Relationship between structure and function
Immunology
Pathology
Post prepared lectures, podcasts
Medicine
Organized by complexity (nurses, physicians, patients)
Therapeutics
Slide Catalogue with links/tutorials
Epidemiology
Imaging
Critique of key advances
Surgery
Guidelines from international societies
Best Practice
Health education for young people
Evidence-based Practice
Translation Services
ACADEMIC ANCHOR
The Initiative is expected to be guided by individuals without personal commercial interests. The bMDs consortium
may be based in an academic or non-profit foundation to galvanize participation of an extended global ecosystem
necessary to build an organization that can help build the platform for the BioMedical Decision Systems Initiative. It
is necessary in the self-interest of an affluent aging
19
Western society but offers hope to the less fortunate, as well.
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 6 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
1
2
http://www.devicelink.com/ivdt/archive/01/01/005.html
http://www.healthnet.org/index.php
3
Developing System for Remote Clinical Evaluation of Computed Tomography and B-Mode Ultrasound Images
D. Smutek (Czech Republic, Japan), A. Shimizu (Japan), L. Tesar (Japan, Czech Republic), H. Kobatake, S. Nawano
(Japan), and P. Maruna (Czech Republic)
Keywords:
Image processing, Computed tomography, Ultrasound imaging, Liver lesions, Thyroid disorders
Abstract:
A computed-aided diagnostic system (CAD) for remote automatic classification of CT and B-mode ultrasound
images is described. The system is capable of discriminating among focal liver lesions and chronic thyroid
inflammatory diseases. The application is based on texture analysis. Two types of texture features are used in the
system: 22 first-order features computed from the original gray levels and four different gray-level transformations
of an image and 108 second-order features (computed from co-occurrence matrices) which capture the spatial
organization of texture primitives. The classification of images is performed by network of Bayes classifiers with
majority voting. The classifier was trained by Gaussian mixture model method and classifies images according to
their texture feature values. In testing phase the system achieved 100% classification success rate when using four
principal descriptive features. The results are sufficiently consistent under small changes in image tool setting and
scan type. The implementation of the remote CAD system is promising for automatic classification.
4
www.stakes.fi/verkkojulk/Raportteja6/2006 STAKES Center of Excellence for ICT
Hamalainen, P., Tenhunsen, E., Hypponen, H. And Pajukoski, M. (2005) Experiences on Implementation of the Act
on Experiments with Seamless Service Chains in Social Welfare and Health Care Services. National Research and
Development Center for Welfare and Health, Finland (Helsinki, 2005)
5
www.Internet2.edu (Internet2 health science applications in remote visualization)
6
In the commercial world such aggregation of various functions on a common platform (circa 1980’s) gave birth to
ERP or enterprise resource planning systems that contained under “one roof” all the “departments” of business
(finance, manufacturing, supply, distribution, human resource, procurement, etc).
7
www.nmr.mgh.harvard.edu/martinos/aboutUs/index.php
8
www.crd.ge.com/esl/cgsp/projects/medical/
Three dimensional medical reconstruction
This is a collection of movie clips, showing various medical reconstructions. These images are derived from slice
data from a variety of medical image modalities such as MR or CT. The two dimensional slice data from these
scanners is used as input for the three dimensional reconstructions.The data starts out as slices(images) taken at
regular intervals throughout a portion of the body. The slices are first segmented to separate the various tissues.
An algorithm developed at GE R&D known as marching cubes is then used to create a three dimensional
representation of the structures. Once the 3D model is created, an animation package called LYMB is used to
provide the visualization and animation.
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 7 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
9
Goshtasby, A.A. (2005) 2-D & 3-D Image Registration for Medical, Remote Sensing and Industrial Applications
http://eu.wiley.com/WileyCDA/WileyTitle/productCd-0471649546,subjectCd-BE70.html
10
The scene is familiar to anyone who has seen a TV medical drama. The old man on the gurney goes into cardiac
arrest, his heart monitor emitting an urgent whine. A doctor grabs a paddle in each hand, barks a warning (Clear!)
and applies a jolt of electricity to the patient's bare chest. The body thrashes violently, but a tense moment later,
the monitor resumes its steady beeping. That this procedure is instantly recognizable to people who've never set
foot in an emergency room is largely due to Boston physician Bernard Lown, inventor of the modern
defibrillator and Nobel Prize winner. It wasn't always clear that passing current through a patient's body could
restore a wayward heart. In 1775, a Danish veterinarian used electricity to stun and revive a chicken. It wasn't
until 1955, though, that physician Paul Zoll resuscitated a human patient by applying a burst of alternating current
to his chest. AC defibrillation didn't have a high success rate, but since its recipients were nearly dead anyway, its
drawbacks did not receive much scrutiny.
That began to change in 1959, when 37-year-old Lown faced a desperate situation. A patient arrived at the
emergency room with short breath and a rapid pulse. When the customary drugs failed to slow the man's racing
heart, Lown recalled Zoll's work. Although defibrillation had never been attempted in this type of case, Lown
obtained permission from the man's wife and delivered a shock to his chest. To Lown's relief, the treatment worked.
But three weeks later, when the man returned with the same problem, Lown's attempt to shock the heart back to
normal made the muscle contract erratically instead. Doctors opened the man's chest to apply electrodes directly
to the heart. The patient survived the night but died soon after.
Lown spent a year trying to find out what had gone wrong. He finally realized that the alternating current doctors
were using did enormous damage to the heart muscle. Lown enlisted engineer Baruch Berkowitz to help find a
safer, more effective treatment. The two developed a defibrillator based on direct current, which delivers a single
pulse of electricity instead of a current that repeatedly switches direction. Within a few years, this machine had
replaced its AC predecessor in hospitals, and it has been saving lives, in real life. (MIT Tech Review, Nov 2004)
11
BIOTELEMATICS
ICOMS Vienna 2005: Congress for Oral and Maxillofacial Surgery
Telenavigation with the Vodafone Mobile Connect Card from A1
Within the framework of this congress on 2 September 2005, Dr. Michael Truppe from Karl Landsteiner Institute for
Biotelematics, one of the pioneers of computer-assisted surgery, demonstrates an exciting innovation: for the first
time ever it is possible to interact with an live operation by means of a Vodafone Mobile Connect Card via Laptop.
Any number of these so-called telenavigation clients can simultaneously and independent of one another have live
interactive access – in a visual quality that has previously been unknown – of critical phases of operations. A
normal Vodafone Mobile Connect Card from A1 makes an Internet connection with up to 384 kBit/sec, and doctors
and students can take part in operations across geographic distances, within the scope of their training, as well as
in teleconsultations for particularly complex cases.
For Telekom Austria and mobilkom austria involvement at the ICOMS is a good opportunity to prove their
technology leadership in telemedicine applications in the wireline as well as in the wireless segment, at a highly
regarded forum. The Telekom Austria Group wants to position itself more strongly in the future as a telemedicine
technology partner, particularly against the background of e-Health as a focus of the European "i2010" strategy.
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 8 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
12
Lasers Surg Med.1988;8(1):1-9
Remote biomedical spectroscopic imaging of human artery wall
Hoyt CC, Richards-Kortum RR, Costello B, Sacks BA, Kittrell C, Ratliff NB, Kramer JR, Feld MS.
George R. Harrison Spectroscopy Laboratory, Massachusetts Institute of Technology, Cambridge 02139.
We discuss a technique, laser spectroscopic imaging (LSI), remote acquisition of spectroscopic images of biological
tissues and tissue conditions. The technique employs laser-induced spectroscopic signals, collected and transmitted
via an array of optical fibers, to produce discrete pixels of information from which a map or image of a desired
tissue characteristic is constructed. We describe a LSI catheter that produces spectral images of the interior of
human arteries for diagnosis of atherosclerosis. The diagnostic is based on the fact that normal artery wall and
atherosclerotic plaque exhibit distinct fluorescence spectra in the 500-650 nm range when excited by 476-nm laser
light; the fluorescence from blood is minimal. The catheter is composed of 19 optical fibers enclosed in a
transparent, protective shield. Argon ion laser radiation is used for excitation, and an optical multichannel spectral
analyzer is used for detection. Sequential sampling is used to minimize crosstalk among fibers and reduce blurring
of the image. Computer-processed 19-pixel spectroscopic images are produced of fresh cadaver artery in vitro.
Regions of normal tissue, plaque, and blood are identified. Diagnoses are confirmed histologically and by direct
spatial correlation. The results demonstrate the concept of using this laser catheter system for real-time imaging.
13
www.landsteiner.org
www.miralab.unige.ch
www.cis.rit.edu/content/view/102/166/;
www.ee.surrey.ac.uk/CVSSP/index.php
14
http://www.meduniwien.ac.at/maxillo-facial/bio.htm
http://archive.nlm.nih.gov/proj/bita/trans-pacific.php
http://www.umbc.edu/rssipl/research_publication/cad.php
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 9 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
15
http://www.nlm.nih.gov/research/visible/animations.html
The Visible Human Project
eMedTool- "dissection" and rotation of body regions and systems, from Merck Medicus.
Virtual Reality in medicine Fly- through examples from the Female dataset appearing in "Virtual Reality in
medicine", BMJ 1999, 319: 1305 (13 November) by Szekely and Satava.
Clips from the Visible Human Project (Fly throughs of crysection generated views from the University of
Portsmouth Centre for Radiology).
Human Tongue Atlas (grant from the National Institute of Deafness and other Communicaation Disorders).
Marching Through the Visible Man (paper by Bill Lorensen of the GE Imaging & Visualization Lab).
Marching Through the Visible Woman. (paper by Bill Lorensen of the GE Imaging & Visualization Lab).
3D Virtual Colonoscopy (static images and fly-through animations from the Departments of Radiology and
Computer Science at SUNY Stony Brook).
Interactive Knee Program (University of Pennsylvania Medical Center).
Visible Human Research (Arctic Region Supercomputing Center; includes renderings and animations based
upon CT and RGB data from both the male and female datasets).
A Guided Tour of the Visible Human (Washington University Medical School - The MAD Scientist Network).
The Virtual Human (Argonne National Laboratory and the University of Chicago).
The Vesalius Project (Creating a Computer-Based Anatomy Curriculum - Columbia University).
MEET Man Project (Models for Simulation of Electromagnetic, Elastomechanic and Thermic Behaviour of
Man - University of Karlsruhe)
Computer Graphics and Medicine (MIRALab - University of Geneva)
MPIRE (Massively Parallel Interactive Rendering Environment - an interactive, distributed, direct volume
rendering system for Cray T3D, Cray T3E, or SGI workstations, from the San Diego Supercomputer
Center).
Voxel- Man Gallery ( images and animations of the torso from the University of Hamburg).
Award winning 3D browser for the head (Tom Conlin of the University of Oregon - requires a Java 1.1
enabled Web client).
U-SCALE Computer Aided Learning Environment for Human Gross Anatomy Project (DEMO from the
University of Saskatchewan - Netscape or Internet Explorer 4.x: Java and javascript enabled, and
accepting cookies).
The Virtual Anatomy Explorer (3D VRML browsing prototypes from the University Clinic Giessen, Germany
- Cosmo player recommended for VRML viewing).
The Visible Human Project-Reduced Data Sets (Normal [Fresh] CT scan data sets for both the male and
female in reduced resolution - the University of Wisconsin).
3D Anatomy for Medical Students (interactive VRML models, with contol for structure displayed,
transparency, labels, embedded images, etc. - McGill University).
daVinci (prototype simulator for performing vascular catherterization and interventional radiology
procedures - Kent Ridge Digital Labs, Singapore).
Animating the Visible Human Data Set (Rutgers University, VIZLAB).
Peel-away Visible Human Movies for the Palm Pilot (Medical Multimedia Systems).
Football player animations (Anatomical Travelogue).
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 10 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
16
NON-INVASIVE IMAGING FOR DIABETIC RETINOPATHY
RELEASE DATE: July 12, 2004 (see addendum NOT-EB-04-004)
RFA Number: RFA-EY-04-001
EXPIRATION DATE: February 15, 2005
Department of Health and Human Services (DHHS)
PARTICIPATING ORGANIZATION:
National Institutes of Health (http://www.nih.gov)
COMPONENT OF PARTICIPATING ORGANIZATION:
National Eye Institute (NEI) (http://www.nei.nih.gov)
CATALOG OF FEDERAL DOMESTIC ASSISTANCE NUMBER: 93.867
APPLICATION RECEIPT DATE: February 14, 2005
THIS RFA CONTAINS THE FOLLOWING INFORMATION
o
o
o
o
o
o
o
o
o
o
o
o
Purpose of this RFA
Research Objectives
Mechanism(s) of Support
Funds Available
Eligible Institutions
Individuals Eligible to Become Principal Investigators
Where to Send Inquiries
Submitting an Application
Peer Review Process
Review Criteria
Receipt and Review Schedule
Award Criteria
PURPOSE OF THIS RFA
Diabetes is a prototypical chronic disease that imposes a large public health
burden. One of its many complications is diabetic retinopathy (DR). Early
detection of DR is an important problem because many Americans with diabetes
do not get regular eye exams due mainly to distance, geography, and economic
status. Studies conducted in the 1980’s and 1990’s demonstrated that vision
loss from DR could be prevented if treatment with laser photocoagulation
therapy was initiated based upon regular screening and subsequent diagnosis
of retinal disease. In this Request for Applications (RFA) the NEI seeks to
develop, apply, and evaluate noninvasive technology that is practical,
affordable, and accessible so that patients with diabetes can benefit from
remote site disease screening.
RESEARCH OBJECTIVES
Nature of the Research Problem
Early detection is critical for the prevention and treatment of the blinding
complications of retinopathy in persons with diabetes. Evidence suggests
that the retina begins to malfunction very early in diabetes, and that the
sooner treatment is initiated the greater the chance for medical benefit.
Ophthalmologists started treating diabetic retinopathy with lasers in the
1960s in an attempt to seal leaking blood vessels and limit neovascular
growth. While laser treatment is effective, unfortunately about half of
diabetic patients do not receive this treatment because they do not have the
recommended regular dilated eye examination. Currently, new retinal imaging
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 11 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
technologies are being developed which have potential utility in remote-site
clinics. If interfaced with the internet, these technologies could allow
primary health care providers to identify patients with retinopathy. There
are at least two barriers however to effective screening: the cost of the
noninvasive instrument and the ease and simplicity of its use. Noninvasive
technologies coupled with internet-based telemedicine communication platforms
might facilitate remote expert clinical diagnosis and subsequent treatment
recommendations, thus increasing access to medical care.
Background Information
Noninvasive assessment of retinal function in patents with diabetes has
traditionally been done using fluorescein angiography, which is especially
valuable in assessing advanced stages of retinal neovascularization. This
procedure allows an ophthalmologist to determine the course of further
treatment. Retinal fundus photography is another technique to assess retinal
status so that the clinician can visualize the microvasculature features of
DR. However, complete examination with fluorescein angiography and/or fundus
photography requires the patient, the photographer, and the retinal
specialist to be physically present in the clinic. Although this may be the
best way to examine an individual, it may not be the best approach for mass
screening for early detection of DR in diabetics. Newer noninvasive
technologies are currently being developed. These include functional
magnetic resonance imaging (fMRI), laser Doppler ultrasound, electron spin
resonance spectroscopy (ESR), and fluorescence spectrometry. These
technologies are expensive, and diagnostic improvements are only one aspect
of the problem of how to deliver the benefits of early detection to the
widest possible range of Americans with diabetes. The screening data needs
to be interpreted, and treatment recommendations made. With the increased
ability to digitize images and transfer data over the internet it is possible
for a retinal specialist at a major medical center to diagnose disease with
the patient at a remote location. This could allow screening of persons with
diabetes by a qualified specialist, thus bringing the benefits of advanced
surface feature and tissue imaging and high bandwidth Internet platforms to
remote clinical settings.
Scientific Knowledge to be Achieved
Proliferative DR involves the formation of new blood vessels that develop
from the retinal blood supply. Much recent research points to the
involvement of several factors in the etiology of DR, and several biological
therapies are either in or close to entering clinical trials. But no drug
treatment for DR is currently available. While laser photocoagulation and
vitrectomy for proliferative retinal disease have been shown to be effective
by large-scale, randomized, controlled clinical trials, these treatments
require clinical observation of overt retinopathy. Noninvasive techniques to
enhance eye health care delivery at remote sites could add to the current
knowledge base of clinical diagnosis and treatment and help increase our
understanding of pathogenic mechanisms. As therapies are developed, these
could be made available to a larger number of patients if individuals with
the disease were effectively screened and then diagnosed by an expert.
Objectives
The need to develop human biomedical imaging techniques has been identified at
several recent workshops and conferences on diabetic complications. For
example, a March 30-31, 1999, workshop entitled "Biomedical Imaging Symposium:
Visualizing the Future of Biology and Medicine" which was coordinated by the
NIH Bioengineering Consortium (BECON), and addressed three scientific areas:
(1) imaging at the cellular and molecular levels for early detection of
disease; (2) imaging for the clinical diagnosis, staging and recurrence of
disease; and (3) imaging applied to therapeutic applications and monitoring
for various disease processes. This symposium also emphasized the need to
support fundamental methodological development of imaging techniques before
disease-oriented and organ-specific applications are determined.
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 12 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
A conference entitled “JDRF-NASA Diabetic Retinopathy Research Workshop” in
1999, sponsored in part by the NEI; the National Institute of Diabetes and
Digestive and Kidney Diseases; and the National Institute of Neurological
Disorders and Stroke, reviewed current technologies for assessing retinal
function as well as new approaches to therapy. A 2001 workshop on “Screening
and Eye Exams for Diabetic Retinopathy” addressed the need for screening and
eye exams for retinopathy and the way new technologies and telemedicine could
enhance patient access to eye exams.
This RFA is designed to encourage submission of research grant applications
that create, utilize, and evaluate noninvasive retinal imaging technologies
that when joined with telemedicine provide retinopathy screening and eye exams
at remote locations. At present only about 50% of all diabetic patients are
involved in an appropriate eye care program. Telemedicine would allow eye
imaging at the point of care, eliminating the need for a separate screening
appointment. Images can be captured and transmitted to a central resource for
accurate medical assessment with diagnostic information and recommendations
transferred back to the provider at the point of contact. Telemedicine
technology makes it possible to remove the barriers of distance, time,
geography, and economics and bring services to the patients rather than the
patients to the services. Telemedicine is currently being tested and utilized
in a number of state programs, including the California Indian health program.
Most current noninvasive technologies fall into several broad categories:
retinal imaging, neurological function, and vascular function. For retinal
imaging, current technologies include the Retinal Thickness Analyzer (RTA)
which produces three-dimensional maps of the retina. A similar approach is
used in optical coherence tomography (OCT) to obtain a high-resolution cross
sectional map of the retina. Spectral imaging of the retina can reveal
tissue composition using chemical and spectroscopic signatures. Spectroscopy
can potentially map and/or classify clinical features such as perfusion,
capillary dropout, hemorrhage, and retinal metabolism (by mapping cytochrome
oxidative state). Although potentially limited by spatial resolution, fMRI
has the potential to be clinically powerful in measuring blood flow and
oxygen content. Sensitivity to inner retinal neuronal dysfunction is
important for detection and study of early local changes in the retinas of
diabetics and retinal neural circuitry can be assessed with the multifocal
electroretinogram that tests the retina’s adaptive properties based on
multifocal and geographically precise flashes. Certain aspects of vascular
structure and function can be studied noninvasively by fundus photography but
the currently available cameras are bulky and require pharmacological
dilation of the pupil for optimal image quality. New advances in optical
technology that could theoretically produce a sharper image include adaptive
optics, ballistic imaging, and scanning laser ophthalmoscope.
These advancements can benefit diabetic patients by improving the technology
needed to assess the functional status of the retina. But they are currently
expensive and have not been adapted for low cost use as screening
technologies. To be useful the application of the technology needs clinical
and technological validation before it can be adopted for clinic application.
Project Organization
The NEI anticipates that applications will likely require the involvement of
engineers, physicists, computer specialists, and bio-scientists knowledgeable
in interfacing technology with disease applications.
MECHANISM OF SUPPORT
This RFA will use the NIH Research Project (R01) award mechanism. As an
applicant you will be solely responsible for planning, directing, and
executing the proposed project. This RFA is a one-time solicitation. Future
unsolicited, competing-continuation applications based on this project will
compete with all investigator-initiated applications and will be reviewed
according to the customary peer review procedures. The anticipated award
date is December 1, 2005.
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 13 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
17
www.olympusmicroimaging.com/index.cfm/page/products.index.cfm/cid/492/navid/180/parentid/1
18
GE Healthcare - Medical Imaging and Information Technologies, Medical Diagnostics and Patient Monitoring
Systems
GE Healthcare provides transformational medical technologies that will shape a new age of patient care. GE
Healthcare's expertise in medical imaging and information technologies, medical diagnostics, patient monitoring
systems, disease research, drug discovery and biopharmaceuticals is dedicated to detecting disease earlier and
tailoring treatment for individual patients. GE Healthcare offers a broad range of services to improve productivity in
healthcare and enable healthcare providers to better diagnose, treat and manage patients with conditions such as
cancer, Alzheimer's and cardiovascular diseases.
VIVID 7 DIMENSION DELIVERS INTEGRATED REAL-TIME 4D AND COMPLETE MULTI-DIMENSIONAL IMAGING
CAPABILITIES
GE Healthcare's new Vivid 7 Dimension cardiovascular ultrasound system is the world's first system to give
clinicians the first fully integrated real-time 4D and multi-dimensional imaging capabilities. The Vivid 7 Dimension,
a latest member of the Vivid 7 family, provides clinicians with new tools for detecting and diagnosing heart
conditions such as heart failure in patients.
Vivid 7 Dimension brings new echocardiography acquisition, reconstruction and analysis techniques that allow
physicians to view images in multiple planes simultaneously. As a result, physicians can more precisely detect and
diagnose pathology that they may not have been able to see before while improving clinical productivity. With its
increased capabilities, Vivid 7 Dimension can help physicians better manage their cardiovascular patients and
enhance their diagnostic confidence.
MULTI-DIMENSIONAL IMAGING AND 4D TISSUE SYNCHRONIZATION IMAGING (TSI)
Facilitating efficient communications of findings, physicians benefit from user-friendly formats and functionality,
while increasing productivity in all modes of the standard adult echo exam.
Vivid 7 Dimension system delivers cutting edge performance by reaching new levels of image quality, contrast
functionality and new clinical tools including:
•
•
•
•
Multi-dimensional imaging - acquisition technology that acquires bi-plane and tri-plane images with colour
and doppler to significantly reduce exam times and simultaneously view more myocardial segments for
both echo exams and stress echo exams
4D Imaging - real-time display of 3D images of the heart to provide more information on cardiac function
and better communicate of the heart's structure and function
4D Tissue Synchronization Imaging (TSI) – a dynamic 4D parametric imaging model of dyssynchrony from
a single heartbeat to better understands and communicates a patients heart condition. Additional
reporting format of a "bull's eye report" to communicate dyssynchrony in a format more easily understood
by the EP physician
Bloodflow Imaging (BFI) – vascular imaging mode that gives clinicians a better understanding and
delineation of flow in vessels
HIGH PERFORMANCE MEETS MINIATURIZATION WITH GE'S NEW VIVID I CARDIOVASCULAR ULTRASOUND
SYSTEM
GE Healthcare's new Vivid i is the world's first miniaturized cardiovascular ultrasound system to provide highperformance, full-featured imaging in a lightweight design.
Vivid i address one of the biggest challenges in cardiovascular care - access to complete, real-time diagnostic
information. The new system expands the reach of echocardiography by offering all the functionality and high
performance of full-featured premium scale systems - but in a completely portable and wireless design that weighs
30 times less.
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 14 of 15
BioMedical Decision Systems (bMDs) Initiative: Suggested by Dr Shoumen Palit Austin Datta, Massachusetts Institute of Technology
VIVID I: BREAKTHROUGH TECHNOLOGY
Within GE Healthcare, Vivid i is considered an "Imagination Breakthrough" product - a major new invention with
significant clinical and productivity value for clinicians and patients, and of great importance to GE's growth.
Vivid i leverages GE's research and development investment in the areas of software-based platforms, hardware
miniaturization and connectivity in a new, patented design. The new product is based on the company's TruScan
architecture and features a software-driven PC-based platform, raw data storage with new post-processing
capabilities, complete connectivity and compatibility with the entire GE family of cardiovascular ultrasound systems.
VOLUME ULTRASOUND - THE REVOLUTIONARY NEW APPROACH TO THE ULTRASOUND EXAMINATION
We are proud to be part of an exciting new step in the development of ultrasound imaging. Volume Ultrasound
means the only ultrasound unit providing virtually all necessary tools to perform a complete ultrasound "check-up"
of the entire ROI (region of interest) from one single volume sweep. Volume Ultrasound includes:
•
•
Hardware - based on a very high computer power and memory capacity, as well as a complete range
micro 4D transducers dedicated to all applications, from abdomen, small parts to endocavitary and
pediatrics
Software - developed for real-time and post processing analysis of the entire ROI. It offers tools enabling
you to virtually "extract" from the ROI the information you need (surface abnormalities, fetal bone
structures, vascularization of organs, flow information, fetal heart advanced analysis and much more)
GE Healthcare is a $14 billion unit of General Electric Company (NYSE: GE) that is headquartered in the United
Kingdom. Worldwide, GE Healthcare employs more than 42,500 people committed to serving healthcare
professionals and their patients in more than 100 countries.
19
An Aging Western Society
% 25
COMPARE
Fertility rate 2000
China 1.5
India 3.0
Median Age 2015
China ~ 45
India ~ 25
Population 2025
1.5 billion each
EU <19
EU >65
US >65
% 20
% 15
% 10
European Scenario
»
»
»
»
% 5
0
1950
Europe
Population
(millions)
2000
Labour shortage
65+ prefer own homes
65+ are ICT users
Healthcare & Pension
2050
Under 19 year
(millions)
Over 65 year
(millions)
Over 65 year
% population
1950
350
70
14
4.0
2000
450
60
40
8.9
2050
400
40
90
22.5
bMDs Proposal by Dr Shoumen Palit Austin Datta, MIT <shoumen@mit.edu)>
Page 15 of 15