CIMIT FORUM AGENDA
September 11, 2007
4:00 – 6:00 PM
4:00PM
Application of Cardiac Non-Invasive Imaging in Clinical Decision-Making
Speaker: Patrick
O'Gara,
MD, Director
of Clinical Cardiology and Vice-Chairman of Medicine for Clinical Affairs at Brigham
and Women’s Hospital, and Associate Professor of Medicine at Harvard Medical
School, pogara@partners.org
Moderator: Ahmed Tawakol,
MD, Co-Director, Cardiac MR-PET-CT Program and Associate Director, Nuclear
Cardiology, MGH; Co-Leader, CIMIT Cardiovascular Disease Program,
atawakol@partners.org
The cardiologist’s chief task is often not to make a
diagnosis but to weigh the risks of a heart condition and to determine how
aggressively it should be treated. Over
the past decades, imaging techniques have proved vital to cardiologists seeking
to make these decisions. As expensive
new imaging methods proliferate, however, it has become important that
physicians understand when each is warranted.
Dr. Patrick O’Gara, a noted cardiologist from Brigham and Women’s
Hospital, suggests that imaging techniques should be cost-effective and safe
and that they should provide results that are visually familiar, accurate, and
reproducible.
The cost of imaging has become a major concern for both
clinicians and insurers. In a few cases,
private practitioners have been suspected of recommending unnecessary imaging
to increase revenue. In more cases,
patient care has probably suffered due to the lack of clear guidelines
specifying when certain imaging techniques are and are not appropriate.
The question of how best to reap the benefits of imaging
technology will only become more pronounced as new procedures are
developed. One area of active research
involves using molecular biomarkers to shed light on a patient’s cardiac
health. Such techniques may soon be used
to detect early signs of ischemia and to identify dangerous arterial plaques. These methods may someday allow cardiologists
to diagnose heart problems before they clinically manifest themselves.
Cardiologists are slowly moving toward a more personalized
approach to medicine. In addition to
clinical data, scientists are beginning to use gene expression profiles and
proteomics to assess risk. With new
techniques, however, come new challenges.
Costs can be high, and care providers sometimes lack the training to
correctly interpret results. If the
potential of modern imaging is to be fully realized, the training that
cardiologists receive will have to be altered.
Click
here to view this presentation
5:00PM
Modeling of Optimal ICD Electrode Placement in Children and Adults
Speakers: Matthew Jolley, MD, Assistant
in Medicine, Children's Hospital; Instructor in Pediatrics, Harvard Medical
School, matthew.jolley@cardio.chboston.org; and Jeroen Stinstra,
PhD, Technical Manager/Research Associate, Center of Integrative
Biomedical Computing, Scientific Computing and
Imaging Institute, University of Utah, jeroen@cvrti.utah.edu
Moderator: Ron Kikinis, MD, Director, Surgical
Planning Laboratory of the Department of Radiology, BWH and Harvard Medical
School; Professor of Radiology, Harvard Medical School; Co-Program Leader,
Image Guided Therapy, CIMIT, kikinis@bwh.harvard.edu
The number
of cases in which cardiac defibrillators are implanted in children is increasing,
but the procedure is still performed in a largely ad hoc manner. Compared to similar procedures in
octogenarians, implanting defibrillators in kids is difficult. Children are small and will grow over time,
and their cardiac anatomy differs from that of adults. Hoping to improve the efficacy of implanted
defibrillators in children, Matthew Jolley, MD, and Jeroen Stinstra, PhD, have created a modeling system capable of mapping an
individual’s chest and then determining the optimal position for an implanted
cardiac defibrillator.
To model an individual’s thorax, the
software uses pre-existing surgical planning applications such as 3Dslicer and SCIRun. It segments
images from CT and MRI scans and creates a meshwork model that is good for
visualization.
Jolley and
Stinstra’s software uses myocardial voltage gradients
to predict the likelihood of successful defibrillation. According to the critical mass hypothesis,
defibrillation is effective – rendering the heart temporarily inexcitable – if it produces a threshold voltage gradient
in a large fraction of the myocardial mass.
Usually, a gradient of three to five volts per centimeter is needed in
95 % of the heart. Voltage gradients of
over 60 V/cm can damage tissue. When
modeling different defibrillator placements, Jolley
and Stinstra sought to obtain safe voltage gradients
above the defibrillation threshold.
Their model suggests that small changes in defibrillator electrode
position and length have large effects on voltage gradients in the heart.
To test their software, Jolley and Stinstra compared its
predictions to defibrillation metrics obtained in a catheterization lab. They found a good, albeit not perfect,
correlation between the model’s predictions and the data.
Additional challenges must be
overcome if their modeling system is to be used clinically. First, the CT and MRI scans needed to model a
patient’s chest are difficult and slightly dangerous to obtain. These scans are usually taken only of
patients who have cancer or who have suffered serious trauma. From an engineering perspective, the user
interface of the software needs to be improved, and the software must be
adapted to run on PC’s.
Despite
these challenges, the modeling system developed by Jolley
and Stinstra promises to help guide the placement of
implanted defibrillators in children, and it nicely illustrates the uses to
which computer modeling may be put. Click
here to view this presentation