Discovering (predicting) new cardiac physiology/function from cardiac imaging,mathematical modeling and first principles

Sandor Kovacs
Washington University in St. Louis

The information obtained by imaging the heart (echocardiography, MRI, CT etc) in the
clinical setting is virtually always used for clinical diagnosis or for advancement of
imaging technology, rather than to gain further insight into the physiologic rules that
govern cardiac function at the organ system level. Because the four-chambered heart is an
almost perfect constant-volume pump, and because the heart is a mechanical suction-pump
during early diastole - aspects of global and segmental cardiac function can be modeled
and characterized mathematically via conservation principles (volume, energy) and
equations of motion. If the model approximates the physiology with sufficient accuracy
and is invertible -it has practical utility in quantitatively characterizing presence and
severity of dysfunction. Another beneficial consequence of modeling is its potential to
generate new biologic insight and predict 'new' physiology. Several such new insights,
motivated by kinematic modeling and validated by simultaneous cardiac imaging and
physiologic recording in humans will be highlighted. Some 'unsolved problems' in cardiac
physiology will also be discussed.

Audio (MP3 File, Podcast Ready) Presentation (PowerPoint File)

Back to Heart Modeling: Image Acquisition. Segmentation, Modeling and Analysis