It all started at
Duke when Henry McIntosh gave me a job in 1962 as an engineer in the Cardiac
Cath Lab. Soon after that, Gene Stead, then
Chairman of Medicine, began to push me in directions I’d never been
pushed. Finally, he pushed me out to
finish my educational program at Rice and UNC.
I came back to Duke in 1968 as the first Ph.D. Assistant Professor in
the Department of Medicine. It was an
experiment in Stead’s classical manner:
people chemistry. He was a
genius at mixing folks up in ways that were unconventional, adding heat and
watching the reactions. The results
were usually extraordinary – witness the birth of the Physician’s Assistant
program. Here are some of the view of
Duke: the Chapel, the Old Chemistry
Building where I had my first lab and the Science Research Building
I really had a lab –
first in the old Chemistry building (middle above) and then in the “tin hut” –
a building for Ivan Brown and Wirt Smith to build the first hyperbaric facility
at Duke (in the early 60s). From 1968
until 1980 I worked on building the Duke Cardiology database with Bob Rosati
and a host of fellows (Galen Wagner, Rob Calif, Fred McNeer, etc etc etc). In parallel, we ( Tom Gallie, Merrell
Patrick and Dee Ramm) started the
Department of Computer Science in 1971.
It was an interesting adventure because Bill Anlyan (Dean of the Medical
School) threatened to start the program in the Medical School if the University
remained quiet on this matter. Harold Lewis (Dean of the Graduate School)
had a meeting with Tom, Merrill,
Gene Stead, Bill Anlyan and myself and decided that the time was right
for something new and different. So,
Duke CS was born.
Frank, Tom, Dee and where is Merrill?
In 1980 Bill Anlyan
gave me some space old Chem to begin to
expand our computer science and bioengineering activities. At the same time Joe Greenfield took over
Cardiology and Jim Wyngaarden was running Medicine. To continue, I learned that I needed to have some outlet for
energy – and Ellen (my wife) decided I should start running. Maddy Spach made it possible and he and I
started running together in 1977 – and suddenly I had a 1 hour seminar each day
about discontinuous propagation and a chance to decide how to untangle (during
the afternoon) all that I messed up (in the morning). Between Maddy, Joe, Jim and Gene Stead, I had probably the most
unique group of mentors that one could
imagine. All the research skills I
learned from them and from Jim Grizzle, who was chairman of Biostatistics at
UNC Chapel Hill. When I finally moved
to the old Chem building, I actually built a wet lab for tissue studies, to
complement the facilities for cellular studies that Gus Grant (my long time
friend and collaborator). During this
time we did lots of interesting things.
We acquired UNIX in 1973 – and have a letter signed by Ken Thompson when
we acquired an update in 1978. I was
sitting on the board of the Triangle Universities Computer Center and became
interested in porting UNIX to the 370 and Amdahl machines. Fred Brooks (chairman of UNC computer
science) started a project to do this.
It was too early, though – sigh.
In addition, I had the good fortune to work with friends at Roche in
Basel, and this started my international adventures. It was a very serious virus that infected me – resulting in work
in Russia (USSR), India, Greece, Spain, France, UK, Germany, Nepal and
Egypt. The NIH was gracious to fund
many of these adventures. Below are
photos that capture some of our lab without walls. First, Maddy, Gus (co
architect of the guarded receptor paradigm) and David (Wendt).
Here is Anselmo
(Lastra), Beth (Rusnak) and David (Cherveny) -
these guys did all the UNIX and network magic that made our
international collaborations really happen
We lived in the crazy
tin-hut, designed to optimize people chemistry. As you see, no one took anything seriously. Here is Yvonne (Walker), Ed (Darken), Jo (Smaltz), Margie (Dietz) –
and Marge without a doubt has the most profound effect on everything we
did. Without her imagination and
problem solving – we would still be crawling in the dark
I had two homes: one
in the tin hut and one in the SRB. Here
I found another mentor. Xiaobai (Sun)
was my friendly neighbor in the Computer Science department and my teacher of
numerical methods and linear algebra. When
I had algorithm problems (e.g. how to efficiently solve a problem with a
circulant matrix) Xiaobai stopped what she was doing and led the way. Her thinking was fun to watch. Xiaobai sat in on my seminar-course:
Computing, communication and memory in biological systems. She had no biological background but
suddenly becamse a net contributor to the class. We explored models of excitable cells (cardiac and neuronal) and
demonstrated many interesting problems that are inherent in cellular arrays,
e.g. that the circuitry associated with horizontal cells in the retina computes
an approximate laplacian of the visual field, a neat way to detect motion. She and I discussed teaching strategies
where the role of the teacher and the students was blurred, i.e. sometimes we
were the learners and sometimes others were learners. To test these ideas, Xiaobai and I explored problem-based
learning in her numerical methods course.
Actually, I challenged Xiaobai to come up with a problem whose solution
would touch all the essential areas in numerical methods. She initially thought it was not possible,
but then decided an excitable single-cell would be a good model. With our first group of students, Xiaobai
took the role of the facilitator and I watched from the side lines and was the
cheer leader. The strategy was a
success and now her courses, although tough as nails, are among the most
popular computer science courses at Duke.
Xiaobai and her gentle manner, had a profound influence on the way I thought about problems and
how to understand the nature of different types of errors.
It’s a little known
fact – I really did some experiments.
Here are two students, Frank
(Tong) and Melanie trying to keep me from tripping over the apparatus. We explored vulnerability in isolated rabbit left atrium where we were able
to demonstrated drug-induced prolongation of the vulnerable period which was predicted
by Anselmo’s Ph.D. work (Proarrhythmic response to Na channel blockade: Theoretical model and numerical
experiments. Circ. 84: 1364-1377,
1991.). It was one of those interesting
times where the model and computer studies actually predicted something not
previously observed. We were pleased
and put these results together in a paper for
the American J. Physiology:
Cardiac instability amplified by use-dependent Na channel blockade. Amer. J. Physiol 262:H1305-1310, 1992. As you can see from the expression on
Frank’s face, we were quite lost in the fascinating world of antiarrhythmic
drugs, proarrhythmic responses and reentrant arrhythmias.
Because of our work
in Moscow and Pushchino, we were fortunate to acquire Yuri (Zilberter) and
Josef (Starobin). Yuri’s experimental
skills were unbelievable and he worked through ideas of ion channel blockade
that had escaped us. Josef brought an
unbelievable degree of mathematical sophistication to our group, and worked out
the quasi-analytical characterization of the vulnerable period. His mathematical and computing skills led us
to a new appreciation of the importance of the wavefront in propagation and
also he developed on his own, the interaction between waves and obstacles.
In parallel with the
Duke work were our collaborations: here
is our Egyptian team with David and Marge and to the right is a demonstration
of our indestructible medical record in Kathmandu, Nepal
Here is our Moscow /
French link: Nail (Burnashev) and Guy
(Vassort) and Rene (Clapier) in the
Moscow lab and in Heidelberg
Here is Jorg
(Weirich) in Freiburg and T. Sada, Ban and colleagues in Japan
Finally here are
Valentin (Krinsky), Vicente
(Perez-Munuzuri) and his wife Natalia in Santiago de Compestello Spain. Valentin liked this photo with a potato
digger: the fate of many Pushchino scientists in the post Gorbachev era. I came to know Vicente through Valentin
Krinsky’s lab in Pushchino. Natalia’s
sister, Olga, was Valentin’s secretary. Valentin taught me the power of
identifying generic properties of complex systems, in our case, generic
properties of an excitable cell or an array of excitable cells. Whether BZ chemical media, cardiac cells or
coupled nonlinear differential equations (FHN), the generic properties were
always there. It was a refreshing insight
because it said that adding complexity to a mathematical model in order to
achieve more “realism” may be a distraction.
Rather, study the generic properties of a minimal model and from here,
develop your insights. Example of
generic properties of an excitable cell (or array of cells): threshold of
excitation, spatial threshold of excitation (liminal region) and
vulnerability. None of these properties
disappear when you increase the complexity of the model. Vicente deeply understood this and took our
ideas to his laboratory in Spain. Vicente was a chemical and circuit magician, and transported our
ideas about vulnerability to the excitable BZ chemical medium and arrays of
Choa circuits. He demonstrated
experimentally to us, that vulnerability was truly a generic property in a
series of elegant studies published in Phys Rev.
Although I’ve moved to MUSC, our lab still functions. Gus
and I continue to collaborate. Maddy
and I continue to talk about propagation and when together, we have our usual
noon seminar on the Duke track (or somewhere else). My current work
continues to focus on the ion channel blockade (with Gus), vulnerability and the
math aspects of excitable media with Tassos (Bountis) and two special
Romanian students at the University of Patras (Adi and Laura, and look at our
camping trips).
I continue to search
for new ideas, new insights and new adventures. Stay tuned for more.