I became interested in things Bayesian when I was contemplating the impact of probabilistic knowledge on medical decisions. As a physicist, I had been introduced to standard Fischer statistics, but had never studied probability and/or statistics formally. As I became more involved in decision making under uncertainty, Bayesian networks seemed to me the obvious choice, and that, as they say, was that.
My interest in medical decision making stemmed from my work in multiobjective optimization, which arose from working on optimization in intensity modulated radiation therapy (IMRT). I got involved in IMRT early in its development but more on the hardware side and measuring dose distributions. Originally I was an atomic physicist who moved to charged particle biophysics, then to clinical heavy charged particle therapy.