Applications of Adaptive Simulated annealing algorithms to Intensity Modulated Radiotherapy(IMRT) planning problems A.Tadesse , [email protected] F. Fondjo, [email protected] J. Snow, [email protected] A. Bucki, [email protected] Langston University, Langston, OK 73050 NSF-EPSCOR RII Track1- Plenary Session, November,17, 2011 Contents The IMRT Principles(problem  statement) Methods  Model 2D-IMRT problem  Preliminary results  Future directions  Questions and discussions  IMRT PRINCIPLES IMRT PRINCIPLES – 3D “Classical” conformation vs Intensity Modulation Conventional Planning vs Inverse planning Methods: Combinatorial Optimization Algorithms. Genetic algorithms  Projection onto convex sets  Bayesian optimization  Maximum entropy  Gradient-based optimizations  Simulated annealing algorithms  Monte-Carlo based simulated annealing  …etc.  Simulated Annealing(SA): Algorithm (Kirkpatrick et al. 1983) Ensemble based simulated annealing(EBSA): (George Ruppeiner et al , 1991) The EBSA hypothesis on IMRT settings The EBSA hypothesis states that given  only thermodynamic information, constant thermodynamic speed cooling schedule (EBSA) is optimal. ( Frost Concepts -- Facts, Conjectures, and Improvements for Simulated Annealing, Peter Salamon, Paolo Sibani, Richard Frost)