This paper presents an algorithm to calculate gamma dose rates for Virtual Reality (VR) applications, which dynamically adapts the method to cope with both accuracy and time requirements. Given the real-time constraints imposed by VR applications, more accurate, but computationally intensive stochastic algorithms (e.g. Monte Carlo) cannot be used. On the opposite end, a Point Kernel (PK) method can be effective in some cases with as little as one point (Mono PK) to define a source, in contrast with the millions of points that Monte Carlo computes.
Simple Mono PK codes may lack the desired accuracy in some circumstances, requiring a more detailed source representation. In this work a novel method is presented which automatically estimates the appropriate level of detail for the source volumetric representation in order to reduce accuracy error. It then generates a non-regular mesh model and subsequently computes the dose rate via a PK method, performing this 3-step process in real time.