Table 2.

Comparison of a FE simulator based on ANSYS [26] and a proprietary FD simulator [17].

Comparison Criteria	FE-Based Simulator		FD-Based Simulator
	Pros	Cons	Pros	Cons
Geometrical representation of complex shapes	Natural conformity of grid points to curved surfaces			Difficult with the more intuitive orthogonal grid system
Simulation of the classical bioheat equation		Requires a special operations to account for the heating effect due to blood perfusion	Similar difficulty to the ordinary conduction problem; blood perfusion increases system stability [15]
Computation cost at each time step		High, requires matrix solvers and high number of operations	Low for an explicit scheme, straightforward calculations
Memory resources		High demand, matrices size is orders of magnitude larger than the number of grid points	Low demand, matrices size and number of grid points are of the same order
Defect region calculations		Very high cost	Low, can be calculated simultaneously
Time step interval for bioheat simulation	Relatively long	Stability does not guaranty convergence nor consistency; have to be evaluated iteratively	Stability is a necessary and sufficient condition for convergence and consistency	Explicit formulation requires very small time steps to ensure stability
Computational Complexity	Linear with numerical grid points		Linear with numerical grid points
Pre-processing & Post-processing		Complex due to grid points registration	Trivial with an orthogonal grid system
Scalability	Parallel kernels possible	Bottlenecks due to matrices operation	Trivial for an explicit scheme, no bottlenecks
Availability of commercial codes for heat transfer simulations	A selection is available			None
User friendliness	High when an FE commercial code is used by a proficient user with technical background	Low when a commercial code is used by a novice user with no technical background	High when a proprietary code is tailored for a novice user from a medical background	Impractical without developing a tailored system