Fig. 4

The generic cost function term formulation used throughout the Model Personalization and Treatment Optimization toolsets. The formulation utilizes a maximum allowable error and error center to make the calculated cost physically meaningful. If the difference between the current value and the error center is greater than the maximum allowable error, then the quantity in parentheses will be greater than 1, and squaring this quantity will amplify the cost (e.g., 1.1^2 = 1.21). In contrast, if the difference between the current value and the error center is less than the maximum allowable error, then the quantity in parentheses will be less than 1, and squaring this quantity will attenuate the cost (e.g., 0.9^2 = 0.81). For cost function terms that track time-varying experimental data, the time-varying experimental value is used implicitly as the error center. For cost function terms that minimize errors relative to a static reference value, the specified error center value is used explicitly as the error center